Someone paid to have one of the commercial Lisps open sourced?
How much would this cost?
Then someone paid to have the Lisp libraries modernized to CPAN(etc)
standards?
How much would this cost?
Then someone paid to do something like Rails in the foregoing new
modernized OS Lisp?
How muh would this cost?
·········@gmail.com" <········@gmail.com> writes:
> Someone paid to have one of the commercial Lisps open sourced?
> How much would this cost?
Assume the company earns 200k/year for the share holders with its
CL implementation. It would be worth at least 4M.
If the share holders are also programmers working on CL
implementation, you'd have to add their salaries, so it would be worth
at least 9M.
> Then someone paid to have the Lisp libraries modernized to CPAN(etc)
> standards?
> How much would this cost?
Depends on the size of your team. At least 50k/year/person.
Let's say 25 man.year: 1.25M.
> Then someone paid to do something like Rails in the foregoing new
> modernized OS Lisp?
> How muh would this cost?
Depends on the size of your team. At least 50k/year/person. Let's
say 10 man.year, 0.5M.
So total you're speaking of something like 10M at least.
(monetary unit being either USD or EUR or similar).
--
__Pascal Bourguignon__ http://www.informatimago.com/
NOTE: The most fundamental particles in this product are held
together by a "gluing" force about which little is currently known
and whose adhesive power can therefore not be permanently
guaranteed.
On Jun 10, 3:57 am, Pascal Bourguignon <····@informatimago.com> wrote:
> Depends on the size of your team. At least 50k/year/person.
> Let's say 25 man.year: 1.25M.
That's probably a fairly significant underestimate of the cost of
employing a decent programmer for a year. I don't know what
programmer salaries are like, but I know that the daily rate for a
decent SA come to something significantly more than 50k/year (assuming
dollars: even for GBP a reasonable SA should expect that or more in
salary I think, somewhat location-dependent). That's before all the
overheads etc are factored in which would probably be a factor of
1.5-2.
In the UK, a reasonable contract rate for a decent SA might be GBP 300/
day which is (220 days/year) GBP 66,000, or say $130,000 / year.
That is without overheads, so perhaps the total cost might be $200,000/
year.
Let's just in order to get past this detail, agree that a fully loaded
engineer is $300k/year. I don't actually think that that is the
significant factor here because that's a fairly small recurring cost.
The big cost here will be to buy the product lock-stock-and-barrel
from the vendor. THAT will be the most expensive aspect because, of
course, they have hundreds of person-years behind their products, but
we don't have to use THAT total cost, but rather their current cash
flow and a multiplier (that is, ~ their p/e ratio). Of course, I don't
know what their cash flow is, but lets just make a guess and say that
one of these gets, say, $5K/license and that they sell 2000 licenses/
year. So that's 2000 * $5000 = $10,000,000, times a multipler of, say,
5 ~= $50M; that's more like the calculation I'm looking for.
Is that a reasonable way to calculate this?
On Jun 10, 11:25 am, Tim Bradshaw <··········@tfeb.org> wrote:
> On Jun 10, 3:57 am, Pascal Bourguignon <····@informatimago.com> wrote:
>
> > Depends on the size of your team. At least 50k/year/person.
> > Let's say 25 man.year: 1.25M.
>
> That's probably a fairly significant underestimate of the cost of
> employing a decent programmer for a year. I don't know what
> programmer salaries are like, but I know that the daily rate for a
> decent SA come to something significantly more than 50k/year (assuming
> dollars: even for GBP a reasonable SA should expect that or more in
> salary I think, somewhat location-dependent). That's before all the
> overheads etc are factored in which would probably be a factor of
> 1.5-2.
>
> In the UK, a reasonable contract rate for a decent SA might be GBP 300/
> day which is (220 days/year) GBP 66,000, or say $130,000 / year.
> That is without overheads, so perhaps the total cost might be $200,000/
> year.
On Jun 10, 8:15 pm, ·········@gmail.com" <········@gmail.com> wrote:
> Let's just in order to get past this detail, agree that a fully loaded
> engineer is $300k/year. I don't actually think that that is the
> significant factor here because that's a fairly small recurring cost.
That recurring cost *is* what it costs to run the company.
To answer your other question, what you are proposing is essentially a
hostile takeover of a privately-held company. There are standard ways
of doing that kind of valuation.
(I think this idea, which has been proposed before, is just as
implausible as it was last time BTW...)
--tim
> > Let's just in order to get past this detail, agree that a fully loaded
> > engineer is $300k/year. I don't actually think that that is the
> > significant factor here because that's a fairly small recurring cost.
>
> That recurring cost *is* what it costs to run the company.
No, it's only a part of what it costs to run the company. But
regardless, it is small relative to the investment required to buy out
the product, and moreover may be recoupable in, for example, support
contracts (as was mentioned elsewhere in this thread).
> To answer your other question, what you are proposing is essentially a
> hostile takeover of a privately-held company. There are standard ways
> of doing that kind of valuation.
No, it is not. You cannot do a hostile take over of a privately-help
company. All you can do it to try to offer them what they think their
product is worth (or perhaps what the whole company is worth) to sell
the product (or company) to you. And although I believe that there may
be "standard" ways of reading tea leaves for valuation, you apparently
don't intend to share any of your insights on this with us.
> (I think this idea, which has been proposed before, is just as
> implausible as it was last time BTW...)
It would be nice to have a pointer with that claim. But regardless,
the market is in rapid flux, esp. since the exponential rise of Ruby,
and so I believe that the theory is becoming rapidly more plausible.
It is also likely that the previous proposers were simply trying to
talk the commercial vnedors into giving it away, whereas I'm talking
about buying it from them. Since you haven't provided a pointer for
your claim that this was previously discussed, I could be wrong. But
if I'm not, then this should be a very different sort of conversation.
On Jun 11, 6:32 am, ·········@gmail.com" <········@gmail.com> wrote:
> > That recurring cost *is* what it costs to run the company.
>
> No, it's only a part of what it costs to run the company.
Ah yes, I forgot all the heavy plant and so on that software companies
need. Silly of me.
>
> No, it is not. You cannot do a hostile take over of a privately-help
> company.
That was my point.
> It would be nice to have a pointer with that claim. But regardless,
> the market is in rapid flux, esp. since the exponential rise of Ruby,
> and so I believe that the theory is becoming rapidly more plausible.
Let's look at it from the point of view of someone who as money and
might want to invest in such a scheme (me, say). What, actually,
you're saying is that commercially developed lisps are far more
valuable than open-source ones. So you propose to buy a commercial
Lisp house and *make their product open source*. OK, well, thanks,
I'll go and put my money somewhere else, I think.
> Let's look at it from the point of view of someone who as money and
> might want to invest in such a scheme (me, say). What, actually,
> you're saying is that commercially developed lisps are far more
> valuable than open-source ones. So you propose to buy a commercial
> Lisp house and *make their product open source*.
No. What I propose is to build valuable verticals on an excellent open
source platform that has guaranteed support (because we are supporting
it ourselves). It may be that at some point in the future, the
platform support subdivision can be once more spun out as a self-
sustaining company, but that would be far into the future. To be more
precise, I claim that Lisp is uniquely positioned for the semantic web
market, and that a semantic web company built on an excellent Lisp
platform can be highly profitable in semantic web verticals, and that
the possible much smaller profit in Lisp engines is negligible and so
the engine (and some support) can be essentially given away. Moreover,
by giving it away, you garner additional support for the engine
through the open source community. In the long run, if your model is
not to make money by selling, but rather by using, Lisp engines, then
I claim that the optimal strategy is to give away the engine in order
to garner an open source community for it, and make your money where
you can.
·········@gmail.com" <········@gmail.com> writes:
> Someone paid to have one of the commercial Lisps open sourced?
> How much would this cost?
> Then someone paid to have the Lisp libraries modernized to CPAN(etc)
> standards?
> How much would this cost?
> Then someone paid to do something like Rails in the foregoing new
> modernized OS Lisp?
> How muh would this cost?
>
Well, how long is a piece of string....
I don't quite understand the motivation/thoughts underlying this question. What
would be the advantage of paying (probably a lot) to have a commercial lisp
made into open source compared to putting that money into developing/expanding
an existing open source version?
If the objective is just to try and have a rails like app, why not just do that
with one of the existing open source lisps?
Rails has obtained quite a lot of interest and having looked at it, its quite
nice. There are some aspects of it that I don't think quite work, at least not
for the way I like to design and build web apps, but it does feel better to use
than other web development frameworks I've used. In particular, it does show
how easily you can build quite sophisticated web apps without having the
complexities associated with other models, such as Java beans etc, which seem
to have way to much overhead for all but the most complex of applications.
However, I'm not sure there is anything technical stopping someone from
developing an similar (maybe even better) framework using one of the existing
open source or commercial CL implementations. At least I'm not aware of
anything in a commercial implementation which would facilitate this over an
open source implementation.
I think the limitation or reason we don't have such a framework (noting that
some may argue we actually already do given the libraries and projects working
in this area) is more related to not having someone who has both the creative
ability, CL skill and the idea for some web app which fills a perceived need at
just the right time for the market to accept it (i.e. a BaseCamp equivalent).
It may even be a mistake to try and develop a rails competitor at this time.
Anything that tries to do the same sort of role as rails is going to just be
seen as an attempt to jump on the band wagon or ride on the coat tails of ruby
and rails and being done in CL is unlikely to gain any real support or get past
the FUD out there concerning lisp.
If on the other hand, someone came up with a truely original idea which has a
superior implementation that is largely due to being written in CL and it is
supported by some app that fills a real need that isn't being adequatly
addressed already - well then, just maybe.....
--
tcross (at) rapttech dot com dot au
> I don't quite understand the motivation/thoughts underlying this question. What
> would be the advantage of paying (probably a lot) to have a commercial lisp
> made into open source compared to putting that money into developing/expanding
> an existing open source version?
I believe that the cost of brining any of the current open lisps up to
the level of any of the commerical ones would be about the same as
just open sourcing a commerical one, and the latter plan would be a
great deal faster and more reliable.
> If the objective is just to try and have a rails like app, why not just do that
> with one of the existing open source lisps?
That is not the goal. The goal is to make Lisp a viable competitor in
the current programming language market, dominated by C++, Java, and
Python/Ruby. The Rails part of my question is merely a part of the
overall picture. In order to play in this domain there needs to be a
single, free, excellent Lisp with a modern library and several killer
apps; Rails seems to be a killer app that would be easy to copy --
that's the only reason I mentioned Rails.
My theory is that is we bought and open-sourced ACL or LispWorks
Lisps, including buying support for their engineers as gatekeepers,
and building out the libraries, we could get the community to rally
around that one, flush the confusing morass of current open source
versions, and focus on building out from that root. I'm fairly certain
that this path is the right one for Lisp; What I'm not sure of is how
to do it in such manner as the commercial vendor doesn't fold in the
process. That vendor would have to transition to being an application
level provider, and which one survives that will depend upon which of
them has the best set of verticals, and which one takes the leap to
open source their lisp first, because then their verticals will
automatically run on the new lisp, and also they get to own the very
important hearts and minds of the market.
On 10 Jun, 16:51, ·········@gmail.com" <········@gmail.com> wrote:
> That is not the goal. The goal is to make Lisp a viable competitor in
> the current programming language market, dominated by C++, Java, and
> Python/Ruby. The Rails part of my question is merely a part of the
> overall picture. In order to play in this domain there needs to be a
> single, free, excellent Lisp with a modern library and several killer
> apps; Rails seems to be a killer app that would be easy to copy --
> that's the only reason I mentioned Rails.
So what features do you think open source lisp is missing that
prevents it from competing efficiently, but which are present in ACL/
lispworks?
> So what features do you think open source lisp is missing that
> prevents it from competing efficiently, but which are present in ACL/
> lispworks?
A team of brilliant and dedicated engineers who have maintained and
built out the software full time for many years, and who, given the
opportunity, will continue to do so.
·········@gmail.com" <········@gmail.com> writes:
> > So what features do you think open source lisp is missing that
> > prevents it from competing efficiently, but which are present in ACL/
> > lispworks?
>
> A team of brilliant and dedicated engineers who have maintained and
> built out the software full time for many years, and who, given the
> opportunity, will continue to do so.
Just for the record, what opportunity are you offering them?
As you note, they've had years of revenue from selling it. How exactly
would you present the case to them that they should trade an ongoing
source of revenue for a fixed cost and assume they are doing something
financially secure? It sounds a little like asking a farmer how much he'd
want to sell his farm. Unless you were going to finance his retirement,
I don't see him selling. But you used the phrase "will continue to do so",
as if the farmer thinks he'll still be a farmer.
Although it's well-known that I have reservations about open/free
software as a general-purpose market paradigm (although it's also
widely misquoted what those reservations are), I'm not trying to argue
any particular case here so please don't misinterpret my remarks. I'm
not challenging you, at least not at this time. I'm just trying to
understand what is being suggested, before I bother to take a position
on it one way or another.
> ... it's well-known that I have reservations about open/free
> software as a general-purpose market paradigm ...
As do I. In fact, I believe that is is confused and incoherent.
Unfortuntely, it is an unstoppable force in some areas of software
engineering, and so if one wishes to survive, one must figure out how
to use that force to one's advantage.
> > A team of brilliant and dedicated engineers who have maintained and
> > built out the software full time for many years, and who, given the
> > opportunity, will continue to do so.
>
> Just for the record, what opportunity are you offering them?
Just for the record, I haven't offered anyone anything. However, were
I rich enough to make this happen, what I would offer them is to
become a part of a larger, much more dynamic and profitable set of
vertical operations built on top of their (thence open source) Lisp,
with a larger, much more dynamic and profitable set of vertical
visions, where they could share in the profitability of that
operation, and at the same time have many more resources with which to
develop their (thence open source) lisp. Sort of like Google does with
Python.
> As you note, they've had years of revenue from selling it.
Although this seems to be one a downward spiral.
> How exactly
> would you present the case to them that they should trade an ongoing
> source of revenue for a fixed cost and assume they are doing something
> financially secure?
That wasn't quite what I had in mind. As above, I would buy them out,
and then continue to pay them to work on the product as a part of a
larger operation, where the Lisp is open source (but the larger
operation is NOT!)
> It sounds a little like asking a farmer how much he'd
> want to sell his farm. Unless you were going to finance his retirement,
> I don't see him selling.
> But you used the phrase "will continue to do so",
> as if the farmer thinks he'll still be a farmer.
Exactly so! The goal is to buy the farm and continue to pay the farmer
to farm it, and at the same time to share in the potential profits of
a larger, potentially more profitable operation that makes, for
example, egg nog from his eggs (and he can give away the eggs too! --
this isn't quite analogous because unlike eggs, you can both use your
open source software, and give it away too! :-)
On Jun 11, 7:29 am, ·········@gmail.com" <········@gmail.com> wrote:
>
> Exactly so! The goal is to buy the farm and continue to pay the farmer
> to farm it, and at the same time to share in the potential profits of
> a larger, potentially more profitable operation that makes, for
> example, egg nog from his eggs (and he can give away the eggs too! --
> this isn't quite analogous because unlike eggs, you can both use your
> open source software, and give it away too! :-)
Has it occurred to you that the Lisp companies might have considered
this idea? What is their cost to open source their product? Why have
they not done so? I mean, it's not as if this idea is some weird
thing any more: look at Sun, say.
> Has it occurred to you that the Lisp companies might have considered
> this idea?
Yes, this has occurred to me.
> What is their cost to open source their product?
I don't know, they haven't replied to my post yet.
> Why have they not done so?
I don't know, but I doubt that it's because they are making so much on
Lisp licenses that they are all out on yatchs.
> I mean, it's not as if this idea is some weird
> thing any more: look at Sun, say.
Yes, look at Sun, say. It sounds like you're supporting my position.
Sun makes money (if they do) by selling other products -- specifically
computers and consulting services -- not on Java.
On Jun 11, 3:49 pm, ·········@gmail.com" <········@gmail.com> wrote:
> > What is their cost to open source their product?
>
> I don't know, they haven't replied to my post yet.
That was a rhetorical question. If they own all the rights to their
product (which is likely to be true to first order) then their cost is
close to zero. In particular it is enormously less than the cost of
buying the company in order to give away its product.
> > Why have they not done so?
>
> I don't know,
Well, that's a question that you need to answer.
>
> > I mean, it's not as if this idea is some weird
> > thing any more: look at Sun, say.
>
> Yes, look at Sun, say. It sounds like you're supporting my position.
Sigh. No, I'm trying to point out that the Lisp companies are not
living inside a chamber sealed off from all outside contact in 1990
and are therefore probably aware that open sourcing their product is
an option. (I wasn't thinking of Java but Solaris, but Java will do.)
So it behooves you to try and explain why they have not taken this
option ("they are all stupid" is not a very good answer).
I'm kind of worried that you seem to have magnificently missed my
points in both my previous articles: perhaps it's because I'm
English. In any case, I'll leave it for others to continue: as I said
this has been gone over many times before in cll and I doubt there is
anything new to say.
--tim
On Jun 11, 10:06 pm, ·········@gmail.com" <········@gmail.com> wrote:
> > I'm kind of worried that you seem to have magnificently missed my
> > points in both my previous articles...
>
> Not to worry.
Wow. At everything above the level of syntax you don't understand
what I say. How odd.
·········@gmail.com" <········@gmail.com> writes:
> > I mean, it's not as if this idea is some weird
> > thing any more: look at Sun, say.
>
> Yes, look at Sun, say. It sounds like you're supporting my position.
> Sun makes money (if they do) by selling other products -- specifically
> computers and consulting services -- not on Java.
Only companies make money on other things. Individuals can't
multitask and break even. So the scheme you describe trades
individual autonomy away from one person for market share in a
different market for others. It's possible to argue there's more
money in it your way, though I think it's not as much of a slam dunk
to do so as some allege, but that's not my point. The money made
other ways goes to other people.
Although they both involve software, some of us don't consider
"design" the same job as "consulting". One is about personal freedom
and expression, the other about doing someone else's bidding and
making enough money to pay the rent. Once in a while, one consults
for someone who wants something freely expressed. But I wouldn't say
that's the norm.
When talking about Lisp companies, I'll wager you're talking the dream
job of the people working there. Doing consulting from those
companies probably pays the bills, but is probably not what they are
there for. And so offering to let them give up the design part, the
one little piece of autonomy they have, in exchange for more ability
to "just pay the bills" sounds like a bizarre thing to think they'd
leap at.
Tim Bradshaw wrote:
> On Jun 11, 7:29 am, ·········@gmail.com" <········@gmail.com> wrote:
>
>> Exactly so! The goal is to buy the farm and continue to pay the farmer
>> to farm it, and at the same time to share in the potential profits of
>> a larger, potentially more profitable operation that makes, for
>> example, egg nog from his eggs (and he can give away the eggs too! --
>> this isn't quite analogous because unlike eggs, you can both use your
>> open source software, and give it away too! :-)
>
> Has it occurred to you that the Lisp companies might have considered
> this idea? What is their cost to open source their product? Why have
> they not done so? I mean, it's not as if this idea is some weird
> thing any more: look at Sun, say.
My $0.02 on the subject. It seems obvious to me that we are witnessing
the rise in popularity of dynamic languages (due to various reasons).
Stimulated by the hype, more and more people/businesses are looking for
a substitute for C++/Java/C#/etc. Once they find their new pet
language/platform, they will most likely stick with it for the next 5 or
more years. In my opinion, these businesses are mostly small businesses
and I believe current commercial CL prices represent a crucial factor in
discarding CL altogether and going with Python/Ruby/whatever.
My point is: maybe commercial CL vendors should reconsider their
strategy and try to use the hype to gain volume. Once the hype is over
it might be very hard to make living by selling CL compilers. Of
course, vendors would most likely need to broaden their portfolio and
sell other stuff besides development environments.
I think Sun got this right. OTOH, they have enough money to afford
longer term investments.
Josip Gracin wrote:
> Tim Bradshaw wrote:
>> On Jun 11, 7:29 am, ·········@gmail.com" <········@gmail.com> wrote:
>>
>>> Exactly so! The goal is to buy the farm and continue to pay the farmer
>>> to farm it, and at the same time to share in the potential profits of
>>> a larger, potentially more profitable operation that makes, for
>>> example, egg nog from his eggs (and he can give away the eggs too! --
>>> this isn't quite analogous because unlike eggs, you can both use your
>>> open source software, and give it away too! :-)
>>
>> Has it occurred to you that the Lisp companies might have considered
>> this idea? What is their cost to open source their product? Why have
>> they not done so? I mean, it's not as if this idea is some weird
>> thing any more: look at Sun, say.
>
> My $0.02 on the subject. It seems obvious to me that we are witnessing
> the rise in popularity of dynamic languages (due to various reasons).
> Stimulated by the hype, more and more people/businesses are looking for
> a substitute for C++/Java/C#/etc. Once they find their new pet
> language/platform, they will most likely stick with it for the next 5 or
> more years. In my opinion, these businesses are mostly small businesses
> and I believe current commercial CL prices represent a crucial factor in
> discarding CL altogether and going with Python/Ruby/whatever.
Whether a language is dynamic or not is just one of many possible
features in the laundry list for a programming language. Whether a
language makes you more productive or not depends more on how well the
various features are integrated, not whether and which features are
available in the first place.
Whether a language is popular and whether it is available for free or
not are two of the most superficial characteristics of a programming
language. If a company bases its decisions on such superficialities,
it's probably not in a very good shape. Companies have to spend so much
money an assets anyway, so the comparatively small costs for software
cannot possibly be a serious stumbling block.
> I think Sun got this right. OTOH, they have enough money to afford
> longer term investments.
No, Sun just got lucky at one particular point in time. Java became such
a success because of its bundling with Netscape Navigator and the
implied promise of a new deployment model for applications (which never
took off).
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
> Whether a language is popular and whether it is available for free or
> not are two of the most superficial characteristics of a programming
> language. If a company bases its decisions on such superficialities,
> it's probably not in a very good shape. Companies have to spend so much
> money an assets anyway, so the comparatively small costs for software
> cannot possibly be a serious stumbling block.
Unfortunately, this is not the reality that we live in. The open
source movement, and various other factors, has/have convinced nearly
everyone that software should be free -- and unfortunately, they mean
free like beer, not free like freedom, which is exactly the opposite
of its initial intent...but let's let that pass; I really don't want
to start the same fight about open source again!
I have worked at many places, including the most famous and wealthy
university on earth, the most famous (and previously wealthy) computer
science research lab on earth, and several very wealthy places in
between, and I can tell you for absolute fact that every time I have
asked for the money to buy CL licences, or put these into my grant
proposals, the response has been a line item veto, or at least a
serious battle. And these are people who will drop $50,000 on a Mass
Spec that they use rarely, or drop the entire license fee to send a
team on cross-country travel for a week to DC to chat up a potential
client with near zero expected ROI. If you really believe that you can
survive by charging anything with 4 digits before the decimal point
for what looks to them like Ruby or Python, or even Java, only worse
in most *apparent* ways, then you're living in another reality. I hate
the fact that this is how things are, but it's how things are.
From: Dimiter "malkia" Stanev
Subject: Re: What would happen if...
Date:
Message-ID: <466EC1B7.7080809@mac.com>
Allegro Student Edition costs $99
http://www.franz.com/downloads/student.lhtml
From the page:
Requirements and Restrictions
1 Available to undergraduate students enrolled full time at a degree
granting university.
2 Individuals interested in purchasing will be required to validate
their student and enrollment status prior to purchase.
3 Full version of Allegro CL 8.0 with no feature restrictions. Upon
purchase of the Student Version the product will function for a period
of 2 years.
4 Not for use in a commercial setting or in university sanctioned
research projects and applications.
I guess what you've described, is case 4, so it might've not worked for
you back then...
Lispworks also has a Academic Version, which restricts you from
delivering applications/DLL's and distributing those applications free
of charge.
http://www.lispworks.com/products/features.html
Prices for the Academic are higher than Allegro's Student, but still
under 4 figures:
http://www.lispworks.com/buy/prices-1a.html
········@gmail.com wrote:
>> Whether a language is popular and whether it is available for free or
>> not are two of the most superficial characteristics of a programming
>> language. If a company bases its decisions on such superficialities,
>> it's probably not in a very good shape. Companies have to spend so much
>> money an assets anyway, so the comparatively small costs for software
>> cannot possibly be a serious stumbling block.
>
> Unfortunately, this is not the reality that we live in. The open
> source movement, and various other factors, has/have convinced nearly
> everyone that software should be free -- and unfortunately, they mean
> free like beer, not free like freedom, which is exactly the opposite
> of its initial intent...but let's let that pass; I really don't want
> to start the same fight about open source again!
>
> I have worked at many places, including the most famous and wealthy
> university on earth, the most famous (and previously wealthy) computer
> science research lab on earth, and several very wealthy places in
> between, and I can tell you for absolute fact that every time I have
> asked for the money to buy CL licences, or put these into my grant
> proposals, the response has been a line item veto, or at least a
> serious battle. And these are people who will drop $50,000 on a Mass
> Spec that they use rarely, or drop the entire license fee to send a
> team on cross-country travel for a week to DC to chat up a potential
> client with near zero expected ROI. If you really believe that you can
> survive by charging anything with 4 digits before the decimal point
> for what looks to them like Ruby or Python, or even Java, only worse
> in most *apparent* ways, then you're living in another reality. I hate
> the fact that this is how things are, but it's how things are.
>
These "student" editions are not useful for real work; in addition to
help and/or time limits and licensing limitations (as you point out,
e.g., case 4!), they typically are not the "enterprise" editions, so
important connectivity and web development tools are missing.
(Although all of this changes variously, so my info on this could be
dated, but not by far; I fairly deeply explored this a couple of years
ago.)
On Jun 12, 8:54 am, "Dimiter \"malkia\" Stanev" <······@mac.com>
wrote:
> Allegro Student Edition costs $99
>
> http://www.franz.com/downloads/student.lhtml
>
> From the page:
>
> Requirements and Restrictions
> 1 Available to undergraduate students enrolled full time at a degree
> granting university.
> 2 Individuals interested in purchasing will be required to validate
> their student and enrollment status prior to purchase.
> 3 Full version of Allegro CL 8.0 with no feature restrictions. Upon
> purchase of the Student Version the product will function for a period
> of 2 years.
> 4 Not for use in a commercial setting or in university sanctioned
> research projects and applications.
>
> I guess what you've described, is case 4, so it might've not worked for
> you back then...
>
> Lispworks also has a Academic Version, which restricts you from
> delivering applications/DLL's and distributing those applications free
> of charge.
>
> http://www.lispworks.com/products/features.html
>
> Prices for the Academic are higher than Allegro's Student, but still
> under 4 figures:
>
> http://www.lispworks.com/buy/prices-1a.html
>
>
>
> ········@gmail.com wrote:
> >> Whether a language is popular and whether it is available for free or
> >> not are two of the most superficial characteristics of a programming
> >> language. If a company bases its decisions on such superficialities,
> >> it's probably not in a very good shape. Companies have to spend so much
> >> money an assets anyway, so the comparatively small costs for software
> >> cannot possibly be a serious stumbling block.
>
> > Unfortunately, this is not the reality that we live in. The open
> > source movement, and various other factors, has/have convinced nearly
> > everyone that software should be free -- and unfortunately, they mean
> > free like beer, not free like freedom, which is exactly the opposite
> > of its initial intent...but let's let that pass; I really don't want
> > to start the same fight about open source again!
>
> > I have worked at many places, including the most famous and wealthy
> > university on earth, the most famous (and previously wealthy) computer
> > science research lab on earth, and several very wealthy places in
> > between, and I can tell you for absolute fact that every time I have
> > asked for the money to buy CL licences, or put these into my grant
> > proposals, the response has been a line item veto, or at least a
> > serious battle. And these are people who will drop $50,000 on a Mass
> > Spec that they use rarely, or drop the entire license fee to send a
> > team on cross-country travel for a week to DC to chat up a potential
> > client with near zero expected ROI. If you really believe that you can
> > survive by charging anything with 4 digits before the decimal point
> > for what looks to them like Ruby or Python, or even Java, only worse
> > in most *apparent* ways, then you're living in another reality. I hate
> > the fact that this is how things are, but it's how things are.- Hide quoted text -
>
> - Show quoted text -
Some entity, AKA ·········@gmail.com" <········@gmail.com>,
wrote this mindboggling stuff:
(selectively-snipped-or-not-p)
> These "student" editions are not useful for real work; in addition to
> help and/or time limits and licensing limitations (as you point out,
Gosh, is that why they're called student-edition?
is it really true , that they really are meant for study ? ...
Cor
--
(defvar MyComputer '((OS . "GNU/Emacs") (IPL . "GNU/Linux")))
The biggest problem LISP has, is that it does not appeal to dumb people
If that fails to satisfy you, start reading the HyperSpec or woman frig
mailpolicy @ http://www.clsnet.nl/mail.php
·········@gmail.com" <········@gmail.com> writes:
> Unfortunately, this is not the reality that we live in. The open
> source movement, and various other factors, has/have convinced nearly
> everyone that software should be free -- and unfortunately, they mean
Really? Nearly everyone? That's quite a strong statement.
> I have worked at many places, including the most famous and wealthy
> university on earth, the most famous (and previously wealthy) computer
> science research lab on earth, and several very wealthy places in
> between, and I can tell you for absolute fact that every time I have
> asked for the money to buy CL licences, or put these into my grant
> proposals, the response has been a line item veto, or at least a
> serious battle. And these are people who will drop $50,000 on a Mass
So this "most famous and wealthy university on earth" didn't have any
commercial software? How did they manage without Mathematica, Gauss,
Matlab, Stata and other similar programs? You must share the secret,
I am sure that less famous and wealthy universities would like to save
money on licenses too.
> Spec that they use rarely, or drop the entire license fee to send a
> team on cross-country travel for a week to DC to chat up a potential
> client with near zero expected ROI. If you really believe that you can
So are you trying to imply that these people are stupid? (that's what
I gather from your description -- spending time and money on a near
zero ROI project qualifies as stupid). Hope you have better luck next
time and find a company with sane people.
Best,
Tamas
On Jun 12, 3:54 pm, Tamas Papp <······@gmail.com> wrote:
> So this "most famous and wealthy university on earth" didn't have any
> commercial software? How did they manage without Mathematica, Gauss,
> Matlab, Stata and other similar programs? You must share the secret,
> I am sure that less famous and wealthy universities would like to save
> money on licenses too.
Not to mention the Oracle system that ran their payroll &c &c.
> So are you trying to imply that these people are stupid? (that's what
> I gather from your description -- spending time and money on a near
> zero ROI project qualifies as stupid). Hope you have better luck next
> time and find a company with sane people.
Always remember the the correct fix for stupid decisions is never
educating people so they can make smart ones. Instead do everything
you can to support the stupidity.
> So this "most famous and wealthy university on earth" didn't
> have any commercial software? How did they manage without
> Mathematica, Gauss, Matlab, Stata and other similar
> programs? You must share the secret, I am sure that less
> famous and wealthy universities would like to save
> money on licenses too.
First off, I didn't say that they didn't do it; there was just often a
serious battle. Also, on the other hand, you're correct that I was not
specific enough (which mislead others as well). Let me rephrase: It
has been my experience more often than not that decision makers who
normally would not balk at dropping a few thousand dollars for
hardware or travel, will often reject (or fight) paying for slightly
expensive programming language software based, I think, upon, among
other things, a series of misunderstandings of what "open source"
means, which are not improved by the fact that there are nearly
equivalent (in these decision makers' limited knowledge) programming
languages out there that (a) "everyone else it using", and (b) are
demonstrably free. Now, I'll admit that my experience here is limited
to Lisp, so I can't speak for how these decision makers decide to buy
Oracle or Mass Specs or send folks on trips for thousands of dollars
to accomplish thing that could mostly be done by email. Obviously they
do sometimes pay for software, but mostly these are incompariable
cases because MatLab, Mathematica, etc. are (for the most part)
unique, and do not have well respected freeware similar tools. Another
case that I do know of, R, which is basically a freeware SPlus has
essentially deleted the market for SPlus.
>>>>> "JS" == ········@gmail com <········@gmail.com> writes:
JS> It has been my experience more often than not that decision
JS> makers who normally would not balk at dropping a few thousand
JS> dollars for hardware or travel, will often reject (or fight)
JS> paying for slightly expensive programming language software
I've fought this fight before.
The principal difference is that the decision maker understands the
difference that paying a few thousand dollars for hardware or travel
makes. The decision maker knows that there are very good languages
and language implementations available for free, and so will balk at
spending any money for them. It's your job to articulate *why*
the commercial implementation is worth spending the extra money for.
"Because you spent $2000 for Joe and Bob to go to that conference in
Washington, DC" does not have any bearing on the merits of a
commercial software package. Neither does "Because you spent $5000
for the server that's currently not doing anything."
And, frankly, if you can't articulate to a decision maker why the
software package is worth the money, it's a reasonable conclusion that
you don't really need it.
Charlton
--
Charlton Wilbur
·······@chromatico.net
········@gmail.com wrote:
> I hate the fact that this is how things are, but it's how things are.
I don't think it's a good idea to make things even more how they appear
to be, especially when you hate their very appearance.
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
Pascal Costanza <··@p-cos.net> writes:
> ········@gmail.com wrote:
>
>> I hate the fact that this is how things are, but it's how things are.
>
> I don't think it's a good idea to make things even more how they appear to be,
> especially when you hate their very appearance.
>
Things are more like they are now than they ever have been before!
--
tcross (at) rapttech dot com dot au
On Jun 12, 10:48 am, Pascal Costanza <····@p-cos.net> wrote:
>
> > I think Sun got this right. OTOH, they have enough money to afford
> > longer term investments.
>
> No, Sun just got lucky at one particular point in time. Java became such
> a success because of its bundling with Netscape Navigator and the
> implied promise of a new deployment model for applications (which never
> took off).
It's fairly important to remember that if any of the Lisp companies
had been through what Sun have they would probably be long, long
dead. Sun are doing better now, but they've been through a very long
run of losing money and their share price is a somewhat over 5% of
what it was at the peak. Sun really didn't get lucky (other than in
the sense that they had enough money in the bank to avoid some awful
takeover).
However, I think the other important thing is that Sun are a *systems*
company and have always been that. I think that means (among other
things) that their value to customers is not really some bit of
software, but in the whole bundle: from well-engineered, supportable
and scalable hardware (which is very profitable at the high end) to
service to software. Sun are *nothing like* a software company, and
that's why it makes sense to them to open source Solaris say.
--tim
(Disclaimer: Sun pay some of my wages at present).
Pascal Costanza wrote:
> No, Sun just got lucky at one particular point in time. Java became such
> a success because of its bundling with Netscape Navigator and the
> implied promise of a new deployment model for applications (which never
> took off).
I was actually talking about their current strategy to open source
everything... the way they are attracting people to Solaris in order to
make money on other things.
On Jun 12, 11:48 am, Pascal Costanza <····@p-cos.net> wrote:
> Whether a language is popular and whether it is available for free
> or not are two of the most superficial characteristics of a
> programming language. If a company bases its decisions on such
> superficialities, it's probably not in a very good shape. Companies
> have to spend so much money an assets anyway, so the comparatively
> small costs for software cannot possibly be a serious stumbling
> block.
I just asked someone, and was told (from imprecise memory) that Franz
wanted both:
* over 10k EUR per developer * platform * version
* over 5% revenue of any product using Franz's lisp
Not to single out proprietary Lisps; obviously our societies are
biased in favor of wealth, and sufficiently sane plans are always a
struggle. (Technically speaking, in this discussion we observe the
free rider problem of market economies.)
There are alternatives to private funding. For example, Lisp obviously
did well from government subsidy, during its most innovative periods.
Some countries offer significant subsidies for employers, like I'm
told the Netherlands does. (Or at least parts of it.)
Also, I would pledge significantly to a Lisp organization which was
run reasonably democratically and transparently, if I thought it had a
chance of effectively allocating resources to free software
production.
Tayssir
On Jun 14, 12:31 am, Tayssir John Gabbour
<············@googlemail.com> wrote:
> On Jun 12, 11:48 am, Pascal Costanza <····@p-cos.net> wrote:
>
> > Whether a language is popular and whether it is available for free
> > or not are two of the most superficial characteristics of a
> > programming language. If a company bases its decisions on such
> > superficialities, it's probably not in a very good shape. Companies
> > have to spend so much money an assets anyway, so the comparatively
> > small costs for software cannot possibly be a serious stumbling
> > block.
>
> I just asked someone, and was told (from imprecise memory) that Franz
> wanted both:
> * over 10k EUR per developer * platform * version
> * over 5% revenue of any product using Franz's lisp
1. I bet they negotiate, and imprecise memory is not the best source
of information.
2. There are other Lisp vendors and at least one of them has very
straightforward terms (no runtime fees).
On Jun 14, 8:42 am, Tim Bradshaw <··········@tfeb.org> wrote:
> On Jun 14, 12:31 am, Tayssir John Gabbour
> > I just asked someone, and was told (from imprecise memory) that Franz
> > wanted both:
> > * over 10k EUR per developer * platform * version
> > * over 5% revenue of any product using Franz's lisp
>
> 1. I bet they negotiate, and imprecise memory is not the best source
> of information.
This was after negotiation. And I conservatively lowered the numbers
-- the numbers I was quoted were HIGHER than 10k and 5%. And I've
heard similar anecdotes about the over-5% revenue elsewhere.
In other forums, Franz representatives absolutely refuse to estimate
their pricing, arguing that you should contact them privately. Despite
repeated public questioning. So Franz is not a source of public
information here.
That said, I also spoke in person with a Franz salesman, and discussed
the issue with them on a general level. (We didn't talk about price
specifics, but rather the broad issue.) I won't go into specifics
about our conversation. But the reasoning, as I understand it, is the
most obvious one: differential pricing.
If anyone has better stats, I welcome them to offer it. I've observed
CEOs talk about this pricing privately, and they just found it weird.
Personally though, I can imagine that paying this cost is justifiable,
depending on your situation. So for example, I expect most people get
pretty low wages. (If they're even on a wage system.) Being in a
situation where you can afford to pay Franz's price isn't such a bad
situation after all, all things considered.
Tayssir
On Jun 14, 9:29 am, Tayssir John Gabbour <············@googlemail.com>
wrote:
>
> This was after negotiation. And I conservatively lowered the numbers
> -- the numbers I was quoted were HIGHER than 10k and 5%. And I've
> heard similar anecdotes about the over-5% revenue elsewhere.
Remember, you're a Lisp person. That means you probably have the
negotiating skills of a bat (as have I before you think I'm being
nasty). My guess is you were likely negotiating from a position of
weakness such as being a small company or a university or something.
>
> If anyone has better stats, I welcome them to offer it. I've observed
> CEOs talk about this pricing privately, and they just found it weird.
>
It's not that uncommon a model in the non-toy software market.
--tim
On Jun 14, 7:44 pm, Tim Bradshaw <··········@tfeb.org> wrote:
> On Jun 14, 9:29 am, Tayssir John Gabbour <············@googlemail.com>
> wrote:
> > This was after negotiation. And I conservatively lowered the numbers
> > -- the numbers I was quoted were HIGHER than 10k and 5%. And I've
> > heard similar anecdotes about the over-5% revenue elsewhere.
>
> Remember, you're a Lisp person. That means you probably have the
> negotiating skills of a bat (as have I before you think I'm being
> nasty). My guess is you were likely negotiating from a position of
> weakness such as being a small company or a university or something.
I didn't negotiate; the costs were told to me by others. But I think
you're right, other than "getting to yes" by paying attention to
peoples' interests, I don't know that much about negotiation.
I doubt Franz plays hardball though. If you can offer them a deal that
makes sense to them, why should they reject it? Taking a chunk of your
revenue means they'd rather you make more revenue.
> > If anyone has better stats, I welcome them to offer it. I've observed
> > CEOs talk about this pricing privately, and they just found it weird.
>
> It's not that uncommon a model in the non-toy software market.
The last company I worked with had kind of similar pricing for one of
their products, actually. Somewhat ironically, they rejected Franz in
favor of Lispworks, partly because of the revenue thing... but also
because of Lispworks' CAPI.
But I suppose most of the people reading this can only afford "toy
software," and are presumably like children.
Tayssir
Tim Bradshaw <··········@tfeb.org> writes:
> On Jun 14, 9:29 am, Tayssir John Gabbour <············@googlemail.com>
> wrote:
>
>>
>> This was after negotiation. And I conservatively lowered the numbers
>> -- the numbers I was quoted were HIGHER than 10k and 5%. And I've
>> heard similar anecdotes about the over-5% revenue elsewhere.
>
> Remember, you're a Lisp person. That means you probably have the
> negotiating skills of a bat (as have I before you think I'm being
> nasty). My guess is you were likely negotiating from a position of
> weakness such as being a small company or a university or something.
>
>>
>> If anyone has better stats, I welcome them to offer it. I've observed
>> CEOs talk about this pricing privately, and they just found it weird.
>>
>
> It's not that uncommon a model in the non-toy software market.
>
My experience is that it is very common in the software market once you move
away from mass produced personal software. The last software package I had to
negotiate a purchase for at work started out at $250,000+ for the software and
$70,000 licensing/support per year. After considerable negotiation, we got what
we wanted for less than $120,000 for the package and $10,000 for licensing and
support per year. I've since found out through talking to others in the sector
who have also purchased this software that while the price I was able to get
wasn't the cheapest in the sector, it was one of the lowest and some other
businesses had payed way more. The cheap price I got wasn't mainly due to my
negotiation skills - it was mainly due to the fact the vendor wanted to get a
foothold in the sector and therefore was willing to drop their price
accordingly.
Tim
On Sun, 10 Jun 2007 23:29:12 -0700, ·········@gmail.com" <········@gmail.com> wrote:
>> As you note, they've had years of revenue from selling it.
>
> Although this seems to be one a downward spiral.
Really? Do you have any evidence for this?
--
Lisp is not dead, it just smells funny.
Real email: (replace (subseq ·········@agharta.de" 5) "edi")
> On Sun, 10 Jun 2007 23:29:12 -0700, ·········@gmail.com" <········@gmail.com> wrote:
> >> As you note, they've had years of revenue from selling it.
>
> > Although this seems to be one a downward spiral.
>
> Really? Do you have any evidence for this?
No. I have the fact that Lisps are far more expensive than pretty much
any other programming languages, and are getting more, not less
expensive. This suggests, although I suppose does not prove my theory.
In article <························@h2g2000hsg.googlegroups.com>,
·········@gmail.com" <········@gmail.com> wrote:
> > On Sun, 10 Jun 2007 23:29:12 -0700, ·········@gmail.com"
> > <········@gmail.com> wrote:
> > >> As you note, they've had years of revenue from selling it.
> >
> > > Although this seems to be one a downward spiral.
> >
> > Really? Do you have any evidence for this?
>
> No.
But that doesn't necessarily mean it isn't true. Generally companies
whose revenues are in a downward spiral do everything they can to
prevent that fact from being known because it's much harder to raise
capital when people know that you're desperate.
rg
In article <························@h2g2000hsg.googlegroups.com>,
·········@gmail.com" <········@gmail.com> wrote:
> > On Sun, 10 Jun 2007 23:29:12 -0700, ·········@gmail.com" <········@gmail.com> wrote:
> > >> As you note, they've had years of revenue from selling it.
> >
> > > Although this seems to be one a downward spiral.
> >
> > Really? Do you have any evidence for this?
>
> No. I have the fact that Lisps are far more expensive than pretty much
> any other programming languages, and are getting more, not less
> expensive. This suggests, although I suppose does not prove my theory.
Actually this is totally wrong.
Nobody escapes the costs. Especially not software companies.
The company has some income:
* licenses
* trainings
* consulting
* applications
* libraries
It has expenses:
* marketing & sales
* staff
* building & equipment, infrastructure
* tax, insurance, ...
* external costs (writers, web designers, ...)
This is absolutely independent if you are selling Lisp or any
other programming languages. The income could be divided
differently. The licenses could be more or less in number and
costs.
If you have a software company with, say, 20 people
it is is independent if you do Lisp or something else.
If the market has a structure and size to a Lisp business
you'll get automatically similar costs for the license.
Don't believe?
Just check the prices for comparable systems (commercial
language implementations with companies behind them).
* Ada compilers
* Java application servers (BEA Weblogic, IBM Websphere, ...)
* Prolog
* Smalltalk (Visualworks, ...)
* Java (Borland, ...)
* Rule systems (Jrules, ...)
You will find that these are as expensive as commercial Lisp and more.
--
http://lispm.dyndns.org
········@gmail.com wrote:
> Just for the record, I haven't offered anyone anything. However, were
> I rich enough to make this happen, what I would offer them is to
> become a part of a larger, much more dynamic and profitable set of
> vertical operations built on top of their (thence open source) Lisp,
> with a larger, much more dynamic and profitable set of vertical
> visions, where they could share in the profitability of that
> operation, and at the same time have many more resources with which to
> develop their (thence open source) lisp. Sort of like Google does with
> Python.
In other words, they would go from being a profit centre to an overhead,
and likely first for the chop should the "set of vertical operations"
turn out to be less profitable than expected.
Not a move I'd /necessarily/ welcome with open arms, were I in their
position.
-dan
>>>>> "JS" == ········@gmail com <········@gmail.com> writes:
>> So what features do you think open source lisp is missing that
>> prevents it from competing efficiently, but which are present
>> in ACL/ lispworks?
JS> A team of brilliant and dedicated engineers who have
JS> maintained and built out the software full time for many
JS> years, and who, given the opportunity, will continue to do so.
I'd concur with this, but I'd wonder where you're going to get the
money to pay the team of brilliant and dedicated engineers what
they're worth.
If you take a look at the languages you're comparing Lisp to here --
Python and Ruby -- you'll see that both of those languages began as
all-volunteer efforts. If you take a look at Perl, another language
taking a similar development path, you'll see that most Perl
development work is done as the result of very small ($500-$1000)
grants or as the side effect of the developer's full-time job. The
developers of PHP aren't paid to work on PHP, but to sell proprietary
addons to PHP that make it faster.
So if you want to successfully buy and open-source a commercial Lisp,
and actively retain the engineers, you need to figure out how to pay
them the same salary to do the same thing without any revenues from
sales of licenses. Otherwise, you lose the principal advantage of
commercial Lisps, which is that engineering and support team. And
you'll lose the experienced engineers to the other Lisp vendor, who is
now benefitting from an influx Lisp users who are not too keen on open
source and who want support contracts with actual engineers behind
them and can now afford to hire more engineers and pay them in
currency rather than good feelings.
Charlton
--
Charlton Wilbur
·······@chromatico.net
In article <··············@mithril.chromatico.net>,
Charlton Wilbur <·······@chromatico.net> wrote:
> >>>>> "JS" == ········@gmail com <········@gmail.com> writes:
>
> >> So what features do you think open source lisp is missing that
> >> prevents it from competing efficiently, but which are present
> >> in ACL/ lispworks?
>
> JS> A team of brilliant and dedicated engineers who have
> JS> maintained and built out the software full time for many
> JS> years, and who, given the opportunity, will continue to do so.
>
> I'd concur with this, but I'd wonder where you're going to get the
> money to pay the team of brilliant and dedicated engineers what
> they're worth.
>
> If you take a look at the languages you're comparing Lisp to here --
> Python and Ruby -- you'll see that both of those languages began as
> all-volunteer efforts. If you take a look at Perl, another language
> taking a similar development path, you'll see that most Perl
> development work is done as the result of very small ($500-$1000)
> grants or as the side effect of the developer's full-time job. The
> developers of PHP aren't paid to work on PHP, but to sell proprietary
> addons to PHP that make it faster.
How about MySQL?
They GPL one version and sell also a version with a
different license.
>
> So if you want to successfully buy and open-source a commercial Lisp,
> and actively retain the engineers, you need to figure out how to pay
> them the same salary to do the same thing without any revenues from
> sales of licenses. Otherwise, you lose the principal advantage of
> commercial Lisps, which is that engineering and support team. And
> you'll lose the experienced engineers to the other Lisp vendor, who is
> now benefitting from an influx Lisp users who are not too keen on open
> source and who want support contracts with actual engineers behind
> them and can now afford to hire more engineers and pay them in
> currency rather than good feelings.
>
> Charlton
--
http://lispm.dyndns.org
On Jun 11, 7:02 am, Rainer Joswig <······@lisp.de> wrote:
> In article <··············@mithril.chromatico.net>,
> Charlton Wilbur <·······@chromatico.net> wrote:
>
>
>
> > >>>>> "JS" == ········@gmail com <········@gmail.com> writes:
>
> > >> So what features do you think open source lisp is missing that
> > >> prevents it from competing efficiently, but which are present
> > >> in ACL/ lispworks?
>
> > JS> A team of brilliant and dedicated engineers who have
> > JS> maintained and built out the software full time for many
> > JS> years, and who, given the opportunity, will continue to do so.
>
> > I'd concur with this, but I'd wonder where you're going to get the
> > money to pay the team of brilliant and dedicated engineers what
> > they're worth.
>
> > If you take a look at the languages you're comparing Lisp to here --
> > Python and Ruby -- you'll see that both of those languages began as
> > all-volunteer efforts. If you take a look at Perl, another language
> > taking a similar development path, you'll see that most Perl
> > development work is done as the result of very small ($500-$1000)
> > grants or as the side effect of the developer's full-time job. The
> > developers of PHP aren't paid to work on PHP, but to sell proprietary
> > addons to PHP that make it faster.
>
> How about MySQL?
> They GPL one version and sell also a version with a
> different license.
>
>
>
> > So if you want to successfully buy and open-source a commercial Lisp,
> > and actively retain the engineers, you need to figure out how to pay
> > them the same salary to do the same thing without any revenues from
> > sales of licenses. Otherwise, you lose the principal advantage of
> > commercial Lisps, which is that engineering and support team. And
> > you'll lose the experienced engineers to the other Lisp vendor, who is
> > now benefitting from an influx Lisp users who are not too keen on open
> > source and who want support contracts with actual engineers behind
> > them and can now afford to hire more engineers and pay them in
> > currency rather than good feelings.
>
> > Charlton
>
> --http://lispm.dyndns.org
Another model is the one employed by Cincom Smalltalk. It's not open
source. You can download the entire thing and you only have to pay
when you start using it commercially. I'm not advocating this for
Franz or LispWorks. I'm sure they've weighed the pros and cons of
various models.
On 11 Jun, 06:42, ·········@gmail.com" <········@gmail.com> wrote:
> > So what features do you think open source lisp is missing that
> > prevents it from competing efficiently, but which are present in ACL/
> > lispworks?
>
> A team of brilliant and dedicated engineers who have maintained and
> built out the software full time for many years, and who, given the
> opportunity, will continue to do so.
Well that would be nice, but do ruby or python have that? How far
behind is a cross platform open source lisp like clisp from matching
the core functionality of the ruby/python interpreters, anyway?
Just remembering that recently (a few years ago) Eclipse CL stopped
beeing sold by its vendor. It was a cross platform CL to C compiler
and you could even license the source.
Maybe this ex-comercial? CL can be bought quite cheaply. However I
don't know much about the implementation, especially compared to ECL.
Chris Russell wrote:
> Well that would be nice, but do ruby or python have that? How far
> behind is a cross platform open source lisp like clisp from matching
> the core functionality of the ruby/python interpreters, anyway?
Infinitely far behind because it's licensed with the GPL and turns all
software you make with it into GPLware automatically unless you restrict
your program to the ANSI CL subset (which you probably wouldn't find
very competitive in the general case).
On 27 Jun, 22:19, Matthias Buelow <····@incubus.de> wrote:
> Chris Russell wrote:
> > Well that would be nice, but do ruby or python have that? How far
> > behind is a cross platform open source lisp like clisp from matching
> > the core functionality of the ruby/python interpreters, anyway?
>
> Infinitely far behind because it's licensed with the GPL and turns all
> software you make with it into GPLware automatically unless you restrict
> your program to the ANSI CL subset (which you probably wouldn't find
> very competitive in the general case).
I'd forgotten about that.
No sockets non-gpl distributed code then.
Oh well, let's hope the SBCL team get their windows kitten of death
under control soon.
Chris Russell <·····················@gmail.com> writes:
> On 27 Jun, 22:19, Matthias Buelow <····@incubus.de> wrote:
> > Chris Russell wrote:
> > > Well that would be nice, but do ruby or python have that? How far
> > > behind is a cross platform open source lisp like clisp from matching
> > > the core functionality of the ruby/python interpreters, anyway?
> >
> > Infinitely far behind because it's licensed with the GPL and turns all
> > software you make with it into GPLware automatically unless you restrict
> > your program to the ANSI CL subset (which you probably wouldn't find
> > very competitive in the general case).
>
> I'd forgotten about that.
>
> No sockets non-gpl distributed code then.
>
> Oh well, let's hope the SBCL team get their windows kitten of death
> under control soon.
I couldn't untangle the hypotheticals, the comparisons, and the other
parts of the sentences in English here well enough to understand
either the current state description or the desire. Can you restate
in direct statements what you see as the present state, the barriers
ahead, and the ultimate goal of what you think should be possible.
(Sorry if it sounds like I'm restating the whole thing about wanting
to understand clear problem descriptions, but that's just how I am.)
(In particular, I'm curious about why someone would need an "open
source lisp" if they're frightened by the GPL. It's not that I think
that's impossible--I'm just trying to understand that in more detail.
Whether it's that some other kind of open source is better, whether
it's that it's ok to be open source in the platform but must go
vertical with non-open-source, etc.)
On Jun 29, 4:36 pm, Kent M Pitman <······@nhplace.com> wrote:
> Chris Russell <·····················@gmail.com> writes:
> > On 27 Jun, 22:19, Matthias Buelow <····@incubus.de> wrote:
> > > Chris Russell wrote:
> > > > Well that would be nice, but do ruby or python have that? How far
> > > > behind is a cross platform open source lisp like clisp from matching
> > > > the core functionality of the ruby/python interpreters, anyway?
>
> > > Infinitely far behind because it's licensed with the GPL and turns all
> > > software you make with it into GPLware automatically unless you restrict
> > > your program to the ANSI CL subset (which you probably wouldn't find
> > > very competitive in the general case).
>
> > I'd forgotten about that.
>
> > No sockets non-gpl distributed code then.
>
> > Oh well, let's hope the SBCL team get their windows kitten of death
> > under control soon.
>
> I couldn't untangle the hypotheticals, the comparisons, and the other
> parts of the sentences in English here well enough to understand
> either the current state description or the desire. Can you restate
> in direct statements what you see as the present state, the barriers
> ahead, and the ultimate goal of what you think should be possible.
> (Sorry if it sounds like I'm restating the whole thing about wanting
> to understand clear problem descriptions, but that's just how I am.)
>
> (In particular, I'm curious about why someone would need an "open
> source lisp" if they're frightened by the GPL.
Because someone is not interested in giving all the source code for
free. It works for some business models but it ain't work for many
others.
Let me tell you my view how do I see an open source philosophy
through the famous car analogy:
Car manufacuters work togather to create some parts that will be
shared between different competitive brands like the engine or some
other parts because if all of them produce separate version result
will be a engine of an expensive and worse in terms of quality than a
produced togather. However they don't share the rest of their know-
how. So I'm ready to share the parts that are better developed
togather but not everything. And the viral nature of GPL force me to
do that. So GPLware is unacceptable for me. I believe in copyleft but
not in viral copyleft as the one presented in GPL.
Slobodan Blazeski
> It's not that I think
> that's impossible--I'm just trying to understand that in more detail.
> Whether it's that some other kind of open source is better, whether
> it's that it's ok to be open source in the platform but must go
> vertical with non-open-source, etc.)
On 2007-06-11, ········@gmail.com <········@gmail.com> wrote:
>> So what features do you think open source lisp is missing that
>> prevents it from competing efficiently, but which are present in ACL/
>> lispworks?
>
> A team of brilliant and dedicated engineers who have maintained and
> built out the software full time for many years, and who, given the
> opportunity, will continue to do so.
>
As someone who is just learning lisp, I don't know what this means. I
can understand the advantage of having a full-time, professional
development team. But what has this team produced that is absent from
the free implementations? Do the proprietary implementations produce
faster or more stable programs, or do they provide more/better
libraries?
Tyler
> > A team of brilliant and dedicated engineers who have maintained and
> > built out the software full time for many years, and who, given the
> > opportunity, will continue to do so.
>
> As someone who is just learning lisp, I don't know what this means. I
> can understand the advantage of having a full-time, professional
> development team. But what has this team produced that is absent from
> the free implementations? Do the proprietary implementations produce
> faster or more stable programs, or do they provide more/better
> libraries?
Uh, let's see: Yes, and Yes, and excellent development environments,
and run on every platform, and ... did I mention a team of brilliant
and dedicated engineers who have maintained and built out the software
full time for many years, and who, given the opportunity, will
continue to do so? Having used both free and commercial Lisps myself
rather extensively, I can tell you that it's like being in another
world entirely!
·········@gmail.com" <········@gmail.com> writes:
>> I don't quite understand the motivation/thoughts underlying this question. What
>> would be the advantage of paying (probably a lot) to have a commercial lisp
>> made into open source compared to putting that money into developing/expanding
>> an existing open source version?
>
> I believe that the cost of brining any of the current open lisps up to
> the level of any of the commerical ones would be about the same as
> just open sourcing a commerical one, and the latter plan would be a
> great deal faster and more reliable.
>
>> If the objective is just to try and have a rails like app, why not just do that
>> with one of the existing open source lisps?
>
> That is not the goal. The goal is to make Lisp a viable competitor in
> the current programming language market, dominated by C++, Java, and
> Python/Ruby. The Rails part of my question is merely a part of the
> overall picture. In order to play in this domain there needs to be a
> single, free, excellent Lisp with a modern library and several killer
> apps; Rails seems to be a killer app that would be easy to copy --
> that's the only reason I mentioned Rails.
You are under the impression that the cost of commercial lisps is the
sole barrier that prevents Lisp from being a "viable competitor". Why
do you think this? When you have some application that is profitable,
a few thousand dollars is peanuts. Or even in a research environment,
software licenses are a relatively small part of your expenditures,
compared to payroll and other costs.
There are nice FOSS Lisps out there, so people can try Lisp before
paying money for a commercial license.
> My theory is that is we bought and open-sourced ACL or LispWorks
> Lisps, including buying support for their engineers as gatekeepers,
> and building out the libraries, we could get the community to rally
> around that one, flush the confusing morass of current open source
> versions, and focus on building out from that root. I'm fairly certain
> that this path is the right one for Lisp; What I'm not sure of is how
> to do it in such manner as the commercial vendor doesn't fold in the
Why would anyone contribute money into this scheme, even if your
theory is correct? (which I doubt). Let people who need a commercial
Lisp pay for it. This way, everyone considers the costs and benefits
of buying a commercial Lisp, and will only do so when they see a
benefit that outweighs the costs. What you are proposing doesn't make
economic sense.
Tamas
·········@gmail.com" <········@gmail.com> writes:
>> I don't quite understand the motivation/thoughts underlying this question. What
>> would be the advantage of paying (probably a lot) to have a commercial lisp
>> made into open source compared to putting that money into developing/expanding
>> an existing open source version?
>
> I believe that the cost of brining any of the current open lisps up to
> the level of any of the commerical ones would be about the same as
> just open sourcing a commerical one, and the latter plan would be a
> great deal faster and more reliable.
>
>> If the objective is just to try and have a rails like app, why not just do that
>> with one of the existing open source lisps?
>
> That is not the goal. The goal is to make Lisp a viable competitor in
> the current programming language market, dominated by C++, Java, and
> Python/Ruby. The Rails part of my question is merely a part of the
> overall picture. In order to play in this domain there needs to be a
> single, free, excellent Lisp with a modern library and several killer
> apps; Rails seems to be a killer app that would be easy to copy --
> that's the only reason I mentioned Rails.
>
> My theory is that is we bought and open-sourced ACL or LispWorks
> Lisps, including buying support for their engineers as gatekeepers,
> and building out the libraries, we could get the community to rally
> around that one, flush the confusing morass of current open source
> versions, and focus on building out from that root. I'm fairly certain
> that this path is the right one for Lisp; What I'm not sure of is how
> to do it in such manner as the commercial vendor doesn't fold in the
> process. That vendor would have to transition to being an application
> level provider, and which one survives that will depend upon which of
> them has the best set of verticals, and which one takes the leap to
> open source their lisp first, because then their verticals will
> automatically run on the new lisp, and also they get to own the very
> important hearts and minds of the market.
>
Hmmm. I don't think its that simple. While I agree that some find the
number/variety of open source CL implementations confusing and I agree that a
good standard library consisting of high level support for common tasks
required for typical applications (networking, web, databases, portable GUIs
etc) would certainly be welcomed, I don't see that making a commercial lisp
open source will address the market share issue or dispel the FUD associated
with lisp. In fact, there is always a risk that we would lose users,
particularly corporate ones. I suspect some companies that are using one of the
commercial lisps do so because they feel there is more security in using a
product that is commercial and which probably has a well defined road
map/direction and guarantee of timely support. While I don't necessarily agree
with such arguments, they are the ones often used by companies when selecting a
product to use.
It isn't the lack of a single open source implementation with high level
features that is preventing CL from obtaining a greater market share. I think
its more to do with the predominate paradigm of procedural languages, the lack
of good intriduction to CL in teaching, the perception that CL is old and
therefore can't be as good as newer languages and possibly even a drop in skill
levels amongst programmers (I'm meeting more and more programmers who have no
real interest in their 'art' - its a job and they do it while they have to in
the same way a clerical assistant does their job. There is no great drive to
improve your skill, breadth of knowledge regarding techniques, algorithms or
data structures etc. This also seems to be encouraged by the management view of
programming being 'grunt work', which you will outsource to somewhere cheap if
you can. There is a strong perception that programming is what you do when you
first start, but if your any good, you will move up into higher positions that
are further and further removed from writing code. It is rare to find a career
path in a company that will allow you to continue programming - in fact, I ve
often seen the attitude that anyone who is a programmer in their late 30s
onwards is someone with no ambition or career aspirations. Obviusly, there are
exceptions, particularly in some specialist domains, but on the whole...)
Putting all that aside, I also don't see any real business case for the
proposal. Where is all this money going to come from to convince any company to
open source their product and replace that lost revenue. While this may work in
for some companies who can offset this through the provision of other services,
I can't see this working in the area of CL, mainly because there isn't a large
enough user base. Keeping the existing engineers on-board in order to continue
development would be expensive - wehre would that revenue come from? Other
companies that have gone down this road have been able to get new revenue from
providing things like consulting services, high level support etc, but you need
a sizable user base to do this and I don't think there is one at present.
Personally, I don't think we will see CL up there with C/C++, Java, Python,
Ruby in the foreseable future, if ever. I think there are enough people using
it and liking it to see it continue for a long time, but suspect it will remain
the domain of specialist applications or small developers/consultancies. I
think its very unlikely we will see it as a mainstream programming language.
This may change if there are more CL based success stories, particularly ones
that succeed because they are done in CL and are difficult to reproduce using
other languages or ones that somehow change the application langscape in such a
way that CLs advantages become even more self evident and which create
expectations that are difficult to meet using other languages.
Tim
--
tcross (at) rapttech dot com dot au
········@gmail.com wrote:
> That is not the goal. The goal is to make Lisp a viable competitor in
> the current programming language market, dominated by C++, Java, and
> Python/Ruby.
Wise man, choosing a goal already reached.
> In order to play in this domain there needs to be a
> single, free, excellent Lisp with a modern library and several killer
> apps; Rails seems to be a killer app that would be easy to copy --
> that's the only reason I mentioned Rails.
<sigh> You were doing so well. What Lisp has to do is not stop being
Lisp. Man, that sounds easy.
>
> My theory is that is we bought and open-sourced ACL or LispWorks
> Lisps, including buying support for their engineers as gatekeepers,
> and building out the libraries, we could get the community to rally
> around that one,....
PWUAUAHAHAHAHAHAHAHAHAHAHAHAHAAHOAOAHAHAOAHHHOOOOOHWHHEHEEEHHEH...
The idea of this Lisp community "rallying" is about as conceivable as a
hootenany down at the cemetery.
hth,kxo
On Tue, 12 Jun 2007 05:45:18 +0200, Ken Tilton <···········@optonline.net>
wrote:
>
> PWUAUAHAHAHAHAHAHAHAHAHAHAHAHAAHOAOAHAHAOAHHHOOOOOHWHHEHEEEHHEH...
>
> The idea of this Lisp community "rallying" is about as conceivable as a
> hootenany down at the cemetery.
>
> hth,kxo
You have a vierd and twisted mind.. :)
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
Ken Tilton wrote:
> PWUAUAHAHAHAHAHAHAHAHAHAHAHAHAAHOAOAHAHAOAHHHOOOOOHWHHEHEEEHHEH...
>
> The idea of this Lisp community "rallying" is about as conceivable as a
> hootenany down at the cemetery.
�Di� de Los Muertos?
On Sun, 10 Jun 2007 06:13:55 +0200, Tim X <····@nospam.dev.null> wrote:
>
> If on the other hand, someone came up with a truely original idea which
> has a
> superior implementation that is largely due to being written in CL and
> it is
> supported by some app that fills a real need that isn't being adequatly
> addressed already - well then, just maybe.....
>
http://www.franz.com/products/allegrograph/
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
(message (Hello 'Tim)
(you :wrote :on '(Sun, 10 Jun 2007 14:13:55 +1000))
(
TX> I don't quite understand the motivation/thoughts underlying this
TX> question. What would be the advantage of paying (probably a lot) to
TX> have a commercial lisp made into open source compared to putting that
TX> money into developing/expanding an existing open source version?
because open sourcing existing has known outcome.
funding development has unknown -- team might not deliver results it will
have too. as you might know, _most_ software projects either run slower than
expected, or completely fail.
as for expanding existing open source implementation, the basis they are
written on might be not appropriate for the desired product. CMUCL and SBCL
have big problems porting them to Windows, and i suspect there are also
glitches with other platforms, because of unusual memory handling. ACL and
Lispworks were ported on Windows long time ago and live fine there.
also, time might be crusial for the idea the Shrager wants to implement --
if project will deliver totally cool implementation in, say, two years, it
will have to compete not with existing languages and tools, but with future
ones -- they might be superior. OTOH open sourcing existing implementation
is much faster, if you have the money :).
by the way, i think open sourcing will not totally kill profits of
commercial vendors -- there still should be people willing to pay for
support.
the prominent north-american Linux vendor, Red Hat, has 4.5 billion dollars
market capitalization -- and all they do is selling support. anyone can
download their source code or fully compatible binaries from CentOS, for
example, and even get updates from CentOS. but some people prefer to pay for
support, and you see -- it's profitable.
certainly, that would be a loss of profit at first, since people who don't
need support won't pay. but in long drive, it can bring even more profit,
because of (much) larger user base.
)
(With-best-regards '(Alex Mizrahi) :aka 'killer_storm)
"I am everything you want and I am everything you need")
In article <·························@news.sunsite.dk>,
"Alex Mizrahi" <········@users.sourceforge.net> wrote:
> (message (Hello 'Tim)
> (you :wrote :on '(Sun, 10 Jun 2007 14:13:55 +1000))
> (
>
> TX> I don't quite understand the motivation/thoughts underlying this
> TX> question. What would be the advantage of paying (probably a lot) to
> TX> have a commercial lisp made into open source compared to putting that
> TX> money into developing/expanding an existing open source version?
>
> because open sourcing existing has known outcome.
>
> funding development has unknown -- team might not deliver results it will
> have too. as you might know, _most_ software projects either run slower than
> expected, or completely fail.
>
> as for expanding existing open source implementation, the basis they are
> written on might be not appropriate for the desired product. CMUCL and SBCL
> have big problems porting them to Windows, and i suspect there are also
> glitches with other platforms, because of unusual memory handling. ACL and
> Lispworks were ported on Windows long time ago and live fine there.
Franz bought a Windows-based Lisp implementation and merged it with
its own Unix-based offerings...
>
> also, time might be crusial for the idea the Shrager wants to implement --
> if project will deliver totally cool implementation in, say, two years, it
> will have to compete not with existing languages and tools, but with future
> ones -- they might be superior. OTOH open sourcing existing implementation
> is much faster, if you have the money :).
>
> by the way, i think open sourcing will not totally kill profits of
> commercial vendors -- there still should be people willing to pay for
> support.
> the prominent north-american Linux vendor, Red Hat, has 4.5 billion dollars
> market capitalization -- and all they do is selling support. anyone can
> download their source code or fully compatible binaries from CentOS, for
> example, and even get updates from CentOS. but some people prefer to pay for
> support, and you see -- it's profitable.
>
> certainly, that would be a loss of profit at first, since people who don't
> need support won't pay. but in long drive, it can bring even more profit,
> because of (much) larger user base.
>
> )
> (With-best-regards '(Alex Mizrahi) :aka 'killer_storm)
> "I am everything you want and I am everything you need")
--
http://lispm.dyndns.org
"Alex Mizrahi" <········@users.sourceforge.net> writes:
>
> by the way, i think open sourcing will not totally kill profits of
> commercial vendors -- there still should be people willing to pay for
> support.
> the prominent north-american Linux vendor, Red Hat, has 4.5 billion dollars
> market capitalization -- and all they do is selling support. anyone can
> download their source code or fully compatible binaries from CentOS, for
> example, and even get updates from CentOS. but some people prefer to pay for
> support, and you see -- it's profitable.
I thought the jury was still out on Red Hat's profitability? I've not checked
for some time, but last time I looked (maybe two years ago), they still hadn't
made a profit.
>
> certainly, that would be a loss of profit at first, since people who don't
> need support won't pay. but in long drive, it can bring even more profit,
> because of (much) larger user base.
>
Actually, you hit exactly my point in another post of why I don't htink this
will work for CL - you need a large user base and I don't believe CL has one
that is large enough given the size of the initial investment being discussed.
Note also that a big part of Red Hat's success has been their acceptance by
other vendors. For example, products like Oracle are only certified and
supported by Oracle as long as you run them on a specific Linux distro, Red Hat
being one of them (Oracle is possibly an interesting example as they are now
talking about doing their own Linux distro and it will be interesting to see
what they will support). Where I work, we run a lot of Red Hat, not because we
think its the best Linux distro out there, but because its the only one that is
supported by all the different vendor software we run. While they all support
different combinations of Linux, Red Hat Enterprise is the one consistent Linux
distro all the vendors support. Note also that nearly all of these vendors
require the Enterprise version - Fedora Core is not good enough. If your not
running on the enterprise version, your not supported.
I think this is a big pat of Red hat's success so far - they were able to get
support from other vendors. Interestingly, running Red Hat is not that cheaper
than running our old enterprise OS, Tru64. The licensing and support costs for
Red Hat enterprise are a bit cheaper than the discounted price we use to get
from HP (nee Compaq nee Digital), but we have a lot more servers, so that lower
cost is offset by requiring more licenses and Red hat's discount for
educational institutions is not as cheap as the one we got for Tru64 (at least
that was the case when I was managing the data centre a couple of years ago,
things may have changed).
I can't see this working with CL. As stated, I don't believe there is a big
enough user base to offset the costs over a long enough period to possibly
build a larger user base and I can't see the 'lock in' possibilities that Red
Hat was able to get through working with other software vendors. The real
problem is that large initial investment cost. If you were able to get the
software at a very low price, then just maybe it would be possible (but highly
risky IMHO).
Tim
--
tcross (at) rapttech dot com dot au
(message (Hello 'Tim)
(you :wrote :on '(Mon, 11 Jun 2007 18:31:30 +1000))
(
TX> I can't see this working with CL. As stated, I don't believe there is a
TX> big enough user base to offset the costs over a long enough period to
TX> possibly build a larger user base and I can't see the 'lock in'
TX> possibilities that Red Hat was able to get through working with other
TX> software vendors.
there are some products that work only with Allegro CL -- AllegroStore,
AllegroGraph, RACER (iirc), and possibly their web dev products live much
better on the 'real' Allegro.
so there are some killer products -- afaik AllegroGraph and RACER are the
best products for SemanticWeb. and market for Semantic Web products growth
as it gets more widespread.
so it can actually can make sense to make base product more available,
focusing on supporting software for building advanced next-generation web
applications.
but certainly Franz won't like to risk for possible significant expansion,
already having some stable income.
TX> The real problem is that large initial investment cost.
yup, it's obvious that if someone has lotsa money, he can do an interesting
experiment, that could possibly change landscape of programming language
market..
but it's very unlikely to find a source for money, given that it's a highle
risky
)
(With-best-regards '(Alex Mizrahi) :aka 'killer_storm)
"I am everything you want and I am everything you need")
In article <·························@news.sunsite.dk>,
"Alex Mizrahi" <········@users.sourceforge.net> wrote:
> (message (Hello 'Tim)
> (you :wrote :on '(Mon, 11 Jun 2007 18:31:30 +1000))
> (
>
> TX> I can't see this working with CL. As stated, I don't believe there is a
> TX> big enough user base to offset the costs over a long enough period to
> TX> possibly build a larger user base and I can't see the 'lock in'
> TX> possibilities that Red Hat was able to get through working with other
> TX> software vendors.
>
> there are some products that work only with Allegro CL -- AllegroStore,
> AllegroGraph, RACER (iirc), and possibly their web dev products live much
> better on the 'real' Allegro.
Racer runs in other Lisps, too. Racer is a product
of Racer Systems, Hamburg, Germany.
http://www.racer-systems.de/index.phtml?lang
>
> so there are some killer products -- afaik AllegroGraph and RACER are the
> best products for SemanticWeb. and market for Semantic Web products growth
> as it gets more widespread.
>
> so it can actually can make sense to make base product more available,
> focusing on supporting software for building advanced next-generation web
> applications.
>
> but certainly Franz won't like to risk for possible significant expansion,
> already having some stable income.
>
> TX> The real problem is that large initial investment cost.
>
> yup, it's obvious that if someone has lotsa money, he can do an interesting
> experiment, that could possibly change landscape of programming language
> market..
>
> but it's very unlikely to find a source for money, given that it's a highle
> risky
>
> )
> (With-best-regards '(Alex Mizrahi) :aka 'killer_storm)
> "I am everything you want and I am everything you need")
--
http://lispm.dyndns.org
On Jun 9, 8:45 pm, ·········@gmail.com" <········@gmail.com> wrote:
> Someone paid to have one of the commercial Lisps open sourced?
> How much would this cost?
> Then someone paid to have the Lisp libraries modernized to CPAN(etc)
> standards?
> How much would this cost?
> Then someone paid to do something like Rails in the foregoing new
> modernized OS Lisp?
> How muh would this cost?
Oddly enough, I was actually thinking about this after reading an
interview with Shuttleworth. Can Common Lisp get a sugar daddy? But
I'm pretty sure it would be a whole lot cheaper for said sugar daddy
to plonk down $10 million (over say 3,4 years) to bring SBCL up to
speed (including a top-notch IDE), then to buy out Allegro or
LispWorks. Nathan Myrvold has Intentional Software. Can anybody
convince Paul Allen to spare $10 million in pocket change?
In article <························@n15g2000prd.googlegroups.com>,
·········@gmail.com" <········@gmail.com> wrote:
> Someone paid to have one of the commercial Lisps open sourced?
This experiment has already been done. OpenMCL is the result.
rg
> This experiment has already been done. OpenMCL is the result.
Interesting. Not being a PPC user I hadn't considered this, but
apparently they've ported it to x86 Linux now as well (although it's
only in pre-released), so it seems that its market share may well
expand. I'll definitely look more closely into this. Thanks! (I wasn't
aware that the OpenMCL folks did a set of verticals; how do the full
time developers support themselves?)
In article <························@w5g2000hsg.googlegroups.com>,
·········@gmail.com" <········@gmail.com> wrote:
> > This experiment has already been done. OpenMCL is the result.
>
> Interesting. Not being a PPC user I hadn't considered this, but
> apparently they've ported it to x86 Linux now as well
And OS X as well. I run it on my Macbook Pro. 32-bit processors are so
last-millennium.
> (although it's only in pre-released)
You wanted open-source, this is what you get.
> how do the full time developers support themselves?)
They have day jobs, just all other open-source developers.
rg
> > > This experiment has already been done. OpenMCL is the result.
>
> > Interesting. Not being a PPC user I hadn't considered this, but
> > apparently they've ported it to x86 Linux now as well
>
> And OS X as well. I run it on my Macbook Pro. 32-bit processors are so
> last-millennium.
>
> > (although it's only in pre-released)
>
> You wanted open-source, this is what you get.
Hmmm. Well, I didn't actually want open source in the sense of not
professionally supported. What I wanted (although you'd have to study
the thread to get this) is a company that supports an excellent CL on-
the-side and makes their money through verticals, like Google supports
Guido to do python, and Sun/Java. So, MCL probably doesn't actually
fit the bill afterall.
> > how do the full time developers support themselves?)
>
> They have day jobs, just all other open-source developers.
Right, as above, this isn't what I'm after, so although I will go out
and try it, on your advice (although it looks like I'm going to have
to buy myself a 64 bit machine first!) I'm not sanguine that it's what
I'm looking for.
On Jun 12, 6:52 am, ·········@gmail.com" <········@gmail.com> wrote:
> > > > This experiment has already been done. OpenMCL is the result.
>
> > > Interesting. Not being a PPC user I hadn't considered this, but
> > > apparently they've ported it to x86 Linux now as well
>
> > And OS X as well. I run it on my Macbook Pro. 32-bit processors are so
> > last-millennium.
>
> > > (although it's only in pre-released)
No, there is a 1.0 release available on several platforms. It is true,
however, that most people seem to use the pre-release versions,
presumably since those have more interesting feature sets, and usually
aren't terribly buggy.
> > You wanted open-source, this is what you get.
>
> Hmmm. Well, I didn't actually want open source in the sense of not
> professionally supported.
OpenMCL is professionally supported (assuming you have the money to
pay Clozure to support it).
> What I wanted (although you'd have to study
> the thread to get this) is a company that supports an excellent CL on-
> the-side and makes their money through verticals, like Google supports
> Guido to do python, and Sun/Java. So, MCL probably doesn't actually
> fit the bill afterall.
Well, Clozure develops and supports OpenMCL, and makes its money doing
consulting gigs that sometimes use OpenMCL. Not exactly "verticals",
but other than that, the model seems fairly close to what you
described.
> > > how do the full time developers support themselves?)
>
> > They have day jobs, just all other open-source developers.
>
> Right, as above, this isn't what I'm after, so although I will go out
> and try it, on your advice (although it looks like I'm going to have
> to buy myself a 64 bit machine first!) I'm not sanguine that it's what
> I'm looking for.
Personally, I think it would be unwise for a company like, say,
Clozure, to toss all its eggs in one "vertical" basket, when it can
instead maintain a relatively diverse portfolio of clients.
Conversely, it seems to me that it would be unwise for a company
building a business around some set of "vertical" applications to
start out with the assumption that they are going to commit to some
particular technology platform to do it (winding up with some
particular platform because that's what turned out to work well for
their business is an entirely different matter).
As a relative newbie to Common Lisp, and specifically SBCL, I think
the issues with open source CL implementations is mostly packaging.
There are a lot of open source libraries out there that do most of the
things as the libraries advertised by the commercial lisps. For me,
Emacs (Aquamacs, actually) + SLIME is an outstanding development
environment. Once you understand it, ASDF and ASDF Install is a
pretty good package manager/installer.
However, you need to get your open source lisp, Emacs, and SLIME
separately and figure out how to make them work together. You need to
find the open source libraries you need and investigate which ones are
any good, then confirm they work well with your Lisp. You need to
learn Emacs is you don't already know it. And there's no one to call
if things don't go the way you suspect.
So I think integration and support are the two real services
commercial lisps offer over free ones. Granted, this is based on
playing around with just the free LispWorks and Allegro versions.
There may be more benefits to the paid versions I have not
experienced. There are also things like AllegroCache that look
totally awesome, but I kinda think of those as a separate product
bundled with the high end versions. If the world went to SBCL or some
other open source Lisp, I think Franz could get by just selling
AllegroCache and associated products.
I tried the free versions, but quickly blew through the heap
limitations because I was using Lisp for homeworks that required
processing large datasets. So it was on to SBCL and no looking back.
I'm also somewhat surprised that no one has mentioned the ultimate
candidate for Sugar Daddy and how he took himself out of the running
by deciding to write a totally new Lisp altogether. Someone who wrote
extensive books on Common Lisp, started and then sold a multi-million
dollar software company on Lisp, and denigrates lesser languages by
calling them "Blub".
-jimbo
On 12 Jun, 16:59, ········@gmail.com wrote:
> As a relative newbie to Common Lisp, and specifically SBCL, I think
> the issues with open source CL implementations is mostly packaging.
> There are a lot of open source libraries out there that do most of the
> things as the libraries advertised by the commercial lisps. For me,
> Emacs (Aquamacs, actually) + SLIME is an outstanding development
> environment. Once you understand it, ASDF and ASDF Install is a
> pretty good package manager/installer.
Well there are things out there that integrate all of this by default.
http://www.cliki.net/Lisp%20in%20a%20Box
But maybe more could be done to make them more obvious.
On Jun 12, 1:25 pm, Chris Russell <·····················@gmail.com>
wrote:
> On 12 Jun, 16:59, ········@gmail.com wrote:
> Well there are things out there that integrate all of this by default.http://www.cliki.net/Lisp%20in%20a%20Box
>
> But maybe more could be done to make them more obvious.
Actually, if you combined LispBox with Starter Pack:
http://weitz.de/starter-pack/
into a single, double clickable application, I think you would answer
the vast majority of problems getting started with Common Lisp. It
would be a lot closer to what people get out of the box with Python
and Ruby.
I'd love to help, but unfortunately don't know enough yet to be much
use.
I actually did start with LispBox, and graduated to AquaMacs/SBCL. As
I mentioned previously the free Allegro version I downloaded with
LispBox hit the heap limit early on. AquaMacs has SLIME pre-
installed, which is pretty sweet.
For the Mac, a single bundle that properly installed SBCL, Aquamacs,
and the Starter Pack libraries, would rival most development
environments out there, Lisp or otherwise, IMHO. As long as someone
knows how to get around Emacs, that is.
-jimbo
On Tue, 12 Jun 2007 21:14:13 +0200, <········@gmail.com> wrote:
> On Jun 12, 1:25 pm, Chris Russell <·····················@gmail.com>
> wrote:
>> On 12 Jun, 16:59, ········@gmail.com wrote:
>> Well there are things out there that integrate all of this by
>> default.http://www.cliki.net/Lisp%20in%20a%20Box
>>
>> But maybe more could be done to make them more obvious.
>
> Actually, if you combined LispBox with Starter Pack:
>
> http://weitz.de/starter-pack/
>
> into a single, double clickable application, I think you would answer
> the vast majority of problems getting started with Common Lisp. It
> would be a lot closer to what people get out of the box with Python
> and Ruby.
Not sure what you mean by Lispbox.
From gigamonkey there is LispInABox (Peter Seibel.)
This is a EMACS with SLIME (Superior Lisp Interaction Mode) preinstalled.
Works with a number of free and comercial Lisp's.
However the starter pack you mention works with LispWorks.
It provides some SLIME functionality in LispWorks IDE + Web library,
regexp's etc.
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
On Jun 13, 4:27 am, "John Thingstad" <··············@chello.no> wrote:
> Not sure what you mean by Lispbox.
http://www.gigamonkeys.com/lispbox/
"Lispbox is a version of Lisp in a Box, which was originally created
by Matthew Danish and Mikel Evins, customized for use with Practical
Common Lisp."
> However the starter pack you mention works with LispWorks.
Exactly. If it worked with, say, SBCL, and you packaged the Starter
Pack interface and SBCL into a LispBox/in a Box distribution, I think
you would have something almost as good as the low end Lispworks and
Allegro CL products.
-jimbo
In article <························@g37g2000prf.googlegroups.com>,
·········@gmail.com" <········@gmail.com> wrote:
> > > > This experiment has already been done. OpenMCL is the result.
> >
> > > Interesting. Not being a PPC user I hadn't considered this, but
> > > apparently they've ported it to x86 Linux now as well
> >
> > And OS X as well. I run it on my Macbook Pro. 32-bit processors are so
> > last-millennium.
> >
> > > (although it's only in pre-released)
> >
> > You wanted open-source, this is what you get.
>
> Hmmm. Well, I didn't actually want open source in the sense of not
> professionally supported.
Clozure supports OpenMCL to the best of their ability. And they are
some of the most professional software engineers I have ever had the
privilege of working with.
> What I wanted (although you'd have to study
> the thread to get this) is a company that supports an excellent CL on-
> the-side and makes their money through verticals, like Google supports
> Guido to do python, and Sun/Java. So, MCL probably doesn't actually
> fit the bill afterall.
I see. You want someone to make a pile of money and then use some of it
to develop and maintain a product that you can use without paying. Is
that a fair restatement of your position?
rg
> I see. You want someone to make a pile of money and then use
> some of it to develop and maintain a product that you can
> use without paying. Is that a fair restatement of your position?
I'm not sure; it doesn't sounds like what I mean, so I guess perhaps
not. A fair restatement would be this: I want to invest in one of the
commercial Lisp vendors to repurpose them to develop a specific set of
semantic web verticals which I believe Lisp is uniquely positioned
for. That company would change their business model to build and sell
those verticals instead of building and selling Lisps. Since they will
have to have a well supported Lisp (which, of course, they already
have, but can't sell for nearly the margin that the verticals will
sell for), they will continue to support their Lisp for their own
purposes, and will significantly addition enhance that support by
spinning off the lisp as open source freeware. The implementors, now
working for a semantic web verticals company, not a Lisp products
company, although still doing more-or-less exactly what they did
before. (Note that I'm investing additional money to buy additional
engineers to add the Semantic Web products layer!) The result would be
a freeware lisp with a set of highly motivated, full time gatekeepers,
supported initially by my investment, and eventually by the income
from the semweb apps. Of course, there are a number of difficult
questions that immediate come to mind: What are the semweb apps? What
is the business plan basde upon these that could convince an investor
or set of investors to go for this? Unfortuntely, unless you are a
serious interested investor, I can't go into that (and even if you
were, we obviously wouldn't do it in this forum!) What I do not know,
and what the community might be able to help me figure out, is (are)
the question(s) that I started out asking: If one of the commercial
Lisps was made into open source freeware, etc. would the community be
likely to rally around that? I shouldn't have gone down the valuation
path for the commercial vendors; that was a red herring; as has been
pointed out, that's really a matter that has to be figured out by
analysis of the specific companies involved. I did want to get a sense
of what the community thought it might cost, and how much time it
would take, to do the same starting from one of the freeware bases.
The pointer for the new, more broadly useable OpenMCL is a very useful
one, for example, so I consider that I'm getting useful information
from this thread. (Again, given that asking for a valuation of Franz
or Lispworks was a stupid question for this forum!)
On 13 Jun, 07:13, ·········@gmail.com" <········@gmail.com> wrote:
> > I see. You want someone to make a pile of money and then use
> > some of it to develop and maintain a product that you can
> > use without paying. Is that a fair restatement of your position?
>
> I'm not sure; it doesn't sounds like what I mean, so I guess perhaps
> not. A fair restatement would be this: I want to invest in one of the
> commercial Lisp vendors to repurpose them to develop a specific set of
> semantic web verticals which I believe Lisp is uniquely positioned
> for. That company would change their business model to build and sell
> those verticals instead of building and selling Lisps. Since they will
> have to have a well supported Lisp (which, of course, they already
> have, but can't sell for nearly the margin that the verticals will
> sell for), they will continue to support their Lisp for their own
> purposes, and will significantly addition enhance that support by
> spinning off the lisp as open source freeware. The implementors, now
> working for a semantic web verticals company, not a Lisp products
> company, although still doing more-or-less exactly what they did
> before. (Note that I'm investing additional money to buy additional
> engineers to add the Semantic Web products layer!) The result would be
> a freeware lisp with a set of highly motivated, full time gatekeepers,
> supported initially by my investment, and eventually by the income
> from the semweb apps. Of course, there are a number of difficult
> questions that immediate come to mind: What are the semweb apps? What
> is the business plan basde upon these that could convince an investor
> or set of investors to go for this?
Ok, so a grossly simplistic version of your investment scheme, would
be:
1)Spend lots of money.
2)Give valuable things away to everyone.
3)Sell other things.
4)Profit.
And frankly you've lost me at the second half of step 2, why bother
giving your lisp away to everyone? I can see the value in bundling a
site licence for the lisp in with your semantic web goodness, and in
fact that is what some of the semi web guys offer (I think), but that
doesn't mean you need to open source it.
As investment schemes go, it makes me worry about losing the remaining
lisp companies to a coming web2.0 winter.
> Ok, so a grossly simplistic version of your investment
> scheme, would be:
>
> 1)Spend lots of money.
> 2)Give valuable things away to everyone.
> 3)Sell other things.
> 4)Profit.
Yes, you're right. That's grossly simplistic. But let's run with it
for the moment...
> And frankly you've lost me at the second half of step 2,
> why bother giving your lisp away to everyone? I can see
> the value in bundling a site licence for the lisp in
> with your semantic web goodness, and in fact that is
> what some of the semi web guys offer (I think), but that
> doesn't mean you need to open source it.
>
> As investment schemes go, it makes me worry about losing the
> remaining lisp companies to a coming web2.0 winter.
And Iraq has weapons of mass destruction!
But scare phrase marketing aside, you apparently don't realize that
your grossly simplistic model is the same as the Open Source
(misunderstood as "freeware" -- i.e., free as in beer) model. "Why
would anyone give away valuable software?" Indeed, why would they?
Maybe if you answer that you'll be closer to answering some of your
other questions for yourself.
OTOH, since I don't personally think that the freeware model makes
sense, I won't argue very strongly for it. The only reason that it
might make sense in this case is that we're trying to chop our way out
through a forest of freeware competitors that operate according to
this model. I'm just tyring to find a path out of that forest.
On 13 Jun, 13:48, ·········@gmail.com" <········@gmail.com> wrote:
> > Ok, so a grossly simplistic version of your investment
> > scheme, would be:
>
> > 1)Spend lots of money.
> > 2)Give valuable things away to everyone.
> > 3)Sell other things.
> > 4)Profit.
>
> Yes, you're right. That's grossly simplistic. But let's run with it
> for the moment...
>
> > And frankly you've lost me at the second half of step 2,
> > why bother giving your lisp away to everyone? I can see
> > the value in bundling a site licence for the lisp in
> > with your semantic web goodness, and in fact that is
> > what some of the semi web guys offer (I think), but that
> > doesn't mean you need to open source it.
>
> > As investment schemes go, it makes me worry about losing the
> > remaining lisp companies to a coming web2.0 winter.
>
> And Iraq has weapons of mass destruction!
>
> But scare phrase marketing aside, you apparently don't realize that
> your grossly simplistic model is the same as the Open Source
> (misunderstood as "freeware" -- i.e., free as in beer) model. "Why
> would anyone give away valuable software?" Indeed, why would they?
> Maybe if you answer that you'll be closer to answering some of your
> other questions for yourself.
>
> OTOH, since I don't personally think that the freeware model makes
> sense, I won't argue very strongly for it. The only reason that it
> might make sense in this case is that we're trying to chop our way out
> through a forest of freeware competitors that operate according to
> this model. I'm just tyring to find a path out of that forest.
You are not suggesting the the open source model, you have not talked
about selling support. Your idea as you've presented it is to take a
profitable company, and have them churn out the same goods, at the
same rate, and providing the same level of support for free.
You intend to finance this by selling a marvellous web 2.0 product,
which unfortunately is too secret to describe in public. And
apparently you want to do this because there are too many open-source
lisps out there already.
Chris Russell <·····················@gmail.com> writes:
<snip>
> You intend to finance this by selling a marvellous web 2.0 product,
> which unfortunately is too secret to describe in public. And
> apparently you want to do this because there are too many open-source
> lisps out there already.
I don't think he means open-source lisps, I think he means other
open-source and/or free/low-cost dev tools such as python, ruby, perl,
java, visual studio, etc.
········@gmail.com wrote:
> OTOH, since I don't personally think that the freeware model makes
> sense, I won't argue very strongly for it. The only reason that it
> might make sense in this case is that we're trying to chop our way out
> through a forest of freeware competitors that operate according to
> this model.
No you're not. If the value is in the verticals, the freeware lisps are
not competing with you in the "valuable" space.
-dan
>>>>> "JS" == ········@gmail com <········@gmail.com> writes:
JS> But scare phrase marketing aside, you apparently don't realize
JS> that your grossly simplistic model is the same as the Open
JS> Source (misunderstood as "freeware" -- i.e., free as in beer)
JS> model. "Why would anyone give away valuable software?" Indeed,
JS> why would they? Maybe if you answer that you'll be closer to
JS> answering some of your other questions for yourself.
In the early days, because the software was developed as research
projects or as ancillary to other concerns, and was given away. This
is where BSD, early gcc, Emacs, early Linux, Perl, Apache, and the
like came from.
In the later days, because companies like IBM and Apple determined
that it was more cost-effective to take a solid open-source foundation
and contribute the specific improvements they needed to the public
good than it was to develop the whole operating system and toolchain
from the ground up. This is where later Linux and gcc come from.
Buying a closed-source in order to give it away and make money on
hypothetical poorly-defined "verticals" matches neither of these.
The root problem is that programmers need to eat. The successes of
open source happen where the programmers get to eat, because they are
getting paid by research grants or student loans, because they are
paid for doing something else and gave away this nifty tool they
created to make doing that thing easier, or because they can create
something they can sell for money for less because some of the work is
already done.
Your "buy a closed source Lisp and make it open source" fantasy does
not explain how the programmers will eat. This is a major failing.
Charlton
--
Charlton Wilbur
·······@chromatico.net
On Jun 13, 12:29 pm, Chris Russell <·····················@gmail.com>
wrote:
>
> As investment schemes go, it makes me worry about losing the remaining
> lisp companies to a coming web2.0 winter.
Although there will undoubtedly be some kind of web 2.0 winter, I
think it's clear that there is genuine value in a lot of web 2.0
stuff. TiddlyWiki, for instance, has changed the way I work to the
extent that when I recently started a contract where it wasn't going
to be realistic to run it on my desktop (bank, change control, locked
down machines etc) I went out and bought a laptop to run it without
thinking twice: even though I will not be able to connect the machine
to the bank's network it was still a no-brainer.
But the semantic web is kind of a different issue. It's just
astonishing how closely it is repeating the early AI hype: the idea
that once there is a basically sufficient representational system
(sexps then, XML now) everything else will just be easy. How wrong
they were then, how wrong they are now. I guess the hope must be
that, as before, some vast wash of government money will appear. This
is kind of moderately plausible as there's going to be a lot of
surveillance stuff coming out of the invented war on terror. But the
winter is going to be hard when it comes.
Tim Bradshaw <··········@tfeb.org> writes:
> On Jun 13, 12:29 pm, Chris Russell <·····················@gmail.com>
> wrote:
>
>>
>> As investment schemes go, it makes me worry about losing the remaining
>> lisp companies to a coming web2.0 winter.
>
> Although there will undoubtedly be some kind of web 2.0 winter, I
> think it's clear that there is genuine value in a lot of web 2.0
> stuff. TiddlyWiki, for instance, has changed the way I work to the
> extent that when I recently started a contract where it wasn't going
> to be realistic to run it on my desktop (bank, change control, locked
> down machines etc) I went out and bought a laptop to run it without
> thinking twice: even though I will not be able to connect the machine
> to the bank's network it was still a no-brainer.
>
> But the semantic web is kind of a different issue. It's just
> astonishing how closely it is repeating the early AI hype: the idea
> that once there is a basically sufficient representational system
> (sexps then, XML now) everything else will just be easy. How wrong
> they were then, how wrong they are now. I guess the hope must be
> that, as before, some vast wash of government money will appear. This
> is kind of moderately plausible as there's going to be a lot of
> surveillance stuff coming out of the invented war on terror. But the
> winter is going to be hard when it comes.
>
Its interesting you make the comparison between the AI hype and what some may
call the semantic web hype. I too think there are a lot of similarities. some
see the concept of the semantic web as something great and new thats just
around the corner. However, people have been working on this and doing research
etc on the semantic web since not long after the whole web stuff kicked off. I
originally came across these ideas back in 95 and ironically, many of these
ideas came from the same people doing research in AI, particularly in the areas
of knowledge representation, ontologies, inductive reasoning etc.
I'm not discounting the potential of a semantic web, but have to put it into
the same box as many of the goals of AI - would be nice, but I'm not convinced
we understand things enough yet to get to that next level or can even agree
enough to see the sort of consistent adoption that would be required to move
the web from its rather primitive pattern matching indexing to something that
would provide more high level representations and potentially more
sophisticated links/relationships that could provide the basis for that next
level of searching and retrieval of data/information.
I do think we will move towards a semantic web and I also still believe that we
will make advances in areas that were traditionally thought of as AI, but I
think the progress will be a lot slower and take a lot longer than many are
arguing. Like AI, I think there are some fundamental issues which need to be
resolved and which many people seem to be glossing over. Until some of these
are resolved, I don't see any real progress being made - we will probably see
some advances and from the outside, many of them will possibly look quite
amazing. However, just like with the AI stuff, a lot of these 'advances' will
likely be little more than already existing solutions that were not practicle
originally, but because of increased processing power, lower storage costs and
faster networks, some of these previously impracticle solutions become
feasible.
Going back to the OPs original idea of getting a high quality commercial lisp,
making it open source and thereby hopefully bringing people back/over to lisp
from other open source languages - I can see where he is coming from and can
understand the reasoning. Possibly too many people have concentrated on the
costs and other aspects of such a proposal and missed the underlying
motivation. I thought the comment regarding trying to find a path through the
forest of all other open source languages was interesting as I think this could
be where the proposal has shaky foundations. I think the error here may be in
trying to find a new path in a forest that is already rich in paths.
Using a different analagy, Java, python, ruby et. al. are the new team
favorites. They are seen as young, fit, powerful and sexy. they are the home
town favorites with some high profile wins. CL is seen as old, a bit fat, slow
and out classed which hasn't had a high profile win in recent memory. giving
them new uniforms is unlikely to make any difference. Trying to compete with
the home town favorites on their own turf under their rules is very unlikely to
succeed.
What CL needs to do is get some high profile wins based on its strengths
playing by the rules it knows best and using the tactics and advantages it has
to get some high profile wins. this is likely to be in a new area (possibly
even the semantic web) where the tactics haven't already been set by the other
teams and possibly, if we are lucky, establish rules and expectations which
favor CL and put the young fast, but inexperienced and often brute force power
of the others at a disadvantage.
However, having said that, I'm extremely skeptical we will ever see CL with the
sort of uptake/popularity of these other languages. I find it far more likely
that CL will maintain a respectable level of support and will likely outlast
some of these newer languages, but I don't see it ever having the popularity of
these languages again (though I don't discount the possibility of a lisp like
language gaining considerable popularity at some point in the future).
I do feel that arguing that the main reason CL isn't more popular is because
there isn't a high quality open source implementation is too simplistic. The
open source languages aren't popular just because they are open source - they
are popular because they were/are as good/productive as anything that was/is
available commercially. In fact, I suspect Java, python and ruby would still be
more popular than CL even if the whole open sorce movement had never occured
and you had to pay for them (like we use to pay for Borland C, C++ and Prolog,
which I suspect were more popular than lisp back then as well - well, maybe not
prolog). other factors also need to be taken into account, such as what venture
capital is prepared to invest in, what is taught and the early experiences
programmers have with various paradigms, the attitudes of managers to
innovation and change, the types of problems we are trying to solve and the
differences in time to master (I think Cl is possibly the easiest language I've
ever learnt, but one of the hardest I've tried to master - this is likely to
impact on adoption as well). Any attempt to increase the market share of CL
needs to consider all of these factors (and others) and not just focus on one
aspect. A high quality open source CL is certainly a worthy goal, but its not
sufficient to somehow reserect CL and put it up there with today's popular
languages.
Tim
--
tcross (at) rapttech dot com dot au
Chris Russell <·····················@gmail.com> writes:
> Ok, so a grossly simplistic version of your investment scheme, would
> be:
>
> 1)Spend lots of money.
> 2)Give valuable things away to everyone.
> 3)Sell other things.
> 4)Profit.
If this is correct, I tend to believe the claim more when the person
proposing it (a) has money to start with, (b) is the one planning to
give the money away, and (c) plans to profit on net when the endeavor
is done.
That is, incidentally, what the Lisp vendors have done... the ones
we're suggesting should be doing something else.
The thing is that when it's your own money, and you really need to
come out ahead, you're much more inclined to not ignore the details.
On 2007-06-13 02:13:31 -0400, ·········@gmail.com" <········@gmail.com> said:
> If one of the commercial
> Lisps was made into open source freeware, etc. would the community be
> likely to rally around that?
Some part of the lisp using community would, but others who are deeply
comitted to existing open source lisps might very well not. You might
want to look into merging an existing open source list (e.g., sbcl)
into an existing commercial lisp which you would buy and open source -
this would definitely increase mindshare. i.e., you'd have more of a
juggernaut with an open source lispworks-sbcl or allegro-sbcl than with
an open source lispworks or open source allegro alone.
In article <························@g37g2000prf.googlegroups.com>, ·········@gmail.com" <········@gmail.com> wrote:
> What I wanted is a company that supports an excellent CL on- the-side and
> makes their money through verticals, like Google supports Guido to do
> python, and Sun/Java.
Ron Garret wrote:
>> I see. You want someone to make a pile of money and then use
>> some of it to develop and maintain a product that you can
>> use without paying. Is that a fair restatement of your position?
Ron's summary seems to match JShrager's wish list, to me at least.
·········@gmail.com" <········@gmail.com> wrote on Tue, 12 Jun 2007:
> I want to invest in one of the commercial Lisp vendors to repurpose them to
> develop a specific set of semantic web verticals which I believe Lisp is
> uniquely positioned for. That company would change their business model to
> build and sell those verticals instead of building and selling Lisps.
If you have a great business opportunity in semantic web verticals, why do
you need the Lisp company? Just license some existing Lisp, and go make your
huge profits.
Then, after you're rich, you can get into philanthropy. Perhaps you'd like
to save tigers and pandas. Or maybe you want to support some open source
Common Lisp.
But I really don't see any necessary connection between your philanthropic
desires, and some profitable business opportunity.
> Since they will have to have a well supported Lisp (which, of course, they
> already have, but can't sell for nearly the margin that the verticals will
> sell for), they will continue to support their Lisp for their own purposes,
> and will significantly addition enhance that support by spinning off the
> lisp as open source freeware.
But you agree that there are already commercial Lisp companies providing a
product of sufficiently high quality? Why do they need to be involved at all
in your new business venture?
> Of course, there are a number of difficult questions that immediate come to
> mind: What are the semweb apps? What is the business plan basde upon these
> that could convince an investor or set of investors to go for this?
Uh, yeah. That's always the hard part, isn't it? Creating a successful,
profitable business out of nothing.
> Unfortuntely, unless you are a serious interested investor, I can't go into
> that (and even if you were, we obviously wouldn't do it in this forum!)
Yet, everything else you've talked about is useless if you can't deliver on
this part of the problem. And we know that lots of people already want to
get rich, but few succeed. You really have zero credibility on this issue.
Which makes the rest of your concerns moot.
> What I do not know, and what the community might be able to help me figure
> out, is (are) the question(s) that I started out asking: If one of the
> commercial Lisps was made into open source freeware, etc. would the
> community be likely to rally around that?
> I did want to get a sense of what the community thought it might cost, and
> how much time it would take, to do the same starting from one of the
> freeware bases.
These are all silly hypotheticals without some separate, independent way to
make tons of money. Meanwhile, if you have that separate thing, then these
questions are all irrelevant anyway.
Unless you're already rich, and this is your idea of a charity. Why don't
you come back to us when that's the situation, and we'll solve it for you
then.
-- Don
_______________________________________________________________________________
Don Geddis http://don.geddis.org/ ···@geddis.org
I wish my name was Todd, because then I could say, "Yes, my name's Todd. Todd
Blankenship." Oh, also I wish my last name was Blankenship.
-- Deep Thoughts, by Jack Handey
On Jun 13, 9:14 pm, Don Geddis <····@geddis.org> wrote:
>
> If you have a great business opportunity in semantic web verticals, why do
> you need the Lisp company? Just license some existing Lisp, and go make your
> huge profits.
I expect better of you: this is an absolutely standard trick which is
played by people on themselves all the time. We all have many
versions of it, one of mine is a variant on "if only I had a '59 Les
Paul I could play like Jimmy Page". Well, you know that's bullshit:
what's stopping me from playing like that is not being willing to
practice 5-10 hours a day and fucking myself up with speed to be able
to do so, not being willing to spend the next 5 years in a van, and
finally lack of talent. But I'd rather not face those awkward truths.
Probably almost half of the traffic on cll is people doing this
trick. Just kill them.
In article <························@g37g2000prf.googlegroups.com>,
·········@gmail.com" <········@gmail.com> wrote:
> > I see. You want someone to make a pile of money and then use
> > some of it to develop and maintain a product that you can
> > use without paying. Is that a fair restatement of your position?
>
> I'm not sure; it doesn't sounds like what I mean, so I guess perhaps
> not.
Doesn't "perhaps not" mean the same as "perhaps so"?
> A fair restatement would be this: I want to invest in one of the
> commercial Lisp vendors to repurpose them to develop a specific set of
> semantic web verticals which I believe Lisp is uniquely positioned
> for.
Do *you* want to invest, or do you want *someone else* to invest? If
the former, why don't you just do it instead of wasting your time (not
to mention everyone else's) chatting about it?
> That company would change their business model to build and sell
> those verticals instead of building and selling Lisps. Since they will
> have to have a well supported Lisp (which, of course, they already
> have, but can't sell for nearly the margin that the verticals will
> sell for), they will continue to support their Lisp for their own
> purposes, and will significantly addition enhance that support by
> spinning off the lisp as open source freeware.
OK, but why does it have to be a commercial Lisp? Why can't you do what
you propose with SBCL or CLisp? There are lots of existence proofs that
commercial software is not a prerequisite to financial success (and
indeed, some evidence that it might actually be a hindrance).
> Of course, there are a number of difficult
> questions that immediate come to mind: What are the semweb apps?
Yeah. It's those damn annoying little details, like "how do we make
money?" that shoot down the brilliant ideas every time. Damned annoying.
> What
> is the business plan basde upon these that could convince an investor
> or set of investors to go for this? Unfortuntely, unless you are a
> serious interested investor, I can't go into that
See http://rondam.blogspot.com/2006/10/top-ten-geek-business-myths.html
myth #3.
> If one of the commercial
> Lisps was made into open source freeware, etc. would the community be
> likely to rally around that?
Since you ask, IMO the answer is pretty clearly no. The Lisp community
hasn't rallied around anything in twenty years, and as I pointed out
elsewhere in this thread, the experiment has already been done at least
once with (so far) negative results.
I think you are making a very fundamental strategic error: you are
conflating the mechanism by which you intend to make money with your
plans for what to do with that money once you've made it. I also think
that if your overarching desire is to unify the Lisp community then
you're tilting at a mighty big windmill.
rg
Ron Garret <·········@flownet.com> wrote on Wed, 13 Jun 2007:
> In article <························@g37g2000prf.googlegroups.com>, ·········@gmail.com" <········@gmail.com> wrote:
>> What is the business plan basde upon these that could convince an investor
>> or set of investors to go for this? Unfortuntely, unless you are a serious
>> interested investor, I can't go into that
>
> See http://rondam.blogspot.com/2006/10/top-ten-geek-business-myths.html
> myth #3.
Similar thoughts have been echoed by Paul Graham, for example here:
http://paulgraham.com/ideas.html
[People] overvalue ideas. They think creating a startup is just a
matter of implementing some fabulous initial idea. And since a
successful startup is worth millions of dollars, a good idea is
therefore a million dollar idea.
[...]
Actually, startup ideas are not million dollar ideas, and here's an
experiment you can try to prove it: just try to sell one. Nothing
evolves faster than markets. The fact that there's no market for
startup ideas suggests there's no demand. Which means, in the narrow
sense of the word, that startup ideas are worthless.
And also perhaps
http://paulgraham.com/start.html
An idea for a startup, however, is only a beginning. A lot of
would-be startup founders think the key to the whole process is the
initial idea, and from that point all you have to do is
execute. Venture capitalists know better. If you go to VC firms with
a brilliant idea that you'll tell them about if they sign a
nondisclosure agreement, most will tell you to get lost. That shows
how much a mere idea is worth. The market price is less than the
inconvenience of signing an NDA.
Another sign of how little the initial idea is worth is the number of
startups that change their plan en route. Microsoft's original plan
was to make money selling programming languages, of all things. Their
current business model didn't occur to them until IBM dropped it in
their lap five years later.
Ideas for startups are worth something, certainly, but the trouble
is, they're not transferrable. They're not something you could hand
to someone else to execute. Their value is mainly as starting points:
as questions for the people who had them to continue thinking about.
What matters is not ideas, but the people who have them. Good people
can fix bad ideas, but good ideas can't save bad people.
_______________________________________________________________________________
Don Geddis http://don.geddis.org/ ···@geddis.org
If you ever get whipped by a bullwhip, try to breathe _in_ as the whip is going
back, and _out_ as it hits your back. Or is it the other way around? Anyway,
you'll figure it out. -- Deep Thoughts, by Jack Handey [1999]
·········@gmail.com" <········@gmail.com> writes:
> > I also think that if your overarching desire is to unify
> > the Lisp community then you're tilting at a mighty big
> > windmill.
>
> Apparently so.
Probably. But then, I don't think you've really got the data needed
to adopt what sounds like a pessimistic tone here.
I don't generally like speaking for others, but you seem to be having
trouble getting a gestalt on this discussion, and while I'd not
presume to speak for others, I'll tell you what I personally hear them
saying, though paraphrasing and aggregating according to my own personal
sense of things. (Being truly objective is hard, in other words.)
It seems to me that, yes, people have made some critical comments.
But I don't perceive that any of the comments have been intended to
say "don't try to find a way to help". I believe the comments have
been focused on "are you sure you're looking in the right place",
"what problem are you really trying to solve", etc.
People like to have help when they perceive it as help. But it's one
thing to say "how do I rearrange the icons on my desktop?" and have
someone show you. It's quite another to have someone walk in and say
"I'm here to help you clean up your desktop since it's obviously a
mess" when you've just finished setting up your icons in a way that
you like but you suddenly find that someone else doesn't.
You probably should have a discussion here on what you perceive the
problem is, if there is one, before you set off to solve it. You just
dove straight into a solution to an unstated problem, and no wonder
you got pushback.
If you skip that step, you'll end up thinking people are rejecting
your help (as I perceive you're saying above) when perhaps they don't
see it as help because they don't perceive the problem as you do.
It's much easier to evaluate the usefulness of a solution if you know
what problem it's trying to solve.
Look at how the X3J13 issues (attached to the Common Lisp HyperSpec)
are written. (I'm pretty sure it was Larry Masinter who forced us to
separate problem descriptions from proposals for solutions, and it was
a HUGE simplification of discussion.) If people can't agree on a
problem description, they have no business talking solutions, since
competing "solutions" to problems that are not agreed upon in the
first place cannot be evaluated usefully.. it just melts down with
people thinking the others are irrational because they can easily see
that the solution doesn't match their description of the problem.
It's very important to have a shared understanding of a problem if you
want to avoid that notion of people feeling others are irrational.
What the process we adopted while designing CL caused us to see was
that "the language lacks a foo construct" is NOT a proper problem
description, since linguistically/socially it only accommodates
solutions like "add a foo construct". It barely admits discussion
on the question of whether the problem description is wrong (after
all, it DOES lack a foo construct, so how COULD it be wrong? ... and
since it's a description of a problem, and the problem is clear, how
could that not be the answer? nothing to think about, right? let's
just do it.) Good problems are not described that way.
Good problem descriptions are written in a way that invite discussion
on the question of whether the problem is characterized correctly, and
whether there is a problem at all. Good problem descriptions allow
others to join the problem description with additional information
without fighting over the problem description. For example, if it's a
problem that person A wants to do tail recursive calls, person B can
also have that problem, or can want some other kind of related thing.
But person B cannot say "person A doesn't have a problem". So
eventually when everyone has spoken, you get to a place where the
problem description correctly says what the problem is.
In article <·············@nhplace.com>,
Kent M Pitman <······@nhplace.com> wrote:
> Good problem descriptions are written in a way that invite discussion
> on the question of whether the problem is characterized correctly, and
> whether there is a problem at all. Good problem descriptions allow
> others to join the problem description with additional information
> without fighting over the problem description. For example, if it's a
> problem that person A wants to do tail recursive calls, person B can
> also have that problem, or can want some other kind of related thing.
> But person B cannot say "person A doesn't have a problem". So
> eventually when everyone has spoken, you get to a place where the
> problem description correctly says what the problem is.
OK, I'll bite since Jeff doesn't seem to want to pick up the ball...
Here's a problem statement:
The current landscape of Lisp implementations seems to have a big hole
in the middle, in the sense that there are a fairly large number of free
implementations, all of which are deficient in one way or another [1],
and a smaller number of commercial implementations which are less
deficient but all very expensive. There is no Lisp of commercial
quality which is affordable by the new breed of self-funded web 2.0
startups running on a shoestring, the result of which is that most such
startups are using Python, Ruby, or PHP instead. The result is a
vicious cycle where those startups, once they succeed, don't pump any
money back into the Lisp economy, which makes it even harder for vendors
to lower their prices.
And here's a proposed solution:
If I were starting a company I would approach the commercial Lisp vendor
of my choice and propose to them that they 1) give me licenses in
exchange for some equity in my company and 2) make it more widely known
that they were amenable to such arrangements. This is almost a no-lose
situation for the vendor because the incremental cost of the license is
zero, the opportunity cost is also most likely zero (since the
alternative is Python) but the upside potential is substantial. That,
it seems to me, is a much more defensible strategy for getting into
verticals that trying to raise a bunch of capital to open-source a
commercial implementation.
rg
On Jun 14, 6:16 pm, Ron Garret <·········@flownet.com> wrote:
> In article <·············@nhplace.com>,
> Kent M Pitman <······@nhplace.com> wrote:
>
> > Good problem descriptions are written in a way that invite discussion
> > on the question of whether the problem is characterized correctly, and
> > whether there is a problem at all. Good problem descriptions allow
> > others to join the problem description with additional information
> > without fighting over the problem description. For example, if it's a
> > problem that person A wants to do tail recursive calls, person B can
> > also have that problem, or can want some other kind of related thing.
> > But person B cannot say "person A doesn't have a problem". So
> > eventually when everyone has spoken, you get to a place where the
> > problem description correctly says what the problem is.
>
> OK, I'll bite since Jeff doesn't seem to want to pick up the ball...
>
> Here's a problem statement:
>
> The current landscape of Lisp implementations seems to have a big hole
> in the middle, in the sense that there are a fairly large number of free
> implementations, all of which are deficient in one way or another [1],
> and a smaller number of commercial implementations which are less
> deficient but all very expensive. There is no Lisp of commercial
> quality which is affordable by the new breed of self-funded web 2.0
> startups running on a shoestring, the result of which is that most such
> startups are using Python, Ruby, or PHP instead. The result is a
> vicious cycle where those startups, once they succeed, don't pump any
> money back into the Lisp economy, which makes it even harder for vendors
> to lower their prices.
>
> And here's a proposed solution:
>
> If I were starting a company I would approach the commercial Lisp vendor
> of my choice and propose to them that they 1) give me licenses in
> exchange for some equity in my company and 2) make it more widely known
> that they were amenable to such arrangements. This is almost a no-lose
> situation for the vendor because the incremental cost of the license is
> zero, the opportunity cost is also most likely zero (since the
> alternative is Python) but the upside potential is substantial. That,
> it seems to me, is a much more defensible strategy for getting into
> verticals that trying to raise a bunch of capital to open-source a
> commercial implementation.
>
> rg
On Jun 14, 6:16 pm, Ron Garret <·········@flownet.com> wrote:
> Here's a problem statement:
>
> The current landscape of Lisp implementations seems to have a big hole
> in the middle, in the sense that there are a fairly large number of free
> implementations, all of which are deficient in one way or another [1],
> and a smaller number of commercial implementations which are less
> deficient but all very expensive. There is no Lisp of commercial
> quality which is affordable by the new breed of self-funded web 2.0
> startups running on a shoestring, the result of which is that most such
> startups are using Python, Ruby, or PHP instead. The result is a
> vicious cycle where those startups, once they succeed, don't pump any
> money back into the Lisp economy, which makes it even harder for vendors
> to lower their prices.
I'm not Jeff, but I'll agree that you describe a real problem.
> And here's a proposed solution:
Another solution: SBCL, Lisp-in-a-Box, Edi Weitz Starter Pack (ported
to SBCL), combined into a double click and run, self contained package
on Linux, Windows, Mac OS.
I guess the hard step is finishing the port of SBCL to Windows. Would
switching CLisp for SBCL solve that problem?
I suspect this is not very hard, just beyond my expertise.
Would such a distribution be lacking compared to commercial CLs?
Would it have enough "batteries included" to get a second look from
those tempted by Python and Rails?
I guess my proposal requires more capital and effort than your
solution, but less than Jeff's.
-jimbo
On 2007-06-14 23:10:14 -0400, ········@gmail.com said:
> Another solution: SBCL, Lisp-in-a-Box, Edi Weitz Starter Pack (ported
> to SBCL), combined into a double click and run, self contained package
> on Linux, Windows, Mac OS.
I think you miss the fact that the commercial lisps that Jeff is
talking about are capable of building native GUI apps on the platforms
on which they run (Windows, Mac and Linux for LispWorks, and Windows
and Linux for Allegro). SBCL doesn't do native Mac OS X GUI apps, and
isn't ready for prime time on Windows either. If your only concern is
writing servers then fine - maybe - but I think that it is gaps such as
these in the better open source lisp implementations that led Jeff to
suggest open sourcing an existing, polished, commercial common lisp in
the first place.
Personally, I think you're on the right track though - anyone with 10
million to invest in an open source lisp would be much better off
spending it getting sbcl up to snuff on Windows, and getting sbcl's
threads and objective-c bridge fully done on Mac OS X.
On Jun 15, 1:10 am, Raffael Cavallaro <················@pas-d'espam-
s'il-vous-plait-mac.com> wrote:
> If your only concern is
> writing servers then fine - maybe - but I think that it is gaps such as
> these in the better open source lisp implementations that led Jeff to
> suggest open sourcing an existing, polished, commercial common lisp in
> the first place.
OK, that clarifies things a lot. But I think being a good web server
development platform is the sweet spot for an open source development
platform to be right now. Python and Ruby may have GUI building
tools, but the sweet spot for them is web app development. If we had
an easy-to-get-started and use open source Lisp distro that was good
for building web apps, I think it would compare favorably to Python,
Ruby, and Perl.
-jimbo
Oops! Raffael Cavallaro <················@pas-d'espam-s'il-vous-plait-mac.com> was seen spray-painting on a wall:
> On 2007-06-14 23:10:14 -0400, ········@gmail.com said:
>
>> Another solution: SBCL, Lisp-in-a-Box, Edi Weitz Starter Pack (ported
>> to SBCL), combined into a double click and run, self contained package
>> on Linux, Windows, Mac OS.
>
> I think you miss the fact that the commercial lisps that Jeff is
> talking about are capable of building native GUI apps on the
> platforms on which they run (Windows, Mac and Linux for LispWorks,
> and Windows and Linux for Allegro). SBCL doesn't do native Mac OS X
> GUI apps, and isn't ready for prime time on Windows either.
I question the importance of this: How many famous startups have been
created to generate shrink-wrapped Win/Mac software lately?
The sorts of startup companies that have tended to make inroads
commercially have NOT been fighting the "how do I get boxes into
retail computer stores", indeed, the size of that market seems to be
shrinking steadily. The only thing selling terribly much dollar value
there is Windows Vista, and I'm not sure that's even a really material
market for Microsoft.
Keeping the "hard problems" under the sort of proprietary control that
allows a startup to later collect a pile of money seems to require
keeping the software on their own private servers, which takes all of
the things you're describing (native GUI, shrink-wrappable apps)
pretty much out of the picture, and points to the notion that the
sorts of things that even the less functional OSS Lisps support are
fairly likely to be sufficient.
When I look at the sorts of "new Mac stuff" that has been emerging,
what I'm seeing tends to be small shop production, with "nothing over
$49.95," and I'm not sure that'll make a startup rich at all fast.
--
(reverse (concatenate 'string "moc.liamg" ·@" "enworbbc"))
http://cbbrowne.com/info/lisp.html
"I have stopped reading Stephen King novels. Now I just read C code
instead." -- Richard A. O'Keefe
>>>>> "CB" == Christopher Browne <········@acm.org> writes:
CB> When I look at the sorts of "new Mac stuff" that has been
CB> emerging, what I'm seeing tends to be small shop production,
CB> with "nothing over $49.95," and I'm not sure that'll make a
CB> startup rich at all fast.
Counterexample: Delicious Library, although it depends on what your
criteria for "rich" are.
Charlton
--
Charlton Wilbur
·······@chromatico.net
On 2007-06-15 10:06:43 -0400, Christopher Browne <········@acm.org> said:
> I question the importance of this: How many famous startups have been
> created to generate shrink-wrapped Win/Mac software lately?
1. Not many no doubt, but while everyone is chasing the next big web
2.0 thing it is certainly possible that the next killer app will run on
the major desktop GUI platforms (or have an essential component that
does) and not in a web browser. c.f., iTunes - it certainly has a huge
server component, but it doesn't run in a web browser. The inability to
create apps like this traps you in the, imho dubious, web 2.0 browser
ghetto.
2. I was pointing out this defict as an example of the gaps in
existing, otherwise high quality, open source common lisps.
On Jun 14, 5:16 pm, Ron Garret <·········@flownet.com> wrote:
> Here's a problem statement:
>
> The current landscape of Lisp implementations seems to have a big hole
> in the middle, in the sense that there are a fairly large number of free
> implementations, all of which are deficient in one way or another [1],
> and a smaller number of commercial implementations which are less
> deficient but all very expensive. There is no Lisp of commercial
> quality which is affordable by the new breed of self-funded web 2.0
> startups running on a shoestring, the result of which is that most such
> startups are using Python, Ruby, or PHP instead.
Here's the problem with that last statement. Most such startups would
not use the Common Lisp (that we have today) even if SBCL was up to or
close to the quality of the commercial offerings, or if Franz or
Lispworks offered something much cheaper. You think the vast majority
of people that choose PHP are going to be interested in Common Lisp?
I don't think so. You might shave off a few of the Ruby, Python
folks, but you'd shave off more if there was a killer IDE + sweet
expressions (oh, I can hear the howls of protest already).
The result is a
> vicious cycle where those startups, once they succeed, don't pump any
> money back into the Lisp economy, which makes it even harder for vendors
> to lower their prices.
>
> And here's a proposed solution:
>
> If I were starting a company I would approach the commercial Lisp vendor
> of my choice and propose to them that they 1) give me licenses in
> exchange for some equity in my company and 2) make it more widely known
> that they were amenable to such arrangements. This is almost a no-lose
> situation for the vendor because the incremental cost of the license is
> zero, the opportunity cost is also most likely zero (since the
> alternative is Python) but the upside potential is substantial. That,
> it seems to me, is a much more defensible strategy for getting into
> verticals that trying to raise a bunch of capital to open-source a
> commercial implementation.
>
> rg
Yeah, can't you already do that (at least in the case of Franz?)
On Jun 15, 3:29 am, ·············@gmail.com" <············@gmail.com>
wrote:
[...]
> Most such startups would not use the Common Lisp (that we have
> today) even if SBCL was up to or close to the quality of the
> commercial offerings, or if Franz or Lispworks offered something
> much cheaper.
Of course not. But when the market share of a language is as low as
Common Lisps' is, you don't need to convince a whole lotta people to
start using it in order to greatly enlarge it. I'm not going to
pretend to have any insight into what start-ups would need in order to
switch to CL, but the idea that you need to have **most** of them come
over strikes me as preposterous.
Cheers, Pillsy
On Jun 15, 7:25 am, Pillsy <·········@gmail.com> wrote:
> On Jun 15, 3:29 am, ·············@gmail.com" <············@gmail.com>
> wrote:
> [...]
>
> > Most such startups would not use the Common Lisp (that we have
> > today) even if SBCL was up to or close to the quality of the
> > commercial offerings, or if Franz or Lispworks offered something
> > much cheaper.
>
> Of course not. But when the market share of a language is as low as
> Common Lisps' is, you don't need to convince a whole lotta people to
> start using it in order to greatly enlarge it.
Yes, if we currently have 8 people that use CL to do real work, and it
goes to 16, then it's doubled!
>I'm not going to
> pretend to have any insight into what start-ups would need in order to
> switch to CL, but the idea that you need to have **most** of them come
> over strikes me as preposterous.
>
> Cheers, Pillsy
Who said anything about "the idea that you need to have most of
them"? My comment was that "most such startups", as Ron Garret put
it, wouldn't even consider Common Lisp even if Franz, Lispworks was
free and/or SBCL was at the same level. It's so much more.
Quoth ·············@gmail.com" <············@gmail.com>:
> On Jun 14, 5:16 pm, Ron Garret <·········@flownet.com> wrote:
>> Here's a problem statement:
>>
>> The current landscape of Lisp implementations seems to have a big hole
>> in the middle, in the sense that there are a fairly large number of free
>> implementations, all of which are deficient in one way or another [1],
>> and a smaller number of commercial implementations which are less
>> deficient but all very expensive. There is no Lisp of commercial
>> quality which is affordable by the new breed of self-funded web 2.0
>> startups running on a shoestring, the result of which is that most such
>> startups are using Python, Ruby, or PHP instead.
>
> Here's the problem with that last statement. Most such startups would
> not use the Common Lisp (that we have today) even if SBCL was up to or
> close to the quality of the commercial offerings, or if Franz or
> Lispworks offered something much cheaper. You think the vast majority
> of people that choose PHP are going to be interested in Common Lisp?
> I don't think so. You might shave off a few of the Ruby, Python
> folks, but you'd shave off more if there was a killer IDE + sweet
> expressions (oh, I can hear the howls of protest already).
PHP programmers aren't even *close* to the market Lisp should be
after.
I am living in an edge of the PostgreSQL realm, these days; not much
direct Lisping for me, and there seem to me to be some relevant
parallels.
The comparison that frequently gets drawn is between PostgreSQL and
MySQL(tm), as they're obviously the two most popular OSS SQL database
systems. But they're NOT very close in intent or in relative
behaviours of communities.
PostgreSQL grew out of advanced DBMS research work at Berkeley, and
interest *began* from people interested in using the "advanced
features" like rules, triggers, funky kinds of indexes, custom data
types, numerous languages for stored functions, and such. It's kind
of like the "Common Lisp" of database systems :-).
MySQL(tm) was basically a SQL processor hacked onto the side of a
B-Tree library, and throwing in "advanced stuff" like basic data
integrity and support for transactions was something that only came in
with rather a lot of kicking and screaming. This very much parallels
the way PHP evolved from a way of adding bits of templates to web
pages into being a full-blown under-designed sorta-clone of Perl.
The "market" for successfully promoting PostgreSQL tends not to be
amongst the sorts of people that (5 years ago) were jumping to
MySQL(tm), looking for a "simple way of stowing web application data."
No, it tends to be more readily sold to people who think they would
sorta like to run Oracle, but can't afford the various costs of it
(it's expensive to buy, expensive to get a DBA to administer, and
rather a pig to install).
They tend to get happily surprised that modern versions *aren't* as
slow as they were told. (Visions of discovering that CL garbage
collection isn't slow, anyone?)
PostgreSQL doesn't take up much space, unless you're storing terabytes
of data. [Indeed, it might even be smaller than a base MySQL[tm]
install, maybehaps :-).] (Just as an install of CLISP or SBCL
probably makes Perl/Python/PHP/GCC/Visual Anything look pretty cheap
on disk and memory space.)
I'd like to think there's a synergy available; there seem to be some
nice CL/PG libraries kicking around these days :-).
But the point is, make sure you're trying to sell/compete in mindspace
the right ways. The people that are picking PHP probably aren't the
right ones to go after...
More likely (wild guess, and I could be wrong), it's the people
writing server apps in Java. And there's *something* missing on the
CL side, in that Java folk tend to expect to have "J2EE"-like library
infrastructures. (Which is why I may be wrong :-(.)
--
(reverse (concatenate 'string "gro.mca" ·@" "enworbbc"))
http://linuxdatabases.info/info/x.html
Inclusion of very old messages from others makes for an impressive show.
-- from the Symbolics Guidelines for Sending Mail
Christopher Browne <········@acm.org> writes:
> But the point is, make sure you're trying to sell/compete in mindspace
> the right ways. The people that are picking PHP probably aren't the
> right ones to go after...
>
> More likely (wild guess, and I could be wrong), it's the people
> writing server apps in Java. And there's *something* missing on the
> CL side, in that Java folk tend to expect to have "J2EE"-like library
> infrastructures. (Which is why I may be wrong :-(.)
Actually, I should correct that. There is a substantial portion of
the Java community that represents people that wandered in from
VB/COBOL/C++ communities to whom certain values, which I would
associate with Lisp, are not of value, notably the value of elegance.
There is (and I think it's valid) a perception that there is a common
desire to create Lisp code that is "elegant"; sometimes the term
"beautiful code" gets used.
In stark contrast, I haven't seen much PHP code that I *wouldn't* call
stolidly ugly. I have seen apps written in much the same style as the
dusty decks of repetitive COBOL code; someone builds one PHP script to
handle one application screen, and then copies that over and over and
modifies each one a bit. Pretty much the antithesis of "elegance."
I think I would apply somewhat the same dichotomy across the
databases' design choices; while they all have the same problem that
SQL has aspects of inelegance, the PostgreSQL community prefers to
implement things in a fashion that retains as much orthogonality of
function as possible (e.g. - not just kludging features on), refusing
changes that don't try to "fit well," whereas MySQL(tm) throws things
in whether they stick together well or not, leading to a lot of
"Gotchas" <http://sql-info.de/mysql/gotchas.html>.
I work on PostgreSQL replication software (see URL in my .sig ;-));
while Slony-I is not without its challenges and 'rough edges'
<http://linuxfinances.info/info/slonyintro.html>, the pages in the
MySQL(tm) docs on what sorts of things do and don't work with
replication (URLs for 4.1, 5.0 below) seem way more, um, capricious.
<http://dev.mysql.com/doc/refman/4.1/en/replication-features.html>
<http://dev.mysql.com/doc/refman/5.0/en/replication-features.html>
This notion of "elegance" suggests that perhaps the people to look for
as would-be Lisp adopters are perhaps the sorts of people that already
have some interest in 'systems with elegance.'
- At some level, that appears to apply to technical MacOS users;
that's not a "language," but it seems plausible that .dmg package
support (whether GUIed or not) would be of some value...
- Of the scripting languages, Python and Ruby are both instances where
the communities seem to have, as one of their values, a sense of
elegance. Ruby on Rails is an interesting case of this, albeit one
where people seem willing to leave out functionality when it
conflicts with "ability to implement readily."
- Users of functional languages like Haskell, ML, Erlang, and Scheme
seem to be a relevant population; their particular *senses* of
elegance tend to differ (pointedly with Scheme, where it is
common for there to be strong preferences for the Lisp/1 namespace
approach, and belief that hygenic macros are forcibly better than
"less hygenic" CL macros).
At any rate, those are communities that have some sort of a nose for
making design choices based on a notion of "elegance," which seems
like more fertile ground than, say, PHP developers who may typically
NOT have this in mind...
--
output = ("cbbrowne" ·@" "linuxdatabases.info")
http://cbbrowne.com/info/slony.html
One good turn gets most of the blankets.
On Jun 15, 12:01 pm, Christopher Browne <········@ca.afilias.info>
wrote:
> - At some level, that appears to apply to technical MacOS users;
> that's not a "language," but it seems plausible that .dmg package
> support (whether GUIed or not) would be of some value...
This seems to have a parallel with Textmate becoming the defacto
editor for Ruby on Rails applications.
-jimbo
Christopher Browne <········@acm.org> writes:
>
> PHP programmers aren't even *close* to the market Lisp should be
> after.
>
Exactly right!
> I am living in an edge of the PostgreSQL realm, these days; not much
> direct Lisping for me, and there seem to me to be some relevant
> parallels.
>
> The comparison that frequently gets drawn is between PostgreSQL and
> MySQL(tm), as they're obviously the two most popular OSS SQL database
> systems. But they're NOT very close in intent or in relative
> behaviours of communities.
>
> PostgreSQL grew out of advanced DBMS research work at Berkeley, and
> interest *began* from people interested in using the "advanced
> features" like rules, triggers, funky kinds of indexes, custom data
> types, numerous languages for stored functions, and such. It's kind
> of like the "Common Lisp" of database systems :-).
>
> MySQL(tm) was basically a SQL processor hacked onto the side of a
> B-Tree library, and throwing in "advanced stuff" like basic data
> integrity and support for transactions was something that only came in
> with rather a lot of kicking and screaming. This very much parallels
> the way PHP evolved from a way of adding bits of templates to web
> pages into being a full-blown under-designed sorta-clone of Perl.
>
> The "market" for successfully promoting PostgreSQL tends not to be
> amongst the sorts of people that (5 years ago) were jumping to
> MySQL(tm), looking for a "simple way of stowing web application data."
>
> No, it tends to be more readily sold to people who think they would
> sorta like to run Oracle, but can't afford the various costs of it
> (it's expensive to buy, expensive to get a DBA to administer, and
> rather a pig to install).
>
having worked as an Oracle developer and as an Oracle DBA, you are spot on
about it being a pig to install - and its not a nice pink domestic pig, it sone
of those big black razor back wild monsters!
> They tend to get happily surprised that modern versions *aren't* as
> slow as they were told. (Visions of discovering that CL garbage
> collection isn't slow, anyone?)
>
> PostgreSQL doesn't take up much space, unless you're storing terabytes
> of data. [Indeed, it might even be smaller than a base MySQL[tm]
> install, maybehaps :-).] (Just as an install of CLISP or SBCL
> probably makes Perl/Python/PHP/GCC/Visual Anything look pretty cheap
> on disk and memory space.)
>
> I'd like to think there's a synergy available; there seem to be some
> nice CL/PG libraries kicking around these days :-).
>
You mirror my experience exactly. 10 years ago I was developing apps with the
early releases of Java JDBC and MySQL. I came to hate MySQL because it was so
primitive and made me do much more at the application/client level rather than
in the database where I think it should have been (i.e. full referencial
integrity checks, foreign keys, triggers etc). I then worked with Oracle for a
time (actually, I still have to a bit), but now, most of the stuff I do with
databases uses postgres and I've grown to quite like it. I certainly prefer it
over MySQL, which on a couple of occasions cost me dearly because of data
loss/curruption. I don't think MySQL is really worth using unless you are doing
something very simple and Berkley DB files (or similar) are not quite enough.
I've been looking a bit at Ruby on Rails and comparing it to using cl-sql and
postgres. To be quite honest, the only real difference I've been able to see is
that rails sets up things automatically for you to start with and creates a
'forced' seperation of functionality and encourages a test driven development
approach. As I've tended to work this way when developing a system anyway, it
seems that rails just makes some of the initial 'grunt work' easier. Doing the
same with CL is quite trivial anyway. Structs seem interesting, but when you
talk to a few rails developers, one of the comments that you see a lot is
'don't use struts'.
> But the point is, make sure you're trying to sell/compete in mindspace
> the right ways. The people that are picking PHP probably aren't the
> right ones to go after...
I think thats a good point. It may also be worth mentioning, with respect to a
powerful open source CL implementation, that this group has possibly the
highest proportion of posters opposed to open source than any other group I've
participated in. If this is a reflection of attitudes in the broader CL
community (of course it may not be), then you would have to question how
successful an open source CL would ever be (I realise that a lot would depend
on what type of open source license was selected).
>
> More likely (wild guess, and I could be wrong), it's the people
> writing server apps in Java. And there's *something* missing on the
> CL side, in that Java folk tend to expect to have "J2EE"-like library
> infrastructures. (Which is why I may be wrong :-(.)
>
I still feel that a significant problem with bringing people from (lets say)
Java is that most Java programmers I've met really just don't get CL. Many of
the concepts are extremely alien to them and the change in thinking and
development process would be a big ask for many of them. This is why I think
possibly a more relevant issue with getting more CL users is an education one
rather than an open source (or favorably priced) sophisticated CL
implementation.
To me, this seems quite self evident when you talk to other developers. The
level of misunderstanding and wrong beliefs surrounding CL is quite amazing.
I'm the only person where I currently work who uses lisp and nearly all of the
other developers think its a joke. they have all the standard criticisms - too
many parens, only good for AI, not capable of dealing with procedural
problems, old and out of date, slow, etc. These users are never going to
consider CL as a viable alternative, regardless of the quality of IDEs,
libraries etc at any cost.
I have managed to get a couple interested enough to have a look at it, but they
find the amount of effort required to change the way they approach problems and
adapting to the CL way of doing things very difficult. In many ways, they can
see the advantages, but at the same time, they can't see how they can get the
time to work at it enough to feel they have reached a sufficient level of
mastery to be confident enough to do real work. I can understand this as I had
a similar experience and to be honest, I will still revert back to perl or java
when I need to get something done in a hurry and I'm not confident I can do it
in CL as quickly (though I definitely find doing it in CL more rewarding and
I'm using it more and more).
When you move from something like Java to PHP or Perl or Python or Ruby, to a
large extent, the big issue is just geting use to the syntax and a few
languages specific idioms, most of which are not that removed from what you
were already use to. However, moving from one of these languages to CL is quite
different. The syntax is almost trivial, but the paradigm shift in how you
develop and code solutions requires more effort.
Tim
--
tcross (at) rapttech dot com dot au
> OK, I'll bite since Jeff doesn't seem to want to pick
> up the ball...
Sorry; I wasn't able to read the list again until just now...
> Here's a problem statement:
>
> The current landscape of Lisp implementations seems to
> have a big hole in the middle, in the sense that there are
> a fairly large number of free implementations, all of
> which are deficient in one way or another and a smaller
> number of commercial implementations which are less
> deficient but all very expensive. There is no Lisp of
> commercial quality which is affordable by the new breed
> of self-funded web 2.0 startups running on a shoestring,
> the result of which is that most such startups are using
> Python, Ruby, or PHP instead. The result is a vicious
> cycle where those startups, once they succeed, don't pump
> any money back into the Lisp economy, which makes it even
> harder for vendors to lower their prices.
Yes, that pretty much summarizes it.
> And here's a proposed solution:
>
> If I were starting a company I would approach the commercial
> Lisp vendor of my choice and propose to them that they
> 1) give me licenses in exchange for some equity in my
> company and 2) make it more widely known that they were
> amenable to such arrangements. This is almost a no-lose
> situation for the vendor because the incremental cost of
> the license is zero, the opportunity cost is also most
> likely zero (since the alternative is Python) but the
> upside potential is substantial. That, it seems to me,
> is a much more defensible strategy for getting into
> verticals that trying to raise a bunch of capital to
> open-source a commercial implementation.
This plan has several problems. First, you are not counting the
incremental costs of supporting all those free users. One of the
important things that the commercial vendors provide is support at all
levels. By open sourcing the Lisp, you spread this load across the
community. Deep problems may still have to be resolved by the vendor,
but many, probably most issues, could be resolved w/o having to go
through the vendor by way of the OS community -- or even oneself.
Second, your plan does not give folks who are not forming companies
any reason to use these quasi-open Lisps -- that is, those who are not
actually building companies: students, personal hackers, etc. The
companies would not have any reason to bother giving those away --
they have only downside (via support as above).
In article <························@x35g2000prf.googlegroups.com>,
·········@gmail.com" <········@gmail.com> wrote:
> you are not counting the
> incremental costs of supporting all those free users.
It doesn't cost that much to tell people to RTFM. And if there are
actual bugs to be fixed you'd think the vendor would want to bear those
costs regardless.
> One of the
> important things that the commercial vendors provide is support at all
> levels. By open sourcing the Lisp, you spread this load across the
> community. Deep problems may still have to be resolved by the vendor,
> but many, probably most issues, could be resolved w/o having to go
> through the vendor by way of the OS community -- or even oneself.
Well, if you think there are deep problems with the commercial
implementations then I am at a loss to understand why you want to use
them at all. Why not just use SBCL or CLisp and be done with it?
> Second, your plan does not give folks who are not forming companies
> any reason to use these quasi-open Lisps -- that is, those who are not
> actually building companies: students, personal hackers, etc. The
> companies would not have any reason to bother giving those away --
> they have only downside (via support as above).
Yes, but so what? I'm not trying to build a Lispian Utopia, I just want
to make money (though I do believe that once people start to make money
using Lisp then the students will start using the free implementations
to learn it so they can go work for those companies.)
rg
Ron Garret <·········@flownet.com> writes:
> In article <·············@nhplace.com>,
> Kent M Pitman <······@nhplace.com> wrote:
>
>> Good problem descriptions are written in a way that invite discussion
>> on the question of whether the problem is characterized correctly, and
>> whether there is a problem at all. Good problem descriptions allow
>> others to join the problem description with additional information
>> without fighting over the problem description. For example, if it's a
>> problem that person A wants to do tail recursive calls, person B can
>> also have that problem, or can want some other kind of related thing.
>> But person B cannot say "person A doesn't have a problem". So
>> eventually when everyone has spoken, you get to a place where the
>> problem description correctly says what the problem is.
>
> OK, I'll bite since Jeff doesn't seem to want to pick up the ball...
>
> Here's a problem statement:
>
> The current landscape of Lisp implementations seems to have a big hole
> in the middle, in the sense that there are a fairly large number of free
> implementations, all of which are deficient in one way or another [1],
> and a smaller number of commercial implementations which are less
> deficient but all very expensive. There is no Lisp of commercial
> quality which is affordable by the new breed of self-funded web 2.0
> startups running on a shoestring, the result of which is that most such
> startups are using Python, Ruby, or PHP instead. The result is a
> vicious cycle where those startups, once they succeed, don't pump any
> money back into the Lisp economy, which makes it even harder for vendors
> to lower their prices.
>
> And here's a proposed solution:
>
> If I were starting a company I would approach the commercial Lisp vendor
> of my choice and propose to them that they 1) give me licenses in
> exchange for some equity in my company and 2) make it more widely known
> that they were amenable to such arrangements. This is almost a no-lose
> situation for the vendor because the incremental cost of the license is
> zero, the opportunity cost is also most likely zero (since the
> alternative is Python) but the upside potential is substantial. That,
> it seems to me, is a much more defensible strategy for getting into
> verticals that trying to raise a bunch of capital to open-source a
> commercial implementation.
>
I think you missed Kent's point. You cannot just say this is the problem and
here is the solution. We need to discuss/debate what the problem is until we
find agreement on the problem definition before putting forward a solution and
the problem needs to be presented in a way that encourages
discussion/contribution.
For example, I find no evidence in your problem definition that indicates the
small startup even considered lisp, little only discounting it because it was
too expensive. That seems to be somewhat of a leap. I actually suspect that few
startups even consider lisp as a possible development platform and never get as
far as looking at the costs of commercial implementations or even evaluate the
open source offerings. This is why I don't beleive a high quality open source
version would make a jot of difference. It is possible that a startup may come
to this conclusion if they did honestly evaluate/consider lisp, but I don't
believe that initial step is actually occuring. Therefore, I feel we need to
start with a much more basic problem definition/question - why is CL not being
considered as a viable development platform for startups?
Tim
--
tcross (at) rapttech dot com dot au
In article <··············@lion.rapttech.com.au>,
Tim X <····@nospam.dev.null> wrote:
> I think you missed Kent's point. You cannot just say this is the problem and
> here is the solution. We need to discuss/debate what the problem is until we
> find agreement on the problem definition before putting forward a solution
At first I thought you were being sarcastic, but upon reflection I
realize that not only are you being serious, you are actually correct.
Hence:
> Therefore, I feel we need to
> start with a much more basic problem definition/question - why is CL not
> being considered as a viable development platform for startups?
That's a question, not a problem statement. Let me try again:
Problem statement version 2: The Lisp economy is shrinking.
(My proposed solution still applies.)
rg
Ron Garret wrote:
> Problem statement version 2: The Lisp economy is shrinking.
You cannot possibly know that (unless you have done an extensive market
study, but I don't believe you did).
I guess what you actually want to say is this: The Lisp economy is not
growing as much as you wish it did; it may even be shrinking.
Pascal
P.S.: The problem in such discussions is that some people seem to
believe that there can be a "big bang" solution, where one single
"clever" move of the Lisp community (whatever that is) suddenly gets
Lisp back to the "top" again. IMHO, this is naive at best.
We have to work at several levels, where the commercial vendors, the
open source vendors, the solution providers, the book writers, the
academic personnel, etc. pp., all contribute to making Lisp more widely
known and more widely used, step by step. Contrary to popular belief,
this is actually happening.
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
> P.S.: The problem in such discussions is that some people seem
> to believe that there can be a "big bang" solution, where
> one single "clever" move of the Lisp community (whatever
> that is) suddenly gets Lisp back to the "top" again.
> IMHO, this is naive at best.
Ah. Thank you, Pascal! I think that you've clarified something very
important, and brought me back to my initial mind-set. You are exactly
correct that that is the problem with discussion of the form Kent
challanged us to engage in -- a challange which Ron took up. You'll
notice that I did NOT say: "Here's the problem, and here's a
solution." In fact, I explicitly avoided that, and Kent took me to
task for that. For a fleeting moment, esp. when Ron responded to
Kent's challange, I lost my way and was sucked into that mode of
thought, but that's NOT where was was coming from, and you're exactly
right that it's NOT where I want to go...which is exactly why I didn't
do it the way Kent would have liked. Instead, I asked a specifically
evolutionary question: What would happen if ...
> We have to work at several levels, where the commercial vendors,
> the open source vendors, the solution providers, the book
> writers, the academic personnel, etc. pp., all contribute
> to making Lisp more widely known and more widely used,
> step by step. Contrary to popular belief, this is
> actually happening.
Although I think that this is probably true to small measure, it's
also true to large measure that Lisp is losing relative mind share.
(Approx. Ron's second problem statement.) My theory was that something
additional has to be added to the mix -- a big bang of some sort.
Recall that what killed the dinosaurs, and enabled us to evolve, was
certainly a complex evolutionary process, but a big part of it was a
big bang! So big bangs needed be thought of as magic -- they are a
part of evolutionary processes....but if one is trying to infuse a big
bang into the process, one would like to know what would happen if ...
I know that it's bad form to reply to my own post, but I wanted to
bring the subject line back in line with the correct (to my mind)
evolutionary vision of this thread.
·········@gmail.com" <········@gmail.com> writes:
> Although I think that this is probably true to small measure, it's
> also true to large measure that Lisp is losing relative mind share.
> (Approx. Ron's second problem statement.) My theory was that something
> additional has to be added to the mix -- a big bang of some sort.
> Recall that what killed the dinosaurs, and enabled us to evolve, was
> certainly a complex evolutionary process, but a big part of it was a
> big bang! So big bangs needed be thought of as magic -- they are a
> part of evolutionary processes....but if one is trying to infuse a big
> bang into the process, one would like to know what would happen if ...
Well you could as well try this:
What if some startup took one free Common Lisp and built a commercial
implementation from it, like RedHat and SuSE did for Linux?
The problem may be that the existing commercial implementors don't
make it to the front page often enough. Perhaps a CL RedHat could
interest the newspapers and the shareholders, like the Linux RedHat
did?
--
__Pascal Bourguignon__ http://www.informatimago.com/
NOTE: The most fundamental particles in this product are held
together by a "gluing" force about which little is currently known
and whose adhesive power can therefore not be permanently
guaranteed.
Pascal Bourguignon wrote:
> What if some startup took one free Common Lisp and built a commercial
> implementation from it, like RedHat and SuSE did for Linux?
Or Scieneer did with CMUCL.
-dan
Daniel Barlow <···@coruskate.net> writes:
> Pascal Bourguignon wrote:
>> What if some startup took one free Common Lisp and built a commercial
>> implementation from it, like RedHat and SuSE did for Linux?
>
> Or Scieneer did with CMUCL.
Indeed. So the question is: How many customers?
--
__Pascal Bourguignon__ http://www.informatimago.com/
NOTE: The most fundamental particles in this product are held
together by a "gluing" force about which little is currently known
and whose adhesive power can therefore not be permanently
guaranteed.
········@gmail.com wrote:
>> P.S.: The problem in such discussions is that some people seem
>> to believe that there can be a "big bang" solution, where
>> one single "clever" move of the Lisp community (whatever
>> that is) suddenly gets Lisp back to the "top" again.
>> IMHO, this is naive at best.
>
> Ah. Thank you, Pascal! I think that you've clarified something very
> important, and brought me back to my initial mind-set. You are exactly
> correct that that is the problem with discussion of the form Kent
> challanged us to engage in -- a challange which Ron took up. You'll
> notice that I did NOT say: "Here's the problem, and here's a
> solution." In fact, I explicitly avoided that, and Kent took me to
> task for that. For a fleeting moment, esp. when Ron responded to
> Kent's challange, I lost my way and was sucked into that mode of
> thought, but that's NOT where was was coming from, and you're exactly
> right that it's NOT where I want to go...which is exactly why I didn't
> do it the way Kent would have liked. Instead, I asked a specifically
> evolutionary question: What would happen if ...
>
>> We have to work at several levels, where the commercial vendors,
>> the open source vendors, the solution providers, the book
>> writers, the academic personnel, etc. pp., all contribute
>> to making Lisp more widely known and more widely used,
>> step by step. Contrary to popular belief, this is
>> actually happening.
>
> Although I think that this is probably true to small measure, it's
> also true to large measure that Lisp is losing relative mind share.
> (Approx. Ron's second problem statement.) My theory was that something
> additional has to be added to the mix -- a big bang of some sort.
> Recall that what killed the dinosaurs, and enabled us to evolve, was
> certainly a complex evolutionary process, but a big part of it was a
> big bang! So big bangs needed be thought of as magic -- they are a
> part of evolutionary processes....but if one is trying to infuse a big
> bang into the process, one would like to know what would happen if ...
I don't get it. Seriously. Do you want a big bang or not?
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
> I don't get it. Seriously. Do you want a big bang or not?
My point is that "big bangs" (hereafter BBs) are a natural part of
evolution, and that evolution absorbs the impact and continues,
hopefully somewhat reshuffled by the impact. Hopefully BBs are less
common and have greater effect than the "small bangs" of everyday
evolution. So, yes, I want a BB, but to distinguish this from
evolution is, I think, a mistake. Moreover, I think that Kent is
completely wrong (and annoyingly "holier than though") when he looks
for a crisp statement of the problem, and a crisp solution. That's
good for programming, but not for complex evolving systems, esp.
social systems! So, what I want is exactly what I asked for: An
analysis by the community of how the evolution of the Lisp community
would be reshuffled if the BB that I described (those three points)
were to take place. Nothing more nor less. (Well, I also asked about
costs, which, as I've said elsewhere, was silly on my part, and I've
retracted these.)
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <uejka7h4x.fsf@nhplace.com>
·········@gmail.com" <········@gmail.com> writes:
> So, yes, I want a BB, but to distinguish this from
> evolution is, I think, a mistake.
I have no problem with trying to analyze systems in complex ways. I do it
all the time.
If you read my 1990 paper on conditions, you'll see a discussion in
there about conditions as a subjective matter, where I point out that
the exact line between what is a "normal situation" and what's an
"exceptional situation" is subjectively drawn, but that I think that
doesn't matter: what matters is that people understand that the
conversation cannot proceed easily without drawing it. People can
even dynamically adjust a need for change of terminology as they go,
but to say there is no terminology and that everything is open season
doesn't help apply rationality to chaos.
A financial system can be modeled as a closed or open system, for
example. Each is a model. One admits the possibilities of things
outside, and one doesn't. Sometimes that matters, sometimes it
doesn't. It doesn't chagne the reality, but it simplifies the
discussion.
It's hard to talk about complex things at all, and doubly hard if you
don't expose what you want. I believe, as a fundamental principle,
that the way one approaches these things, is to try to bound a
discussion on any side in which there is an ability to do so. There
are too many degrees of freedom to get past the complexity if you
don't. So I've said that. That is not me putting down idea of
addressing hard problems, it is me addressing the very serious nature
of them.
Some problems really are systematic, but some only seem so. I suspect,
for example, that global warming is systematic and does not yield to
smaller analysis. However, I think the chief barrier to that is that
to solve it will require global consensus. And if someone doesn't work
through the discussion of why a local solution doesn't work, they will
not end up with the right people in the discussion to bootstrap the will
for change. You're leaping over an obligation to satisfy Ocham's Razor
by insisting, sans apparent proof, that the problems are hard. And I
don't see that this particular problem you're talking about is "obviously"
hard, but that isn't me saying it's not hard. It's me saying that most
things that purport to be obvious are not obvious. Obvious is an elusive,
point of view kind of thing. Many things are non-obviously true even if
not obviously true.
The funny and sad thing is that I don't even know if I disagree with you,
or with Ron (and I can't tell if you agree with him). You both seem to
perceive me as arguing against your positions rather than seeing my questions
as saying "I'm interested but I think I won't be able to follow what you're
saying without some help". You're describing my reaction as "holier than
thou" in a very off-putting way that suggests that you have a right to have
me believe you quietly or not at all. If you have that power and right, why
not just apply it directly. Assert that you, yourself, are holier than we.
And just be done with it. I'm not saying you're doing that, but I'm saying
that you might as well if you're not willing to listen to what people are
saying about why they don't understand what you're going after.
I think you think that by doing what I'm doing, I'm trying to disable
you or make your issue not relevant. I think I'm doing good science
by asking you to first show that the problem does not admit to modular
solutions before you use the extraordinary claim that all things must
be considered at once.
> Moreover, I think that Kent is completely wrong (and annoyingly
> "holier than though") when he looks for a crisp statement of the
> problem, and a crisp solution.
I'm wrong for looking? Hmm. Do you feel the same about truth and
justice? Are people wrong for looking for crisp statements of those?
Or only for insisting that such exist? I think a great deal of very
interesting thought has come from such. I cited one yesterday in Ayn
Rand's book, The Fountainhead. I'd say the same of the American Civil
Liberties Union, and even of Richard Stallman. There is virtue in
seeking crispness, but that is not the same as saying that crispness
is the answer. Clear terminology and perspective sharpens
conversation, but does not substitute for answers.
> That's good for programming, but not for complex evolving systems,
While complex systems are hard to analyze, the one tool we have is
simplification and abstraction. That's all I've asked for. Nor do
you have to listen to me. But why do you have the right to criticize
me with ad hominems ("holier than thou"--that's not a technical
criticism) that in the selfsame breath suggest you have the right to
be right. I'm just expressing my opinion, as are you. But if your
expressing yours is reasonable and mine is not, then you might see
that perhaps what you're doing is not operating on a balanced field of
discussion.
> esp.
> social systems! So, what I want is exactly what I asked for: An
> analysis by the community of how the evolution of the Lisp community
> would be reshuffled if the BB that I described (those three points)
> were to take place. Nothing more nor less.
In my personal opinion, this is a meaningless discussion to have
without some amount of framing. Going back to what you said about
evolution, and working with that as a premise: The world is what it is
because, and only because, of where it was, what it needed at each
moment in between, and what was left. There is no why, there is only
the fact of the moment. That's the real truth. So if it's going to
go another way, you have to say why. Is it an economic incentive? Is
it fiat? Ask the people of a country that is democratic state what
would happen if it went to a socialist state. ... well, the answer
would be different if the reason was "we were invaded by others who
forced us" than if the answer was "we came to this conclusion
ourselves". The case here is analogous. If you don't lay out some
details of your hypothetical, no one can understand the premise, and
it's yours to control. It's a faux discussion.
> (Well, I also asked about costs, which, as I've said elsewhere, was
> silly on my part, and I've retracted these.)
Costs vary depending on reasons, too. Ask me how much it would cost to
take something near and dear to me, like my teddy bear, and I'll tell
you one number. Tell me whether it's going to a charity, to a museum,
or to a grandchild and the answer might be different. Your problem
setup was nonsensical to me (in the strict sense, not trying to
disparage it--"it made no apparent sense") because it didn't distinguish
these utterly different cases. "How would you like to travel to another
country for free?" Would your answer change if I told you that you'd be
kidnapped and sold into slavery to make that happen for you? There are
very real analogs here where people, to buy into, to cooperate with, to
not try to sabotage, to be willing to just act in harmony with the ideas
you're saying would have to know what you were after. And if you insist
they have to give their answer before knowing these things, it sounds like
a kind of 1-2 punch that is setting them up to be quoted inappropriately
as approving something they're not sure they approve. And none of that
is saying you intend it that way. It's just how it came across. So I
was trying to help by suggesting a mechanism that I had seen work in the
past. Call that holier than thou if you like. I've been called worse.
But it isn't my intent.
I work as hard as any person reasonably could here to make sure that
any association between my name and some particular success in the
past is not taken as a credential for me to have a right to speak for
the language. I'm just a person. Fallible. I don't own the language.
I didn't even make the language--I just edited its spec in what I hope
was neutral fashion, and contributed to its design. But, like anyone,
I have some experience, and I like to share that. I expect my ideas
to stand on their own, and to either persuade or not on their own
merit, not because I said them. I have a bit of a name in this very
narrow forum, so that maybe buys more than average people to read a
few of my posts, and it gets me a little ahead. But if I didn't say
something that resonated, it would not be enough to have anyone agree.
If you want to argue with me, please argue with the points I've made.
Please try to stay clear of the ad hominems. It doesn't suit you.
The most important reason is that I'd like to hear what you have to
say, and I don't want you discredited merely for trying to say it. I
just want to hear it in a form I can make some sense of. But if you
disagree, fine, say it your way. But I may not be the only person to
ask. Nor should you take my rejection of what I perceive as an
incoherently presented position as rejection of the position if I've
clearly stated that it's the presentation I'm rejecting, exactly
because it doesn't let me se the position in order to either agree or
disagree with it.
In article <·············@nhplace.com>,
Kent M Pitman <······@nhplace.com> wrote:
> The funny and sad thing is that I don't even know if I disagree with you,
> or with Ron (and I can't tell if you agree with him). You both seem to
> perceive me as arguing against your positions rather than seeing my questions
> as saying "I'm interested but I think I won't be able to follow what you're
> saying without some help".
I don't perceive you that way. My only quarrel with you in this
discussion is that you misrepresented my position (an annoyingly common
reaction here, and not just from you). In general I find what you have
to say to be almost uniformly worthwhile, insightful, and fair (if a bit
long-winded at times). It may appear that we disagree more than we do
because most of the time when I agree with you I often don't have
anything to add and it's considered bad nettiquette to just chime in
with a "me too" so I just lurk.
rg
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <umyyye6au.fsf@nhplace.com>
Ron Garret <·········@flownet.com> writes:
> In article <·············@nhplace.com>,
> Kent M Pitman <······@nhplace.com> wrote:
>
> > The funny and sad thing is that I don't even know if I disagree
> > with you, or with Ron (and I can't tell if you agree with
> > him). You both seem to perceive me as arguing against your
> > positions rather than seeing my questions as saying "I'm
> > interested but I think I won't be able to follow what you're
> > saying without some help".
>
> I don't perceive you that way. My only quarrel with you in this
> discussion is that you misrepresented my position (an annoyingly common
> reaction here, and not just from you).
I'll just observe that I didn't intend at all to represent your
opinion, only to respond to it. As such, I could say you
misrepresented my position, by representing it as a misrepresentation
rather than by representing it as a terminological confusion. I think
misrepresentation is a crime of intent, and I had no intent. Neither
do I think you had such an intent here. And so I won't provoke a
cycle by suggesting you did. I'll just say you're illustrating the
problem of reference in what you say, and that if you can understand
what was bothering you about what I said, maybe you can understand
what I'm saying you just did that bothers me. And then maybe we can
just move on and admit language to be imperfect and intentions to be
good. Without at least presupposing good intentions in others among
us, and working past those to talk about issues, we won't make
progress.
> In general I find what you have to say to be almost uniformly
> worthwhile, insightful, and fair (if a bit long-winded at times).
Sorry about the longwinded part. (I normally expect the cure will be that
people won't read me when they get bored.) I often don't have the time
to write shorter, though sometimes I do and you should see the parts I don't
send sometimes! But I plead basically guiltyon that parenthetical point.
> It may appear that we disagree more than we do because most of the
> time when I agree with you I often don't have anything to add and
> it's considered bad nettiquette to just chime in with a "me too" so
> I just lurk.
Fair enough. Though I have to say I think that sometimes though the
netiquette documents say otherwise, it's important to chime in... when
it achieves a purpose. And in the case where it says something
surprising, I think it's not wasted. While not every post needs a "me
too", I think the case where a "me too" must be done, even in
opposition to a general tendancy not to, is when it indicates a
surprise or shift. In that case, it is only a syntactic "me too", but
the underlying semantics is different. This is good in my book:
(macrolet ((me-too () '(identify-consensus-reached)))
...)
and
(macrolet ((me-too () '(give-weight-to-suspected-irrational-impression)))
...)
That makes the remark provide a service. I think what the normal advice
tries to avoid is:
(macrolet ((me-too () 'nil))
...)
On Sun, 17 Jun 2007 07:17:53 -0700, ·········@gmail.com" <········@gmail.com> said:
| ...
| My point is that "big bangs" (hereafter BBs) are a natural part of
| evolution, and that evolution absorbs the impact and continues,
| hopefully somewhat reshuffled by the impact. Hopefully BBs are less
| common and have greater effect than the "small bangs" of everyday
| evolution. So, yes, I want a BB, but to distinguish this from
| evolution is, I think, a mistake.
This makes me think that it may be worthwhile to clarify "natural
part of evolution". Obviously, if a big bang event strikes an
evolutionary process, the latter will be impacted, and whatever the
outcome, it will be natural. But does evolution _require_ big bang
events in order to even proceed? (Another question would be whether
evolution _produces_ big bang events, but I am not sure if we should
be addressing more than one such question at a time.)
There is an interesting, I think, (pre)historic case, that of the
Neolithic Revolution (which was the "First Economic Revolution", the
Industrial Revolution being the second one). It occurred roughly in
the course of a couple of millennia, in the broad vicinity of 10,000
BC, starting in the Golden Crescent of what we call today the Middle
East. The formulation of the problem was simple: hunger, brought
about by climate changes after the most recent Ice Age, which led to
the extinction of large animals that human hunter-gatherers living
in that region had been subsisting on. The solution, at the high
level, was also simple: start growing and breeding one's own food,
i.e. start farming, which was the first productive economy of human
society.
While any resemblances to any modern situation may be purely
coincidental, the question here is: would productive economy have
come into existence without the big bang event of climate change,
leading to fauna change? Nobody knows the answer with certainty, of
course, but note that in speculation mode, it is possible to argue
both ways:
* it wouldn't, because it had not happened before, even though
Homo sapiens had been in existence for a period of time that
was an order of magnitude larger;
* it would have, eventually, perhaps more slowly, and it hadn't
happen before because humans had not been technologically
advanced enough earlier.
So, I suppose the answer to the question, "to BB or not to BB", is
not quite obvious, especially given that big bang events also
introduce instability which may be difficult to control to ensure
that the outcome stays on the right track.
---Vassil.
--
The truly good code is the obviously correct code.
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <uy7iiefhh.fsf@nhplace.com>
Vassil Nikolov <···············@pobox.com> writes:
> While any resemblances to any modern situation may be purely
> coincidental, the question here is: would productive economy have
> come into existence without the big bang event of climate change,
> leading to fauna change?
It's easy to point to a big bang, which is really just an out of
control chaotic event, as having caused good. After all, the good
that routine big bangs did not enable to survive is not around to talk
to. This is succinctly put by others as "anything that doesn't kill
you outright makes you stronger". But it doesn't follow from that
that it's safe to try to kill yourself outright. You can just get
killed.
It's easy to point to a big bang of climate change as the enabling
force for modern civilization, but it doesn't follow that the one that
might be teetering on the brink is equally benevolent (if even the
previous one can be described as that--I suspect it was quite harsh,
and the time it took to grow back was a lot).
A big bang in Lisp could enable important change. It could also be
very injurious. Mostly though, it's unnecessary in the strict sense.
To make Pepsi, one does not have to convince people to give up Coke.
One instead makes Pepsi and sees whether Coke goes away. (Or maybe it
happened the other way around, I didn't check. It doesn't matter.)
Competition is not about voluntary elimination of one thing on promise
of another. It's about two things existing and competing side by side.
Is that what's being proposed here, in this discussino of a big bang,
because I don't see it. It's true that the introduction of a
really good alternative to something can be VIEWED as a big bang in how
fast people leave the old thing to go to the new... but it's structurally
different. People didn't agree to give up paper on promise of the web.
The web came along and there was a big whoosh of voluntary kind.
I see people in this discussion talking about it like it's something
one could plan and orchestrate, and like one should. But maybe I'm
wrong. I see people who (it seems to me) don't see value in the
present system talking people who (it seems to me) do see value in the
present system (not infinite value--but value that matters to
themselves, and sometimes their ability to feed their famiiles) and
asking those people to give up what they have for the promise that
others will supply a better way. Maybe I'm misinterpreting. That's not
intentional. But I'm trying to say what I'm seeing so if that's wrong you
can adjust what I see accordingly.
Whereas I think I agree with most of what you've said, I don't agree
from here onward:
> Mostly though, [a big bang is] unnecessary in the strict sense.
> To make Pepsi, one does not have to convince people to give
> up Coke. One instead makes Pepsi and sees whether Coke goes
> away. [...] Competition is not about voluntary
> elimination of one thing on promise of another. It's
> about two things existing and competing side by side.
Although I might agree about Pepsi and Coke, the situtation in the
modern programming language world is not analogous. If Coke (or Pepsi)
were free, the analogy would be more accurate.
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <utzt6eans.fsf@nhplace.com>
·········@gmail.com" <········@gmail.com> writes:
> Whereas I think I agree with most of what you've said, I don't agree
> from here onward:
>
> > Mostly though, [a big bang is] unnecessary in the strict sense.
> > To make Pepsi, one does not have to convince people to give
> > up Coke. One instead makes Pepsi and sees whether Coke goes
> > away. [...] Competition is not about voluntary
> > elimination of one thing on promise of another. It's
> > about two things existing and competing side by side.
>
> Although I might agree about Pepsi and Coke, the situtation in the
> modern programming language world is not analogous. If Coke (or Pepsi)
> were free, the analogy would be more accurate.
Why? Suppose WE did muddy the terminological waters by making Free as
in Soda mean the opposite of the perplexing phrase "free as in beer". :)
How would that have changed the interplay between coke and pepsi, or how
would it change the analogy?
On Jun 17, 10:46 am, Kent M Pitman <······@nhplace.com> wrote:
> ·········@gmail.com" <········@gmail.com> writes:
> > Whereas I think I agree with most of what you've said, I don't agree
> > from here onward:
>
> > > Mostly though, [a big bang is] unnecessary in the strict sense.
> > > To make Pepsi, one does not have to convince people to give
> > > up Coke. One instead makes Pepsi and sees whether Coke goes
> > > away. [...] Competition is not about voluntary
> > > elimination of one thing on promise of another. It's
> > > about two things existing and competing side by side.
>
> > Although I might agree about Pepsi and Coke, the situtation in the
> > modern programming language world is not analogous. If Coke (or Pepsi)
> > were free, the analogy would be more accurate.
>
> Why? Suppose WE did muddy the terminological waters by making Free as
> in Soda mean the opposite of the perplexing phrase "free as in beer". :)
> How would that have changed the interplay between coke and pepsi, or how
> would it change the analogy?
You're kidding, right?
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <uir9me68r.fsf@nhplace.com>
·········@gmail.com" <········@gmail.com> writes:
> On Jun 17, 10:46 am, Kent M Pitman <······@nhplace.com> wrote:
> > ·········@gmail.com" <········@gmail.com> writes:
> > > Whereas I think I agree with most of what you've said, I don't agree
> > > from here onward:
> >
> > > > Mostly though, [a big bang is] unnecessary in the strict sense.
> > > > To make Pepsi, one does not have to convince people to give
> > > > up Coke. One instead makes Pepsi and sees whether Coke goes
> > > > away. [...] Competition is not about voluntary
> > > > elimination of one thing on promise of another. It's
> > > > about two things existing and competing side by side.
> >
> > > Although I might agree about Pepsi and Coke, the situtation in the
> > > modern programming language world is not analogous. If Coke (or Pepsi)
> > > were free, the analogy would be more accurate.
> >
> > Why? Suppose WE did muddy the terminological waters by making Free as
> > in Soda mean the opposite of the perplexing phrase "free as in beer". :)
> > How would that have changed the interplay between coke and pepsi, or how
> > would it change the analogy?
>
> You're kidding, right?
Absolutely not. I don't doubt it would affect things. But I don't
think the effect is uniquely determined. I'm genuinely interested in
what you're trying to suggest here.
On Sun, 17 Jun 2007 10:26:47 -0700, ·········@gmail.com" <········@gmail.com> said:
| Whereas I think I agree with most of what you've said, I don't agree
| from here onward:
|| Mostly though, [a big bang is] unnecessary in the strict sense.
|| To make Pepsi, one does not have to convince people to give
|| up Coke. One instead makes Pepsi and sees whether Coke goes
|| away. [...] Competition is not about voluntary
|| elimination of one thing on promise of another. It's
|| about two things existing and competing side by side.
| Although I might agree about Pepsi and Coke, the situtation in the
| modern programming language world is not analogous. If Coke (or Pepsi)
| were free, the analogy would be more accurate.
I think that the issue of "new player on the block" vs. "convert an
existing player" is a different issue from the one of any of the
players charging or not charging to play. (Possibly the dependence
between the two issues is not that high, too.)
Perhaps we ought to look at newspaper distribution for an analogy:
there are the "traditional" dailies and there are those such as
"Metro" that are distributed free of charge. The latter came on
stage as new players; it wasn't the case that an existing newspaper
dropped its price to zero. Would that work better?
---Vassil.
--
The truly good code is the obviously correct code.
> Perhaps we ought to look at newspaper distribution for an analogy:
> there are the "traditional" dailies and there are those such as
> "Metro" that are distributed free of charge. The latter came on
> stage as new players; it wasn't the case that an existing newspaper
> dropped its price to zero. Would that work better?
Okay, let's take this, but the daily's aren't a good comparison (many
are not free, and they generally do not compete in the same market as
the majors); the web and TV are better examples. As a result largely
of alternative, *free* sources of news, newspapers are in very serious
finanacial shape. Many (perhaps all the largest ones) have produced
competing free products and have branched out into additional
"verticals", such as weekly speciality magazines. But, in reality,
this isn't a very good example because the price point isn't
comparable. You get the same NYTimes for 50c as reading new online,
whereas you don't get the same ACL for 99$ as free alternatives.
In article <························@i38g2000prf.googlegroups.com>,
·········@gmail.com" <········@gmail.com> wrote:
> > Perhaps we ought to look at newspaper distribution for an analogy:
> > there are the "traditional" dailies and there are those such as
> > "Metro" that are distributed free of charge. The latter came on
> > stage as new players; it wasn't the case that an existing newspaper
> > dropped its price to zero. Would that work better?
>
> Okay, let's take this, but the daily's aren't a good comparison (many
> are not free, and they generally do not compete in the same market as
> the majors); the web and TV are better examples. As a result largely
> of alternative, *free* sources of news, newspapers are in very serious
> finanacial shape. Many (perhaps all the largest ones) have produced
> competing free products and have branched out into additional
> "verticals", such as weekly speciality magazines. But, in reality,
> this isn't a very good example because the price point isn't
> comparable. You get the same NYTimes for 50c as reading new online,
> whereas you don't get the same ACL for 99$ as free alternatives.
Jeff, the discussion seems to go into the obscure.
What are your plans?
--
http://lispm.dyndns.org
> Jeff, the discussion seems to go into the obscure.
This isn't the least bit obscure: Selling expensive software in a
market full of highly competitive cheaper (esp. free!) products is a
losing battle. One must join'em, and seek to develop new markets in,
for example, verticals. The newspaper analogy is very close!
> What are your plans?
Well, since I've got an answer to my question of what would happen
if..., and, more precisely, to the more specific question: Would the
Lisp community rally around an excellent free Lisp that runs on and is
easy to install on every platform, has an excellent IDE and modern
library set, and is well supported by a small set of brilliant
committed individuals? The answer is clearly "no". So whatever plans I
had, I no longer have.
From: Edi Weitz
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <usl8qpby3.fsf@agharta.de>
On Sun, 17 Jun 2007 13:10:21 -0700, ·········@gmail.com" <········@gmail.com> wrote:
> Well, since I've got an answer to my question of what would happen
> if..., and, more precisely, to the more specific question: Would the
> Lisp community rally around an excellent free Lisp that runs on and
> is easy to install on every platform, has an excellent IDE and
> modern library set, and is well supported by a small set of
> brilliant committed individuals? The answer is clearly "no". So
> whatever plans I had, I no longer have.
So, a Usenet discussion with two or three dozen participants answers
this question for you? Have you also tried crystal balls and reading
the cards just so you have a control group in good scientific
tradition?
--
Lisp is not dead, it just smells funny.
Real email: (replace (subseq ·········@agharta.de" 5) "edi")
> So, a Usenet discussion with two or three dozen participants
> answers this question for you?
Oh, no. This conversation is only a small part of my thinking on this.
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <uejkamhc4.fsf@nhplace.com>
·········@gmail.com" <········@gmail.com> writes:
> > Jeff, the discussion seems to go into the obscure.
>
> This isn't the least bit obscure: Selling expensive software in a
> market full of highly competitive cheaper (esp. free!) products is a
> losing battle. One must join'em, and seek to develop new markets in,
> for example, verticals. The newspaper analogy is very close!
If it's too expensive, why not just decide not to buy it? Why does it
matter that others do?
Why not just create something that is the price you like?
> > What are your plans?
>
> Well, since I've got an answer to my question of what would happen
> if..., and, more precisely, to the more specific question: Would the
> Lisp community rally around an excellent free Lisp that runs on and is
> easy to install on every platform, has an excellent IDE and modern
> library set, and is well supported by a small set of brilliant
> committed individuals? The answer is clearly "no". So whatever plans I
> had, I no longer have.
That's not the question you asked.
I'm not sure what the answer would be. But I'm sure it's not what I
answered. I bet others agree.
And, incidentally, in spite of what I'm sure are all good intentions,
it has the character of a blank check.
If I asked "would you rally around and commit to support a new book
author who wrote on intensely interesting topics with high quality plots
of an undeniably great nature?" would the answer be yes or no, sight
unseen?
If I asked "would the country rally around and commit to support a new
politician who solved real issues, appealed to everyone, didn't overspend,
and finally listened to people, would you say yes or no?"
Or might you say "gee, the details kind of matter. why are you asking?"
All politicians start out that way, but the devil is often in the details.
A pity we didn't get to the discussion of what we did and could agree
on, what we are afraid of, what we are proud of, etc. because we have
failed to just blindly trust someone who comes out of nowhere offended
that we ask a few questions before adopting him as our savior. The
reception you've gotten has been respectful and and interested, but has
indeed not been lemming-like.
> A pity we didn't get to the discussion of what we did and
> could agree on, what we are afraid of, what we are proud of,
> etc. because we have failed to just blindly trust someone who
> comes out of nowhere offended that we ask a few questions
> before adopting him as our savior. The reception you've
> gotten has been respectful and and interested, but has
> indeed not been lemming-like.
You have a most eloquent (and long winded) way of putting words and
intentions into people's mouths. I am not the least bit offened. I am
merely expressing my opinion, based upon the various mostly useful
analyses appearing in this thread, and upon other data and thought,
that the Lisp community will not rally around ... as above. No offense
seen, taken, or offered.
I am allowed to have an opinion, right?
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <uabuym279.fsf@nhplace.com>
·········@gmail.com" <········@gmail.com> writes:
> > A pity we didn't get to the discussion of what we did and
> > could agree on, what we are afraid of, what we are proud of,
> > etc. because we have failed to just blindly trust someone who
> > comes out of nowhere offended that we ask a few questions
> > before adopting him as our savior. The reception you've
> > gotten has been respectful and and interested, but has
> > indeed not been lemming-like.
>
> You have a most eloquent (and long winded) way of putting words and
> intentions into people's mouths. I am not the least bit offened. I am
> merely expressing my opinion, based upon the various mostly useful
> analyses appearing in this thread, and upon other data and thought,
> that the Lisp community will not rally around ... as above. No offense
> seen, taken, or offered.
>
> I am allowed to have an opinion, right?
It's a free forum. I don't run it. I do not exercise authority, but
instead I merely observe the obvious when I say: You're certainly
allowed to have an opinion.
But so am I allowed to have an opinion. And mine is that I disagree
that the situation can be usefully summed up so succinctly. Here's
why I think that:
We're a pluralistic group, so I'm just trying to tell you that if you
think we can speak with one voice to EITHER accept you or reject you
or even divide the possible states into two such clearly delineated
positions, then you don't even believe your own problem description
because you yourself appear to be citing part of the problem as a lack
of unified community, and then concluding you shouldn't attack the
problem because you don't hear a unified community accepting your
unarticulated plan with a single and clear voice.
But beyond that: if you're this easily daunted, then I must say it's
probably good for you to notice that fact early, before you have the
hopes and livelihoods of hundreds or thousands of people hanging on
your success. What if the particular individuals who responded, none
of whom are elected to speak for the community, had said something you
construed as acceptance and then you moved forward only to find others
disagreed? Would after-the-outset dispute have turned others' voices
to lies? Who would you believe? If you continued on, why would that
be different than continuing on now?
And what if we were lucky enough that someone had come along with an
alleged better offer? How would we detect that? Would you have been
offended if we asked questions about that? Would they have been? How
would we discuss it? How do we know the status quo is not such an
offer, given that we don't know the basis for the decisions?
Questions are not the enemy--they are how people sort out hard problems.
> We're a pluralistic group, so I'm just trying to tell
> you that if you think we can speak with one voice to EITHER
> accept you or reject you or even divide the possible
> states into two such clearly delineated positions, then
> you don't even believe your own problem description
> because you yourself appear to be citing part of the
> problem as a lack of unified community, and then concluding
> you shouldn't attack the problem because you don't hear a
> unified community accepting your unarticulated plan with a
> single and clear voice.
> But beyond that: if you're this easily daunted, then I
> must say it's probably good for you to notice that fact
> early, before you have the hopes and livelihoods of
> hundreds or thousands of people hanging on your success.
> What if the particular individuals who responded, none
> of whom are elected to speak for the community, had said
> something you construed as acceptance and then you moved
> forward only to find others disagreed? Would after-the-
> outset dispute have turned others' voices to lies? Who
> would you believe? If you continued on, why would that
> be different than continuing on now?
>
> And what if we were lucky enough that someone had come
> along with an alleged better offer? How would we detect
> that? Would you have been offended if we asked questions
> about that? Would they have been? How would we discuss it?
> How do we know the status quo is not such an
> offer, given that we don't know the basis for the decisions?
Are you off your meds, or something?
········@gmail.com wrote:
> Are you off your meds, or something?
Brilliant retort.
BTW, this whole BB thing is just cargo-cult thinking.
"Oh look, there was a BB in the past, therefore we need
one now!"
You don't get a BB by seeking a BB. You get whatever
evolutionary process is natural under the circumstances,
and you get that by doing whatever you can to help improve
things. If there's a huge store of energy or market demand
built up somewhere, then a BB can occur. But regardless,
the only productive approach is to analyze the situation
and figure out what you can do to improve it. If you think
a BB is likely or inevitable, then you have to explain why
in terms of the current situation, not assume it as an
a-priori goal.
--
Dan
www.prairienet.org/~dsb/
> If it's too expensive, why not just decide not to buy it?
> Why does it matter that others do?
Hunh?
> Why not just create something that is the price you like?
Double hunh??
In article <························@q19g2000prn.googlegroups.com>,
·········@gmail.com" <········@gmail.com> wrote:
> > Jeff, the discussion seems to go into the obscure.
>
> This isn't the least bit obscure: Selling expensive software in a
> market full of highly competitive cheaper (esp. free!) products is a
> losing battle. One must join'em, and seek to develop new markets in,
> for example, verticals. The newspaper analogy is very close!
>
> > What are your plans?
>
> Well, since I've got an answer to my question of what would happen
> if..., and, more precisely, to the more specific question: Would the
> Lisp community rally around an excellent free Lisp that runs on and is
> easy to install on every platform, has an excellent IDE and modern
> library set, and is well supported by a small set of brilliant
> committed individuals? The answer is clearly "no". So whatever plans I
> had, I no longer have.
Oh, I would use it. If I'd be still alive then...
--
http://lispm.dyndns.org
On Jun 18, 7:10 am, ·········@gmail.com" <········@gmail.com> wrote:
> You all seem to think that I have some evil
> intent, or that I don't understand what it means to ask questions or
> something, and that I need a lecture in how to ask appropriate
> questions. As far as I can tell not a single person has actually tried
> to actually answer my questions. (My apologies to the soul or two who
> might actually have done so, only to be drowned out by the Pedant
> Club.)
Daydreaming is nice - but keep it on the low else you won't finish
anything.
Wishful thinking is good but at the and you need to implement the
functionality or find someone who would do it for you. Or , you're
program will be a vapourware.
There's nothing evil in your idea except that there's no way to
implement it. So keep you feets on the ground and do something that
you could do it. Improving some existing lisp, writing & patching
libraries and last but not least making applications in lisp would
help lisp far better than all those daydreaming, ifs, buts and what
would happen ...And it would save a lot of people in here from wasting
time in pointless discussion. Lisp is close to the edge of being
treated as useless academia (and it's even not in the academias any
more , if you count off learning lisp on a pen & paper togather with 5
other languages) so don't make any steps in the wrong direction.
I wish you a happy lisping
Slobodan Blazeski
My apologies to those who like Java, C#, PHP, Delphi, Visual
Basic,Perl, Python, Ruby, COBOL, or any other language. I know you
think you know a better language than lisp. All I can say is I do,
too!
P.S.
I uderstand you have a good intentions but the road to hell is paved
with good intentions.
On Jun 17, 1:10 pm, ·········@gmail.com" <········@gmail.com> wrote:
> Well, since I've got an answer to my question of what would happen
> if..., and, more precisely, to the more specific question: Would the
> Lisp community rally around an excellent free Lisp that runs on and is
> easy to install on every platform, has an excellent IDE and modern
> library set, and is well supported by a small set of brilliant
> committed individuals? The answer is clearly "no".
Did I miss the useful descriptions of "excellent Lisp", "excellent
IDE", and "modern library"?
Also, what does "rally around" mean?
FWIW, my current "platform" is a datacenter full of multi-cpu systems,
each with tons of disk and ram, and some routers. I don't really care
about ease of install, but if you've got something relevant, I'm
interested. (I'm also confused, as there are maybe 1000 organizations
with comparable "platforms".)
I'm curious as to how a lisp that is excellent for my platform will
also be excellent on a 6502 with 1Mbyte.
········@gmail.com wrote:
> Whereas I think I agree with most of what you've said, I don't agree
> from here onward:
>
>> Mostly though, [a big bang is] unnecessary in the strict sense.
>> To make Pepsi, one does not have to convince people to give
>> up Coke. One instead makes Pepsi and sees whether Coke goes
>> away. [...] Competition is not about voluntary
>> elimination of one thing on promise of another. It's
>> about two things existing and competing side by side.
>
> Although I might agree about Pepsi and Coke, the situtation in the
> modern programming language world is not analogous. If Coke (or Pepsi)
> were free, the analogy would be more accurate.
Programming languages are never for free. You may get your first Barbie
doll for free, but then you have to buy all the accessories. Or invest a
lot of time to build your own.
:)
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
On 17 Jun 2007 12:02:34 -0400, Kent M Pitman <······@nhplace.com> said:
| Vassil Nikolov <···············@pobox.com> writes:
|| While any resemblances to any modern situation may be purely
|| coincidental, the question here is: would productive economy have
|| come into existence without the big bang event of climate change,
|| leading to fauna change?
| It's easy to point to a big bang, which is really just an out of
| control chaotic event, as having caused good.
I can see now I should have made it explicit that I did not give
that as an example of something that made things better. In other
words, I should have made it explicit that I gave the Neolithic
Revolution as an example of neither a Good Thing nor a Bad
Thing---just as an example of a Thing that Was what It Was. I
suppose that even today, some ten thousand years later, it is
debatable whether we are better off or worse off than those
hunter-gatherers from before the revolution---possibly neither if
all is taken into account, and certainly neither happier nor sadder.
And, more importantly, for the first farmers life did not
improve---on the contrary, it became worse, it was just that their
previous way of life could not continue. In other words, I should
have made it clear that my point was not to view that big bang of
climate change and its consequences as having caused good (or
bad)---merely as having caused change.
And, of course, we don't know if there was a better solution that
our ancestors did not find or choose.
| After all, the good
| that routine big bangs did not enable to survive is not around to talk
| to. This is succinctly put by others as "anything that doesn't kill
| you outright makes you stronger". But it doesn't follow from that
| that it's safe to try to kill yourself outright. You can just get
| killed.
That's the point I tried to make, perhaps not clearly enough, by
saying that instability may be difficult to control and direct
towards a desired outcome, so I think we are in agreement here.
| It's easy to point to a big bang of climate change as the enabling
| force for modern civilization, but it doesn't follow that the one that
| might be teetering on the brink is equally benevolent (if even the
| previous one can be described as that--I suspect it was quite harsh,
| and the time it took to grow back was a lot).
Indeed. We survived, but if it had been our own deliberate
decision, we might not have chosen for things to happen that way.
---Vassil.
--
The truly good code is the obviously correct code.
········@gmail.com wrote:
>> I don't get it. Seriously. Do you want a big bang or not?
>
> My point is that "big bangs" (hereafter BBs) are a natural part of
> evolution, and that evolution absorbs the impact and continues,
> hopefully somewhat reshuffled by the impact. Hopefully BBs are less
> common and have greater effect than the "small bangs" of everyday
> evolution. So, yes, I want a BB, but to distinguish this from
> evolution is, I think, a mistake. Moreover, I think that Kent is
> completely wrong (and annoyingly "holier than though") when he looks
> for a crisp statement of the problem, and a crisp solution. That's
> good for programming, but not for complex evolving systems, esp.
> social systems! So, what I want is exactly what I asked for: An
> analysis by the community of how the evolution of the Lisp community
> would be reshuffled if the BB that I described (those three points)
> were to take place. Nothing more nor less.
IIUC, your three points were:
+ Someone paid to have one of the commercial Lisps open sourced?
+ Then someone paid to have the Lisp libraries modernized to CPAN(etc)
standards?
+ Then someone paid to do something like Rails in the foregoing new
modernized OS Lisp?
This wouldn't reshuffle anything. The situation would be more or less
similar to the current one:
+ There would be one more open source implementation and one less
commercial one.
+ There are currently plenty of libraries out there. People who want to
get work done with Common Lisp are successful at doing so already.
+ There are even enough alternatives for Rails.
IIRC, at some stage you said that Common Lispers may revolve around that
single open sourced implementation. (I don't find the posting at the
moment. Please correct me if I'm wrong.) I think this is very unlikely.
Lisp is, to a large extent, about having a growable programming language
where everyone can add their own language constructs. This inevitably
leads to a situation where different language extensions with different
trade offs are made. This inevitably leads to diversification. The
single-implementation-language approach doesn't work for Lisp.
Only a scenario that can take advantage of the diversity of Lisp is
likely to be successful, not one that tries to reduce diversity. And I
strongly believe that diversity is an advantage, not a disadvantage.
If you want to do something for Common Lisp's future, pour some money
into the CDR process. This would help to create better portable code and
bring increased reliability to libraries.
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
> + There are currently plenty of libraries out there.
> People who want to get work done with Common Lisp are
> successful at doing so already.
There seems to be disagreement about whether there are plenty of
libraries out there, or at least whether they are out there in some
accessible structure.
> ... you said that Common Lispers may revolve around that
> single open sourced implementation. ... I think this is very
> unlikely.
Noted.
> Lisp is, to a large extent, about having a growable
> programming language where everyone can add their own
> language constructs. This inevitably leads to a situation
> where different language extensions with different
> trade offs are made. This inevitably leads to diversification.
> The single-implementation-language approach doesn't work for
> Lisp. Only a scenario that can take advantage of the
> diversity of Lisp is likely to be successful, not one
> that tries to reduce diversity. And I strongly believe that
> diversity is an advantage, not a disadvantage.
Unfortunately, I think that this is deeply confused. (And you aren't
going to tell me, of all people, that I don't understand Lisp!)
Although I agree that Lisp is, to a large extent, about having a
growable programming language, I strongly disagree that single-
implementation-language approach doesn't work for Lisp. This is like
saying that because digital representation is universal we can't all
agree to use binary at the base. Or, to be a little more reasonable,
like saying that because I86 architectures support both Windows and
Unix OSs, the world can sustain ten randomly broken I86 processor chip
sets. Do you think that if someone were to build an almost-working I86
chip that was slightly different than the standard one, and give it
away for free, anyone would bother with it? Actually, they probably
would if the commercial ones were extremely expensive, and that's more
or less the situation that we have today in Lisp, although isn't not
just one randomly broken free Lisp, it's ten. And, moreover, there are
pretty good, free, non-broken, alternatives, and they are completely
compatible with one another (because there's just one of each)!
········@gmail.com wrote:
>> + There are currently plenty of libraries out there.
>> People who want to get work done with Common Lisp are
>> successful at doing so already.
>
> There seems to be disagreement about whether there are plenty of
> libraries out there, or at least whether they are out there in some
> accessible structure.
We can at least agree that the situation can (always) be improved.
That's an easy nut to crack, though.
>> Lisp is, to a large extent, about having a growable
>> programming language where everyone can add their own
>> language constructs. This inevitably leads to a situation
>> where different language extensions with different
>> trade offs are made. This inevitably leads to diversification.
>> The single-implementation-language approach doesn't work for
>> Lisp. Only a scenario that can take advantage of the
>> diversity of Lisp is likely to be successful, not one
>> that tries to reduce diversity. And I strongly believe that
>> diversity is an advantage, not a disadvantage.
>
> Unfortunately, I think that this is deeply confused. (And you aren't
> going to tell me, of all people, that I don't understand Lisp!)
No, of course I am not. ;)
> Although I agree that Lisp is, to a large extent, about having a
> growable programming language, I strongly disagree that single-
> implementation-language approach doesn't work for Lisp. This is like
> saying that because digital representation is universal we can't all
> agree to use binary at the base. Or, to be a little more reasonable,
> like saying that because I86 architectures support both Windows and
> Unix OSs, the world can sustain ten randomly broken I86 processor chip
> sets. Do you think that if someone were to build an almost-working I86
> chip that was slightly different than the standard one, and give it
> away for free, anyone would bother with it? Actually, they probably
> would if the commercial ones were extremely expensive, and that's more
> or less the situation that we have today in Lisp, although isn't not
> just one randomly broken free Lisp, it's ten.
Why do you characterize the existing open source CL implementations as
being broken?
> And, moreover, there are
> pretty good, free, non-broken, alternatives, and they are completely
> compatible with one another (because there's just one of each)!
I didn't get that part.
Anyway, even if you would succeed in uniting all Common Lispers, there's
still the pesky other Lisp dialects...
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
> Anyway, even if you would succeed in uniting all Common
> Lispers, there's still the pesky other Lisp dialects...
I'd settle for this small vistory. ;-)
On Jun 17, 10:03 pm, ·········@gmail.com" <········@gmail.com> wrote:
> > Anyway, even if you would succeed in uniting all Common
> > Lispers, there's still the pesky other Lisp dialects...
>
> I'd settle for this small victory. ;-)
I have a dream today.
I say today, my friends, that in spite of the difficulties and
frustrations of the moment, I still have a dream. It is a dream deeply
rooted in the Lisp dream.
I have a dream that one day this community will rise up and live out
the true meaning of its creed: "We hold these thruths to be self-
evident: lisp is the greatest language ever created."
I have a dream that one day on the red hills of usenet the users of
free and commercial will be able to sit down together at a table of
broterhood. As there will only one implementation of lisp free and
greatest of all.
I have a dream today.
"Common lispers of the world, unite!",
You rise the flag, I'm going to get some beer.
Slobodan Blazeski
My apologies to those who like Java, C#, PHP, Delphi, Visual
Basic,Perl, Python, Ruby, COBOL, or any other language. I know you
think you know a better language than lisp. All I can say is I do,
too!
On Jun 18, 8:24 am, fireblade <·················@gmail.com> wrote:
> You rise the flag, I'm going to get some beer.
Try Red Stripe - and sing while you're at it!!
http://bc.tech.coop/blog/070514.html.
agt
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <uabuy7gsj.fsf@nhplace.com>
Pascal Costanza <··@p-cos.net> writes:
> Lisp is, to a large extent, about having a growable programming
> language where everyone can add their own language constructs. This
> inevitably leads to a situation where different language extensions
> with different trade offs are made. This inevitably leads to
> diversification. The single-implementation-language approach doesn't
> work for Lisp.
This is a very important point that I think it's great to highlight
because it isn't talked about explicitly much, though I tink we're all
aware of it.
Lisp is, in so many ways, an exercise in a philosophy of pluralism.
It's about multiple namespaces. It's about multiple programming
paradigms (OO, functional, imperative, etc.) It's about multiple
operations that do the same things (conditional, loop, I/O constructs)
in different ways.
> Only a scenario that can take advantage of the diversity of Lisp is
> likely to be successful, not one that tries to reduce diversity. And I
> strongly believe that diversity is an advantage, not a disadvantage.
I've heard it said that "it's easier to learn to like an idea than to
learn to unlike it". I have never been sure if that's exactly the
right way to phrase that thought, but it certainly is close enough to
be relevant and provocative here.
> Appears you just want to generate idle talk; As an
> investor, are you treating "community" as human
> "commodity market"? Your ideas on using opensource
> indicate you do
I have no idea what your last sentence means, but the one before is
completely wrong. I am not trying to generate idle talk, I'm trying to
get opinions as to what would happen if ... etc.
I'm sorry but I just don't understand why so many people seem to think
that there's something wrong with asking a simple "What if" question.
Am I speaking a foreign language or something? Is "What if" somehow a
new form in English? If I asked: "What would happen if the sun was to
explode tomorrow?" you might think it an odd question, but you'd have
no trouble answering it. You all seem to think that I have some evil
intent, or that I don't understand what it means to ask questions or
something, and that I need a lecture in how to ask appropriate
questions. As far as I can tell not a single person has actually tried
to actually answer my questions. (My apologies to the soul or two who
might actually have done so, only to be drowned out by the Pedant
Club.)
From: Don Geddis
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <87r6o9f8vq.fsf@geddis.org>
·········@gmail.com" <········@gmail.com> wrote on Sun, 17 Jun 2007:
> I'm sorry but I just don't understand why so many people seem to think
> that there's something wrong with asking a simple "What if" question.
> You all seem to think that I have some evil intent, or that I don't
> understand what it means to ask questions or something, and that I need a
> lecture in how to ask appropriate questions.
Because you haven't given enough details for any of the answers to mean
anything. Plus, your hypothetical situation is highly unlikely in the
extreme, so there doesn't appear to be much value in wasting time trying to
clarify all the possibilities.
You would make more progress if either: (1) you provided more details, so that
it really did become a "simple" what-if question with a simple answer; or
(2) you offered better arguments on why the situation is feasible (for example,
that some billionaire wants to make this happen as a mere whim), so that there
is value on spending effort to clarify and answer your vague thoughts.
-- Don
_______________________________________________________________________________
Don Geddis http://don.geddis.org/ ···@geddis.org
Love is like racing across the frozen tundra on a snowmobile which flips over,
trapping you underneath. At night, the ice-weasels come. -- Nietchze
> your hypothetical situation is highly unlikely in the
> extreme, so there doesn't appear to be much value in wasting time trying to
> clarify all the possibilities.
That's fair enough, but then why would you (for example) engage in the
conversation at all? Some folks seem to think that it's clear enough,
at least possibly with subsequent clarifying posts, to engage in a non-
meta conversation about it.
> You would make more progress if either: (1) you provided more details, so that
> it really did become a "simple" what-if question with a simple answer;
As I've said before, I don't actually think that it's possible to make
this a simple question, and I don't think that it will have a simple
answer. I could make it more specific: "What if I spent ten millions
dollars of my own money doing X, Y, and Z?" But I doubt that that
would help. Maybe you can give me a template for the sort of question
in this domain that you think has a simple answer. I think that you
will find that that that is not so simple a task.
> or
> (2) you offered better arguments on why the situation is feasible (for example,
> that some billionaire wants to make this happen as a mere whim), so that there
> is value on spending effort to clarify and answer your vague thoughts.
Why do you think that the situation is infeasible? The sort of buy out
I proposed happens every day in industry, and buying out these
products wouldn't be very heavy lifting for the folks that do that
sort of thing. Whether there is enough of a value proposition here to
get one or more investors to actually do that is a different question,
and I did not ask that. But it's extremely easy for me to imagine
gathering enough money to do it if the value proposition was right.
From: Don Geddis
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <874pl3dfxa.fsf@geddis.org>
·········@gmail.com" <········@gmail.com> wrote on Mon, 18 Jun 2007:
> Why do you think that the situation is infeasible?
If I've understood your proposal, it goes something like this:
. Right now, CL has a bunch of open source implementations and a bunch of
commercial ones.
. You think all the open source ones are not "good enough", but you haven't
given any details about what exactly they are lacking, nor have you explained
(in enough detail) what your actual goal is, so nobody else is able to
judge whether a given implementation might be lacking or not either.
. You've proposed that somebody (unknown, rich), "buy out" one of the
commercial implementations, and make it open source. You've offered nobody
who already has money that would want to do this as a whim, nor have you
suggested why it would be a profitable investment on its own, to attract
new investors.
. You've proposed -- but not explained even a little bit -- that some
organization would now start working on "verticals" based on this Lisp.
And all the questions about money eventually come down to this. That
everybody would make it back on the "verticals". But you've given zero
details about what any vertical is, nor any justification for why a
profitable vertical -- even if you knew of one -- depends at all on the
open-source-a-commercial-lisp initial concept. They seem totally
disconnected to me.
Putting it all together, your "question" seems to be of the form "what IF
there was a rich, stupid person, and what IF there were profitable verticals,
and what IF the rich stupid person bought a commercial lisp, and what IF
all the programmers in the commercial company continued working for the new
open source, and what IF the enormous profits from the verticals paid for
this effort for all eternity -- in that case, would the current Lisp
community rally around the new open source Lisp?"
Who knows? Who cares? None of those things is going to happen.
> The sort of buy out I proposed happens every day in industry, and buying
> out these products wouldn't be very heavy lifting for the folks that do
> that sort of thing.
But the people who have the money to do it, don't do it as a charity.
They do it as an investment, because there is a likelihood of profit.
You've given no story to attract any potential investor.
> Whether there is enough of a value proposition here to get one or more
> investors to actually do that is a different question, and I did not ask
> that.
No, you just assumed it.
But you seem to ignore all the responses you've gotten that this is the
hardest part, and none of the other things you claim to care about even
matter if you can't solve this part. Moreover, if you CAN solve this part,
then all your other concerns will go away.
> But it's extremely easy for me to imagine gathering enough money to do it
> if the value proposition was right.
Of course. I could do it too. And if wishes were horses, beggars could ride.
And if pigs could fly...
Fine, let me try to help you. Your questions will be less of a joke if you
work on one of these things:
1. Realize that you get zero credit for "imagine I knew of a secret profitable
vertical". Profitable businesses are very hard to make, and nobody is willing
to just give you this as an assumption. That one assumption is harder than
everything else you've mentioned.
2. Make a necessary connection between the profitable verticals, and the
need for a new open source Lisp. On the face of it, it's highly unlikely
that any profitable business opportunity is so dependant on the state of
Common Lisp that it will fail in the current environment, but will succeed
wildly if only one of the commercial lisps were open source. (If for no
other reason than: if you really can make a ton of money, then that one
business can just buy a regular commercial license from your favorite
commercial lisp. Why do you need to care what happens to "the community"?)
3. Give more details to support your obvious disdain for existing open source
lisps, in comparison to existing commercial ones.
Right now, you're just asserting (without evidence) huge number of claims that
many people here disagree with, and then asking a bizarre and useless question
once we accept all your assumptions. While giving no details at all about
the most interesting part of your proposal (how to make a ton of money using
Lisp), and instead just dismissing that part with "I know how to do it, so
just believe me, but I don't want to talk about it with any of you."
It's hard to take you seriously.
-- Don
_______________________________________________________________________________
Don Geddis http://don.geddis.org/ ···@geddis.org
Based on what you know about him in history books, what do you think
Abraham Lincoln would be doing if he were alive today?
(1) Writing his memoirs of the Civil War.
(2) Advising the President.
(3) Desperately clawing at the inside of his coffin.
-- David Letterman
> 1. Realize that you get zero credit for "imagine I knew of a secret profitable
> vertical". Profitable businesses are very hard to make, and nobody is willing
> to just give you this as an assumption.
Fortunately, logic does not operate on a credit system. It isn't
logically necessary for me to give a mechanism for the antecedent of a
"What if..." question in order to analyze the consequent. If it were
required, no abstract reasoning would be possible at all. You may not
wish to invest in the undescribed scheme, or in spending the effort to
analyze the consequent that I have proposed, but your rhetorical
argument is specious. Oh, and you might consider an anger management
class.
From: Don Geddis
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <873b0m8vqw.fsf@geddis.org>
I wrote:
>> 1. Realize that you get zero credit for "imagine I knew of a secret
>> profitable vertical". Profitable businesses are very hard to make, and
>> nobody is willing to just give you this as an assumption.
·········@gmail.com" <········@gmail.com> wrote on Tue, 19 Jun 2007:
> Fortunately, logic does not operate on a credit system. It isn't
> logically necessary for me to give a mechanism for the antecedent of a
> "What if..." question in order to analyze the consequent. If it were
> required, no abstract reasoning would be possible at all.
Of course it isn't logically necessary. Did you even read what I wrote?
Do you remember my quote, "if wishes were horses, beggars could ride"?
You can imagine all sorts of things. What if gravity were to reverse
tomorrow? Could the Common Lisp community then unify under a single
implementation? Who knows? Who cares?
It is not LOGICALLY necessary to examine the plausibility of the antecedent.
But it is practically necessary. Your hypothetical is so extraordinarily
implausible, that there's not much value in arguing over what might happen in
that case. You're presenting the question as though it matters what the
answer is, as though the Lisp community can learn something about itself by
engaging in the analysis exercise you suggest.
Instead, given the implausibility of your assumptions, there is no value in
spending time exploring your hypothetical. This is not to say that it is
LOGICALLY inconsistent. Merely that it is practically useless.
> You may not wish to invest in the undescribed scheme, or in spending the
> effort to analyze the consequent that I have proposed, but your rhetorical
> argument is specious.
I also recommend that nobody else on this newsgroup analyze the consequent
that you have proposed. Or, at the least, only do it tongue in cheek, as
opposed to mistakenly believing that the result matters in the real world.
> Oh, and you might consider an anger management class.
My previous post was a rather long message, with many different points for
you to consider. Including a number of constructive suggestions at the end.
There was no anger anywhere in the message.
You pruned and ignored all of it, except the brief quote above. Why did you
choose to ignore the substance of all my comments? Why don't you step up and
actually address some of them? For example, I've asked many times -- and
you've yet to even acknowledge -- that you've established no connection at all
between your claimed profitable verticals, and any need for a new open source
Lisp.
I won't bother to repeat my other suggestions, since you'll surely just
ignore them again. But you ought to stop complaining that people aren't
willing to entertain your ideas, when you are the one that isn't engaging in
real discussion.
-- Don
_______________________________________________________________________________
Don Geddis http://don.geddis.org/ ···@geddis.org
Until you stalk and overrun, you can't devour anyone.
-- Tiger aphorism, by Hobbes (Calvin and Hobbes, 11-22-2005)
Don Geddis <···@geddis.org> wrote:
+---------------
| . You've proposed that somebody (unknown, rich), "buy out" one
| of the commercial implementations, and make it open source.
...
| . You've proposed -- but not explained even a little bit --
| that ... everybody would make it back on the "verticals".
+---------------
Hmmm... Where have I heard this business plan before? Oh, right...
To slightly tweak the flip chart from one of
the most famous episodes of South Park:
Step #1: Steal a commercial CL.
Step #2: ???
Step #3: Profits.
One might be safer investing in the underpants gnomes... ;-} ;-}
-Rob
-----
Rob Warnock <····@rpw3.org>
627 26th Avenue <URL:http://rpw3.org/>
San Mateo, CA 94403 (650)572-2607
> +---------------
> | . You've proposed that somebody (unknown, rich), "buy out" one
> | of the commercial implementations, and make it open source.
> ...
> | . You've proposed -- but not explained even a little bit --
> | that ... everybody would make it back on the "verticals".
> +---------------
>
> Hmmm... Where have I heard this business plan before? Oh, right...
>
> To slightly tweak the flip chart from one of
> the most famous episodes of South Park:
>
> Step #1: Steal a commercial CL.
>
> Step #2: ???
>
> Step #3: Profits.
>
> One might be safer investing in the underpants gnomes... ;-} ;-}
I honestly don't understand where you (all) are getting these sorts of
fabricated analyses. How does "buy out and open source X" even vaguely
equate to "steal X"? My proposal may not be realistic, but it's hardly
unethical. If I buy your house from you and give it to a homeless
person, that may be stupid, but I've not stolen it from you.
········@gmail.com <········@gmail.com> wrote:
+---------------
| > Hmmm... Where have I heard this business plan before? Oh, right...
| > To slightly tweak the flip chart from one of
| > the most famous episodes of South Park:
| >
| > Step #1: Steal a commercial CL.
| >
| > Step #2: ???
| >
| > Step #3: Profits.
| >
| > One might be safer investing in the underpants gnomes... ;-} ;-}
|
| I honestly don't understand where you (all) are getting these sorts of
| fabricated analyses. How does "buy out and open source X" even vaguely
| equate to "steal X"?
+---------------
(*sigh*) Kenny may be right. The humor of this group has all leaked away.
*SPOILER WARNING*
*SPOILER WARNING*
*SPOILER WARNING*
*SPOILER WARNING*...
Yes, my first cut at it read:
Step #1: Con an clueless VC into buying a commercial CL for you.
but since in the South Park episode the gnomes' flip chart said:
Step #1: Steal underpants.
I shortened it to match the style. [But then, conning a
clueless VC is *kinda* of "stealing", isn't it?]
Anyway, the exact details of "Step #1" are never important
in these fables. It's the totally missing "Step #2" and the
fantasy of "Step #3: Profits" that contain all of the humor.
Around Silicon Valley, it's even gotten to the point that
all you have to do when somebody at the table has begun
waxing a little too eloquent about their next great thing
is say "Step #3: Profits", and everyone present rolls on
the floor with laughter. [Except the speaker, of course,
who is mortally offended and never invites you to his
parties again. Oh, well...]
-Rob
-----
Rob Warnock <····@rpw3.org>
627 26th Avenue <URL:http://rpw3.org/>
San Mateo, CA 94403 (650)572-2607
On Jun 20, 2:43 pm, ····@rpw3.org (Rob Warnock) wrote:
> ········@gmail.com <········@gmail.com> wrote:
>
> +---------------
> | > Hmmm... Where have I heard this business plan before? Oh, right...
> | > To slightly tweak the flip chart from one of
> | > the most famous episodes of South Park:
> | >
> | > Step #1: Steal a commercial CL.
> | >
> | > Step #2: ???
> | >
> | > Step #3: Profits.
> | >
> | > One might be safer investing in the underpants gnomes... ;-} ;-}
> |
> | I honestly don't understand where you (all) are getting these sorts of
> | fabricated analyses. How does "buy out and open source X" even vaguely
> | equate to "steal X"?
> +---------------
>
> (*sigh*) Kenny may be right. The humor of this group has all leaked away.
>
> *SPOILER WARNING*
>
> *SPOILER WARNING*
>
> *SPOILER WARNING*
>
> *SPOILER WARNING*...
>
> Yes, my first cut at it read:
>
> Step #1: Con an clueless VC into buying a commercial CL for you.
>
> but since in the South Park episode the gnomes' flip chart said:
>
> Step #1: Steal underpants.
>
> I shortened it to match the style. [But then, conning a
> clueless VC is *kinda* of "stealing", isn't it?]
>
> Anyway, the exact details of "Step #1" are never important
> in these fables. It's the totally missing "Step #2" and the
> fantasy of "Step #3: Profits" that contain all of the humor.
>
> Around Silicon Valley, it's even gotten to the point that
> all you have to do when somebody at the table has begun
> waxing a little too eloquent about their next great thing
> is say "Step #3: Profits", and everyone present rolls on
> the floor with laughter. [Except the speaker, of course,
> who is mortally offended and never invites you to his
> parties again. Oh, well...]
>
> -Rob
>
> -----
> Rob Warnock <····@rpw3.org>
> 627 26th Avenue <URL:http://rpw3.org/>
> San Mateo, CA 94403 (650)572-2607
When we started our business we were completely broke.And now we have
10 millions worth debts.
Slobodan Blazeski
Sorry; Rob. I did get that you were joking, but given Don's angry
barrage regarding what became your Step #1, it seemed worth
clarification.
That said, and since you've now amplified your other steps, let me
humorlessly address those as well. You (everyone) seem(s) to be
focused on the antecedent of my question. Logically the content of the
antecedent is irrelevant (which is why I didn't put in an antecedent
to begin with), but since it seems like everyone has fixated on it,
okay, I'll provide an antecedent. (In the following, "you" refers not
to you, Rob, but to those who illogically want me to justify the
antecedent of my question; Don, for example.)
How am I planning to make enough money on a lisp-based vertical to
support a few engineers to gate-keep an open source lisp. Okay, I'll
reveal my evil plan for world domination: First I'm going to build an
spider that will crawl all linked-up web pages and build an inverted
index by words. Of course, it'll start small -- a few million pages --
but it'll eventually have billions of pages. Then I'll build a really
simple web interface that lets users put in keywords to search the
index. I'll use a really simple, well understood, heuristic to order
the hits that are returned to the user. How am I planning to make
money on this? I'll start just on venture funding, but soon I'll let
advertisers tell me what words they want their ads to appear on, and
I'll collect a few cents on click-though overhead. This will support
building out more and more index, and eventually additional products
like Email and spreadsheets, and a bunch of others, and eventually
I'll be making so much money for you, the investor, and my share
holders, and everyone else that you'll be burning the equivalent of
what it cost us to open source and support Lisp in your fireplace
every night, just for the warm glow that it gives your gigantic living
room.
Happy? No, I know, you're going to say: Well, maybe that was plausible
then, or not, but it worked, but anyway this is now. Again, I think
that this whole argument is pointless because justifying the
antecedent isn't required to analyze the consequent, but, okay, so
I'll pick any of the hundreds of Web 2.0 or search business plans that
have been funded in the past year. Happy now? No, I know, you're going
to say: But why is Lisp going to make this a success... at which point
I throw up my hands and realize that I'm arguing with people who just
don't get the logic of argument and are just arguing because they
simply don't want to think about the question I asked and would rather
just make up dumb reasons not to do so.
On Jun 20, 5:43 am, ····@rpw3.org (Rob Warnock) wrote:
> ········@gmail.com <········@gmail.com> wrote:
>
> +---------------
> | > Hmmm... Where have I heard this business plan before? Oh, right...
> | > To slightly tweak the flip chart from one of
> | > the most famous episodes of South Park:
> | >
> | > Step #1: Steal a commercial CL.
> | >
> | > Step #2: ???
> | >
> | > Step #3: Profits.
> | >
> | > One might be safer investing in the underpants gnomes... ;-} ;-}
> |
> | I honestly don't understand where you (all) are getting these sorts of
> | fabricated analyses. How does "buy out and open source X" even vaguely
> | equate to "steal X"?
> +---------------
>
> (*sigh*) Kenny may be right. The humor of this group has all leaked away.
>
> *SPOILER WARNING*
>
> *SPOILER WARNING*
>
> *SPOILER WARNING*
>
> *SPOILER WARNING*...
>
> Yes, my first cut at it read:
>
> Step #1: Con an clueless VC into buying a commercial CL for you.
>
> but since in the South Park episode the gnomes' flip chart said:
>
> Step #1: Steal underpants.
>
> I shortened it to match the style. [But then, conning a
> clueless VC is *kinda* of "stealing", isn't it?]
>
> Anyway, the exact details of "Step #1" are never important
> in these fables. It's the totally missing "Step #2" and the
> fantasy of "Step #3: Profits" that contain all of the humor.
>
> Around Silicon Valley, it's even gotten to the point that
> all you have to do when somebody at the table has begun
> waxing a little too eloquent about their next great thing
> is say "Step #3: Profits", and everyone present rolls on
> the floor with laughter. [Except the speaker, of course,
> who is mortally offended and never invites you to his
> parties again. Oh, well...]
>
> -Rob
>
> -----
> Rob Warnock <····@rpw3.org>
> 627 26th Avenue <URL:http://rpw3.org/>
> San Mateo, CA 94403 (650)572-2607
On Jun 20, 2:55 pm, ·········@gmail.com" <········@gmail.com> wrote:
> You (everyone) seem(s) to be
> focused on the antecedent of my question. Logically the content of the
> antecedent is irrelevant (which is why I didn't put in an antecedent
> to begin with), but since it seems like everyone has fixated on it,
> okay, I'll provide an antecedent.
Actually, that you can say this shows how doomed you are. Yes,
*logically* you can push any hypothesis you like, and logicians do
indeed spend a bunch of time doing that kind of thing. In the real
world when someone says "assume x, then let's see what follows" people
tend to immediately ask "how plausible is x", because that tells them
whether it is worth their time even thinking about what might follow
from x. If x is extremely unlikely, then what follows has to be
*very* important before it's worth thinking about. If x is impossible
then what follows is not worth even reading. If x is not specified
then the only interesting question is "what is x?".
If, for instance, you want to raise capital, you'll find that all they
care about is x.
--tim
On Jun 20, 2:53 pm, Tim Bradshaw <··········@tfeb.org> wrote:
> On Jun 20, 2:55 pm, ·········@gmail.com" <········@gmail.com> wrote:
>
> > You (everyone) seem(s) to be
> > focused on the antecedent of my question. Logically the content of the
> > antecedent is irrelevant (which is why I didn't put in an antecedent
> > to begin with), but since it seems like everyone has fixated on it,
> > okay, I'll provide an antecedent.
>
> Actually, that you can say this shows how doomed you are. Yes,
> *logically* you can push any hypothesis you like, and logicians do
> indeed spend a bunch of time doing that kind of thing. In the real
> world when someone says "assume x, then let's see what follows" people
> tend to immediately ask "how plausible is x", because that tells them
> whether it is worth their time even thinking about what might follow
> from x. If x is extremely unlikely, then what follows has to be
> *very* important before it's worth thinking about. If x is impossible
> then what follows is not worth even reading. If x is not specified
> then the only interesting question is "what is x?".
>
> If, for instance, you want to raise capital, you'll find that all they
> care about is x.
You people are soooooooo confused. Here X *is* the raising capital
part! I'm not TRYING to raise capital, I already HAVE capital; I'm
trying to figure out what to do with it!
From: Don Geddis
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <871wg62cbh.fsf@geddis.org>
·········@gmail.com" <········@gmail.com> wrote on Wed, 20 Jun 2007:
> You people are soooooooo confused. Here X *is* the raising capital
> part! I'm not TRYING to raise capital, I already HAVE capital; I'm
> trying to figure out what to do with it!
Fabulous news! We were confused, because you had neglected to mention
your circumstances.
I had offered a number of scenarios that might make your hypothetical more
interesting. One was, "a billionaire would like to fund this dream because
of a personal whim." Apparently, you're now saying that I wasn't too far off.
Since getting the free money in the first place was by far the hardest part
of everything you mentioned, why didn't you just say this was solved at the
beginning, and then ask for advice on the most effective use of the funds (if
promoting Common Lisp was your goal)?
In any case, I can bet people around here will be able to come up with lots
of ideas for using your money. Although I suspect that your original one, of
buying out a commercial CL and forcing it into open source, will not be at
the top of anyone else's list.
-- Don
_______________________________________________________________________________
Don Geddis http://don.geddis.org/ ···@geddis.org
Nothing in the world can take the place of Persistence. Talent will not;
nothing is more common than unsuccessful men with talent. Genius will not;
unrewarded genius is almost a proverb. Education will not; the world is full
of educated derelicts. Persistence and determination alone are omnipotent.
The slogan "Press On" has solved and always will solve the problems of the
human race.� -- Calvin Coolidge
········@gmail.com wrote:
> You people are soooooooo confused. Here X *is* the raising capital
> part! I'm not TRYING to raise capital, I already HAVE capital; I'm
> trying to figure out what to do with it!
If you already have some funds and you're asking for opinions on
how to use them, then you should state that up front. "I think I have
enough money to do Y. Does Y sound like a good way to promote the
language, or is there some other approach that might be better?"
Otherwise, you sound like an idle dreamer.
--
Dan
www.prairienet.org/~dsb/
On 21 Jun, 00:55, ·········@gmail.com" <········@gmail.com> wrote:
> You people are soooooooo confused. Here X *is* the raising capital
> part! I'm not TRYING to raise capital, I already HAVE capital; I'm
> trying to figure out what to do with it!
Awesome. Well, if you buy me cool stuff, I'll be happy to use it.
Even if I'm some what confused by your business plan.
> You people are soooooooo confused. Here X *is* the raising capital
> part! I'm not TRYING to raise capital, I already HAVE capital; I'm
> trying to figure out what to do with it!
Great, since this thread is dragging on and on I must well give you
the solution. Forget about buying Franz or LispWorks. You don't have
the money for that.
Hire three SBCL guys to work fulltime on core. Hire 3 other guys to
start working on the Eclipse killer (which includes the SWT-type
toolkit). Hire 3 other guys to figure out how "you" can leverage CLOS/
MOP and Macros to create the "Rails killer" (that's part of your
verticals strategy). And if you have anything left over, hire another
guy or two to implement sweet expressions for those that will never,
ever touch CL because of it's scary parantheses.
········@gmail.com wrote:
> You people are soooooooo confused. Here X *is* the raising capital
> part! I'm not TRYING to raise capital, I already HAVE capital; I'm
> trying to figure out what to do with it!
If you want to make CL more popular, you have to create the impression
that there is progress in the language. (Most news about other
programming languages are always along the lines of what features are
added in the newest release of a language.) If you want to make sure
that the progress is valuable for the existing CL users, it should
better be real progress.
Common Lisp is a language whose core is defined by a specification, not
by an implementation (as in the single-implementation languages like
Python, Ruby, Perl, what have you). The advantage of a
specification-driven language is that different vendors (open source and
commercial ones) can focus on different aspects of a specific
implementation and can even make trade offs that wouldn't be possible in
a single implementation. For example, CMUCL/SBCL traditionally focus on
a compiler yielding high performance whereas clisp is an bytecode-based
approach that results in a small footprint. In some cases, clisp can be
even the more efficient choice because its compiler is much faster,
which is beneficial for code that is executed only a few times (for
example, in scripts). These are just two examples - the other CL
implementations all have their respective advantages and disadvantages.
Focusing on one single implementation would reduce the value for CL
users, not increase it. (And the trend in the more popular
single-implementation languages goes towards diversification as well -
once the popularity of a language goes beyond a certain threshold,
people seem to want to have different implementation with specific
characteristics. However, they then have the strong disadvantage that
they cannot rely on a common core across all the implementations.)
The equivalent to a large number of libraries that work out of the box
in single-implementation languages, but in a specification-driven
setting, is a large number of specifications for various purposes.
Specifications that define how more specialized tasks can be tackled in
portable ways, but which are at the same time not as widely used enough
to warrant a change of the underlying core language. A diversification
in such "sub"-specifications can also reinforce one of the strengths of
a specification-driven language, namely the co-existence of alternative
approaches which may even contradict each other. For each concrete
application one can then choose the best combination for the task at
hand. Again, diversity is good.
Scheme seems to have had good experiences with SRFI - see
http://srfi.schemers.org/ - it has even reused some of the SRFIs as
sources for the upcoming revision of the core language itself. A process
like SRFI also does both: create the impression of progress - each SRFI
can be announced in newsgroups, blogs, etc. - while creating actual real
value for developers.
We have had two attempts at something similar for Common Lisp. CLRFI -
http://clrfi.alu.org/ - didn't take off because, IMHO, it had elements
that lead to deadlocks in the process itself. CDR -
http://cdr.eurolisp.org/ - is our own younger attempt and had a better
start. We already have a couple of entries (although far from being a
critical mass), and some others are in the works.
If you want to do something useful for the CL community as a whole, you
could do the following:
- Help identifying libraries or language extensions that could benefit
from (de-facto-)standardization.
- Pay developers / consultants to work out the respective specifications
and example implementations.
- Negotiate with commercial CL vendors to "buy out" some of their
vendor-specific extensions and make them publicly available as
specifications with portable or semi-portable implementations.
Especially the latter would be closest to your original proposal, if I
understand correctly. But also the first two are useful. Writing good
specifications takes time and skills and may benefit if it is not done
as a mere spare-time activity besides all the other work on or with
Common Lisp.
These are my �0.02. I give them here because I am deeply convinced that
this is the right way to go.
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
On Jun 21, 3:10 pm, Pascal Costanza <····@p-cos.net> wrote:
> - Negotiate with commercial CL vendors to "buy out" some of their
> vendor-specific extensions and make them publicly available as
> specifications with portable or semi-portable implementations.
>
> Especially the latter would be closest to your original proposal, if I
> understand correctly. But also the first two are useful. Writing good
> specifications takes time and skills and may benefit if it is not done
> as a mere spare-time activity besides all the other work on or with
> Common Lisp.
I would add that libraries need to not just be publicly available,
but easy to find and more or less bug free as well. I hope a project
like CLAPPA might be the first steps towards a "default library".
If I as a user could go to one location and have a choice of packages
or libraries that have been tested to work with each other and for
"all" implementations it would make getting started easier.
Hopefully it would also become self-reinforcing. If someone wants
other people to use his/her library, he better get it past the
CLAPPA gatekeepers since that is where people would go to find
libraries.
Gatekeepers would also function as a resource for people wanting
to submit their libraries. I imagine most people use a couple of
implementations at most, while the gatekeepers should have all the
implementations (open and commercial) and can help port libraries
to all implementations.
--
-asbjxrn
From: =?utf-8?b?R2lzbGUgU8ODwqZsZW5zbWk=?= =?utf-8?b?bmRl?=
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <0nodj4md21.fsf@kaktus.ii.uib.no>
Pascal Costanza <··@p-cos.net> writes:
> ········@gmail.com wrote:
>
> > You people are soooooooo confused. Here X *is* the raising capital
> > part! I'm not TRYING to raise capital, I already HAVE capital; I'm
> > trying to figure out what to do with it!
>
> If you want to make CL more popular, you have to create the impression
> that there is progress in the language. (Most news about other
> programming languages are always along the lines of what features are
> added in the newest release of a language.) If you want to make sure
> that the progress is valuable for the existing CL users, it should
> better be real progress.
It is not neccessary to change the language itself to archieve that goal.
If one made a bundle of an open source common lisp (say SBCL)with a lot of the
available open source packages out there and made it work out of the box,
that would be a great improvement. This would be much like you can download
python today and get a comprehensive library included in the download.
The Python community calls this "with batteries included".
Common lisp currently comes without batteries included, since most Lisp
implementations (the open source ones at least) comes with little more
than what is required by the standard. Since Lisp is
a "programmable programming language" to begin with one could even
include features that would have to be new language constructs in other
programming languages, like new looping construct (itterate), list
comprehension and a lot of other stuff.
Most of these things already exists. It is just that every developer has to
go around and find the libraries they need manually. This does likely put
some people off, and make them go to Python, Perl or Ruby instead.
A bundle of an open source common lisp that frequently adds new packages
could just be that constant progress you are talking about.
--
Gisle Sælensminde, Phd student, Scientific programmer
Computational biology unit, BCCS, University of Bergen, Norway,
Email: ·····@cbu.uib.no
The best way to travel is by means of imagination
In article <··············@kaktus.ii.uib.no>,
·····@kaktus.ii.uib.no (Gisle SÃõlensminde) wrote:
> Pascal Costanza <··@p-cos.net> writes:
>
> > ········@gmail.com wrote:
> >
> > > You people are soooooooo confused. Here X *is* the raising capital
> > > part! I'm not TRYING to raise capital, I already HAVE capital; I'm
> > > trying to figure out what to do with it!
> >
> > If you want to make CL more popular, you have to create the impression
> > that there is progress in the language. (Most news about other
> > programming languages are always along the lines of what features are
> > added in the newest release of a language.) If you want to make sure
> > that the progress is valuable for the existing CL users, it should
> > better be real progress.
>
> It is not neccessary to change the language itself to archieve that goal.
True. Especially for Common Lisp where you don't have to
wait for the 'benevolent dictator' to add another looping
construct.
> If one made a bundle of an open source common lisp (say SBCL)with a lot of the
> available open source packages out there and made it work out of the box,
> that would be a great improvement.
No, Lisp in a box does not exist.
http://common-lisp.net/project/lispbox/
http://www.gigamonkeys.com/lispbox/
> This would be much like you can download
> python today and get a comprehensive library included in the download.
> The Python community calls this "with batteries included".
> Common lisp currently comes without batteries included, since most Lisp
> implementations (the open source ones at least) comes with little more
> than what is required by the standard. Since Lisp is
You mean there is nothing like http://common-lisp.net/project/clbuild/ .
How sad.
Even worse. http://www.cliki.net/ASDF-Install does not exist.
> a "programmable programming language" to begin with one could even
> include features that would have to be new language constructs in other
> programming languages, like new looping construct (itterate), list
> comprehension and a lot of other stuff.
>
> Most of these things already exists. It is just that every developer has to
> go around and find the libraries they need manually. This does likely put
> some people off, and make them go to Python, Perl or Ruby instead.
You seem to have the impression that Common Lisp is similar
to Perl? It is not. You can do scripting of web pages with CL, but
that's not what it was designed for. It was designed
for very different application domains.
> A bundle of an open source common lisp that frequently adds new packages
> could just be that constant progress you are talking about.
Right, and a bundle of Lisp programmers capable of using google,
finding the projects and then contributing would be even better...
But then whining on comp.lang.lisp is so much easier.
Sorry for being a bit bitter...
--
http://lispm.dyndns.org
Rainer Joswig wrote:
>> It is not neccessary to change the language itself to archieve that goal.
>
> True.
How do you implement decent performance, callcc, tail recursion and most
other modern language features without altering the language?
--
Dr Jon D Harrop, Flying Frog Consultancy
The OCaml Journal
http://www.ffconsultancy.com/products/ocaml_journal/?usenet
On 2007-06-25, Jon Harrop <···@ffconsultancy.com> wrote:
> Rainer Joswig wrote:
>>> It is not neccessary to change the language itself to archieve that goal.
>>
>> True.
>
> How do you implement decent performance, callcc, tail recursion and
> most other modern language features without altering the language?
*sigh*
From: Ken Tilton
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <F9Tfi.12$_S6.2@newsfe12.lga>
Larry Clapp wrote:
> On 2007-06-25, Jon Harrop <···@ffconsultancy.com> wrote:
>
>>Rainer Joswig wrote:
>>
>>>>It is not neccessary to change the language itself to archieve that goal.
>>>
>>>True.
>>
>>How do you implement decent performance, callcc, tail recursion and
>>most other modern language features without altering the language?
>
>
> *sigh*
>
We'll let you off with a warning for that, Larry.
:)
kt
In article <··············@newsfe12.lga>,
Ken Tilton <···········@optonline.net> wrote:
> Larry Clapp wrote:
> > On 2007-06-25, Jon Harrop <···@ffconsultancy.com> wrote:
> >
> >>Rainer Joswig wrote:
> >>
> >>>>It is not neccessary to change the language itself to archieve that goal.
> >>>
> >>>True.
> >>
> >>How do you implement decent performance, callcc, tail recursion and
> >>most other modern language features without altering the language?
> >
> >
> > *sigh*
> >
>
> We'll let you off with a warning for that, Larry.
>
> :)
>
> kt
I think he has to be re-hypnotized. Show him the lambda sign.
--
http://lispm.dyndns.org
From: Ken Tilton
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <HfUfi.11$af7.4@newsfe12.lga>
Rainer Joswig wrote:
> In article <··············@newsfe12.lga>,
> Ken Tilton <···········@optonline.net> wrote:
>
>
>>Larry Clapp wrote:
>>
>>>On 2007-06-25, Jon Harrop <···@ffconsultancy.com> wrote:
>>>
>>>
>>>>Rainer Joswig wrote:
>>>>
>>>>
>>>>>>It is not neccessary to change the language itself to archieve that goal.
>>>>>
>>>>>True.
>>>>
>>>>How do you implement decent performance, callcc, tail recursion and
>>>>most other modern language features without altering the language?
>>>
>>>
>>>*sigh*
>>>
>>
>>We'll let you off with a warning for that, Larry.
>>
>>:)
>>
>>kt
>
>
> I think he has to be re-hypnotized. Show him the lambda sign.
>
I was going to have Frank buddy with him on one of those 12-step deals.
"Frank? Larry. I am dying to blow off Harrop with a one-worder."
"Let's pray for deliverance from this temptation, Lar."
"I tried that, didn't help."
"Well then think about being eviscerated by Kenny."
Pause.
"OK, that works."
hth,kzo
Rainer Joswig <······@lisp.de> writes:
> I think he has to be re-hypnotized. Show him the lambda sign.
This is the proper one:
http://www.xach.com/misc/lambda.html
(Created with zpb-ttf, cl-vectors, and skippy...)
Zach
In article <··············@unnamed.xach.com>,
Zach Beane <····@xach.com> wrote:
> Rainer Joswig <······@lisp.de> writes:
>
> > I think he has to be re-hypnotized. Show him the lambda sign.
>
> This is the proper one:
>
> http://www.xach.com/misc/lambda.html
>
> (Created with zpb-ttf, cl-vectors, and skippy...)
>
> Zach
Awesome, that'll work!
--
http://lispm.dyndns.org
Rainer Joswig wrote:
> In article <··············@unnamed.xach.com>,
> Zach Beane <····@xach.com> wrote:
>
>
>>Rainer Joswig <······@lisp.de> writes:
>>
>>
>>>I think he has to be re-hypnotized. Show him the lambda sign.
>>
>>This is the proper one:
>>
>> http://www.xach.com/misc/lambda.html
>>
>>(Created with zpb-ttf, cl-vectors, and skippy...)
Pretty frickin impressive.
>>
>>Zach
>
>
> Awesome, that'll work!
>
Oh, really? And when exactly do I get to eviscerate anyone?! Did you
think about that for one second?!!
I hate you guys sooo much.
kxo
ps. I hope Dr. <PWUAHAHAHHAHAHAHAHAHAHAHA> Harrop enjoys where this
thread is going. k
Zach Beane wrote:
> Rainer Joswig <······@lisp.de> writes:
>> I think he has to be re-hypnotized. Show him the lambda sign.
>
> This is the proper one:
>
> http://www.xach.com/misc/lambda.html
>
> (Created with zpb-ttf, cl-vectors, and skippy...)
Ran fine for 173 frames and then died with a run-time type error, whatever
that is.
--
Dr Jon D Harrop, Flying Frog Consultancy
The OCaml Journal
http://www.ffconsultancy.com/products/ocaml_journal/?usenet
On Jun 26, 9:09 am, Jon Harrop <····@ffconsultancy.com> wrote:
> Zach Beane wrote:
> > Rainer Joswig <······@lisp.de> writes:
> >> I think he has to be re-hypnotized. Show him the lambda sign.
>
> > This is the proper one:
>
> > http://www.xach.com/misc/lambda.html
>
> > (Created with zpb-ttf, cl-vectors, and skippy...)
>
> Ran fine for 173 frames and then died with a run-time type error, whatever
> that is.
could you post the backtrace of that gif?
/Ties
From: =?utf-8?b?R2lzbGUgU8ODwqZsZW5zbWk=?= =?utf-8?b?bmRl?=
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <0nps3i25cj.fsf@kaktus.ii.uib.no>
Rainer Joswig <······@lisp.de> writes:
> In article <··············@kaktus.ii.uib.no>,
> ·····@kaktus.ii.uib.no (Gisle SÃõlensminde) wrote:
>
> > Pascal Costanza <··@p-cos.net> writes:
> >
> > > ········@gmail.com wrote:
> > >
> > > > You people are soooooooo confused. Here X *is* the raising capital
> > > > part! I'm not TRYING to raise capital, I already HAVE capital; I'm
> > > > trying to figure out what to do with it!
> > >
> > > If you want to make CL more popular, you have to create the impression
> > > that there is progress in the language. (Most news about other
> > > programming languages are always along the lines of what features are
> > > added in the newest release of a language.) If you want to make sure
> > > that the progress is valuable for the existing CL users, it should
> > > better be real progress.
> >
> > It is not neccessary to change the language itself to archieve that goal.
>
> True. Especially for Common Lisp where you don't have to
> wait for the 'benevolent dictator' to add another looping
> construct.
>
> > If one made a bundle of an open source common lisp (say SBCL)with a lot of the
> > available open source packages out there and made it work out of the box,
> > that would be a great improvement.
>
> No, Lisp in a box does not exist.
>
> http://common-lisp.net/project/lispbox/
> http://www.gigamonkeys.com/lispbox/
While it exists, it seems like it's scope is rather limited, and there is just a
handful packages included. While I rather would see a similar package for SBCL,
Lispbox is a useful start and a valuable initiative, but is currently just too
limited as more than a headstart for newbies. All packages have to start small
however, so it can grow into being what I miss.
I must admit that I have not looked too closely on Lispbox simply because of the
impression I got above.
> Even worse. http://www.cliki.net/ASDF-Install does not exist.
While it too exists (and I even use it!), I miss a bundle that comes ready with
asdf install and pointing to a repository with packages. Yes you can do this if
you know about all this, but the package you can download and that just works, and
with nice documentation is not there yet.
In fact asdf-install vs the python library is more like apt-get vs rpm, in that
the first is a package manager and remote update/distribution system, while the
latter is merly a package system / library. It would be nice if one could have a
package that gave people new in the game a head start, and that I feel is missing.
> You seem to have the impression that Common Lisp is similar
> to Perl? It is not. You can do scripting of web pages with CL, but
> that's not what it was designed for. It was designed
> for very different application domains.
I have some problem to see a reasonable interpretation of what you say here. I
could understand it as:
1. Lisp is so special that you don't need to use libraries written by others.
2. Lisp is not intended to be used for web application, so you don't need libraries
like those in Perl/Python/Other scripting languages.
If you mean something else please protest. The first interpretation I find unreasonable.
The development of good packages don't only require good programmers, but also extensive
domain knowledge. The result of this can often be offered as software packages, and
can save a development team for significant effort. I refuse to accept that every lisp
developer is doing something so unique that he don't can find libraries written by others
useful.
When it comes to point 2, that will merly mean that a lisp package will come with
libraries for solving different tasks than those for perl/python/puby etc. That is not
unlikely, even though many still develop web applications.
But even if (2) is the case, you still need to access e.g. files written in file formats
from other software. Then it is useful to have a parser for that file format, even if
your goal is
>
> Right, and a bundle of Lisp programmers capable of using google,
> finding the projects and then contributing would be even better...
> But then whining on comp.lang.lisp is so much easier.
>
> Sorry for being a bit bitter...
It is always easier to complain on a newsgroup. My complaining was more meant constructivly
however:
1. You don't need to change the language (and reopen the standardisation can-of-worms) to
give an impression that there is development in the field. (Or to keep on having real
progress for that sake)
2. Most of the bits are already there (asdf, asdf-install, lispbox, even if I did not mention them)
we just need an effort to assemble them.
I tried to be positive and say - see, we are almost there. It could be interpreted otherwise as
well obviously.
> http://lispm.dyndns.org
--
Gisle Sælensminde, Phd student, Scientific programmer
Computational biology unit, BCCS, University of Bergen, Norway,
Email: ·····@cbu.uib.no
The best way to travel is by means of imagination
In article <··············@kaktus.ii.uib.no>,
·····@kaktus.ii.uib.no (Gisle S�lensminde) wrote:
...
> > Right, and a bundle of Lisp programmers capable of using google,
> > finding the projects and then contributing would be even better...
> > But then whining on comp.lang.lisp is so much easier.
> >
> > Sorry for being a bit bitter...
>
> It is always easier to complain on a newsgroup. My complaining was more meant constructivly
> however:
>
> 1. You don't need to change the language (and reopen the standardisation can-of-worms) to
> give an impression that there is development in the field. (Or to keep on having real
> progress for that sake)
>
> 2. Most of the bits are already there (asdf, asdf-install, lispbox, even if I did not mention them)
> we just need an effort to assemble them.
>
> I tried to be positive and say - see, we are almost there. It could be interpreted otherwise as
> well obviously.
There is a whole ecosystem around SBCL. There is lots of code.
There is a network aware defsystem, you get pre-assembled
libraries. That's a good start!
There is so much infrastructure already (IRC channel, Paste server,
project servers, Wiki, ...) that you can use now. There is not much
need to invent new stuff if you want to use this infrastructure.
It's there. You can use it now and contribute.
I use OpenMCL a bit more and that works fine for me, too. It can
even use a lot of the code that works with SBCL (and other Lisps).
--
http://lispm.dyndns.org
·········@gmail.com" <········@gmail.com> writes:
>
> You people are soooooooo confused. Here X *is* the raising capital
> part! I'm not TRYING to raise capital, I already HAVE capital; I'm
> trying to figure out what to do with it!
>
>
I was one of the earliest to respond to Jeff's original post and my response
was that I basically didn't understand what the benefits of creating another
open source CL would be and I basically don't believe it would have the
rallying impact or draw additional users to CL. I also questioned the benefits
of using CL to try and compete with Java, Python, ruby etc in a market where
they have already set the ground rules.
After watching this thread, I have to say I'm really beginning to see Jeff's
frustration with the irrelevant tangents and challenges raised by many of the
posters - much of which is based on assumptions that were not made in the
original post. The original question was quite simple and I think really just
raised the question of whether making one of the commercial CL implementations
open source would have any significant impact on the number of people using CL
in real world situations and possibly make it a real alternative for
developers. All the posts about venture capital, profit of verticals, how this
would be paid for, motivation, etc are totally irrelevant to the original
question.
for the record and to be totally clear, I don't believe it would have any real
impact on the number of people usinig CL or even make it a real consideration
for shops trying to decide what language to use as the basis for their next
multi-million dollar earning project. This is mainly because other issues, such
as CL FUD, exposure of programmers to CL etc will have a greater impact than
the quality of development environments or standard libraries.
Having said that, a truely smacko commercial quality CL development environment
that was open source would certainly be welcomed by many existing CL users,
though I think putting capital into one of the existing open source
implementations would be a better way to go.
Finally, I can fully understand the OPs increased agressive/frustrated
responses given some of the posts to this thread. However, I think now that
this frustration has possibly over shadowed some of the posts that were
genuinely attempting to address the original question and resulted in increased
defensiveness that is resulting in overly negative interpretation of some of
the posts or questions being asked by others who really seem to want to
understand the motivation underlying the original question (Kent's contribution
is possibly a good example of this).
My advice, for what it is worth, is just ignore responses from those who are
trying to drag the thread off into irrelevant tangents and only respond to
those who you feel add to the discussion in a positive way and remember that
this group is *not* a good representation of the CL community - it is just one
little part of it and many CL users I've met who actually do regularly develop
in CL consider it to be of only minor interest/relevance.
Tim
--
tcross (at) rapttech dot com dot au
EXECUTIVE SUMMARY: Cart. Horse. Before.
GORY DETAILS:
········@gmail.com <········@gmail.com> wrote:
+---------------
| How am I planning to make enough money on a lisp-based vertical
| to support a few engineers to gate-keep an open source lisp.
+---------------
You seem to keep missing the point that others have been *trying*
to convey, which is that you *don't* need an open-source Lisp to
accomplish that!! Given all the *other* costs of your "evil plan
for world domination", the cost of an excellent-quality commercial
Lisp is peanuts.
So... Do please go ahead and dominate the world [using a commercial
Lisp], and then go right on ahead use the outrageous profits that you
will reap to develop [or buy] a Lisp to open-source. Trying to do it
the other way round is just... (*grepping for a non-confrontational
way to express it*)... unlikely to succeed.
+---------------
| First I'm going to build an spider that will crawl all linked-up
| web pages and build an inverted index by words. ...eventually have
| billions of pages. Then I'll build a really simple web interface...
| really simple, well understood, heuristic to order the hits...
| How am I planning to make money on this? I'll start just on venture
| funding, but soon I'll let advertisers tell me what words they want
| their ads to appear on, and I'll collect a few cents on click-though
| overhead. This will support building out more and more [stuff]...
+---------------
Uh... dude... you *do* realize that you're just a few years late
to that particular party, hmmm?
http://www.google.com/corporate/execs.html
http://yhoo.client.shareholder.com/bios.cfm
+---------------
| ...but, okay, so I'll pick any of the hundreds of Web 2.0 or
| search business plans that have been funded in the past year.
| Happy now? No, I know, you're going to say: But why is Lisp
| going to make this a success... at which point I throw up my
| hands and realize that I'm arguing with people who just don't
| get the logic of argument and are just arguing because they
| simply don't want to think about the question I asked and would
| rather just make up dumb reasons not to do so.
+---------------
It's because your antecedent is *so* incredible ["beyond belief
or understanding"] that taking it seriously is ludicrous. It's
as if you had asked:
What would happen if Santa Claus were real and gave me
$500 million dollars, and I used it to buy a commercial Lisp
and make it open-sourced?
We could certainly all debate *that* "What would happen if", too,
but it would be a total waste of our time. And face it, few of us
have much time to spare in useless debate these days...
Look, if you have a way to make money with Lisp, *GO DO IT!!*
You don't need a pet open-sourced implementation to do that
with -- given the economics of a startup, a commercial CL is
just fine.[1] Then use the profits for whatever you like.[2]
-Rob
[1] Though one of the fine existing open-sourced implementations
might already be good enough for you, depending on your
business plan.
[2] Buying up some commercial CL & open-sourcing it seems like
an odd thing to do; but hey, it'll be your money doing it.
-----
Rob Warnock <····@rpw3.org>
627 26th Avenue <URL:http://rpw3.org/>
San Mateo, CA 94403 (650)572-2607
> +---------------
> | First I'm going to build an spider that will crawl all linked-up
> | web pages and build an inverted index by words. ...eventually have
> | billions of pages. Then I'll build a really simple web interface...
> | really simple, well understood, heuristic to order the hits...
> | How am I planning to make money on this? I'll start just on venture
> | funding, but soon I'll let advertisers tell me what words they want
> | their ads to appear on, and I'll collect a few cents on click-though
> | overhead. This will support building out more and more [stuff]...
> +---------------
>
> Uh... dude... you *do* realize that you're just a few years late
> to that particular party, hmmm?
No! Really!? Golly gee, Dr. Science! You don't say! Darn, well, I
guess it's back to the bat cave for me....
> +---------------
> | How am I planning to make enough money on a lisp-based vertical
> | to support a few engineers to gate-keep an open source lisp.
> +---------------
>
> You seem to keep missing the point that others have been *trying*
> to convey, which is that you *don't* need an open-source Lisp to
> accomplish that!! Given all the *other* costs of your "evil plan
> for world domination", the cost of an excellent-quality commercial
> Lisp is peanuts.
No, you (et al) seem to keep missing the point that I'm fully aware of
this; that I'm not doing it for ME -- I'm doing it for YOU! Yes, I can
buy a Lisp, but the 9000 kids in their parents' basement wanting to
start something on peanuts CAN'T -- and can't make due with the
confusing half-supported free Lisps, and so avoid Lisp when they can
get fully supported Ruby (etc), not to mention a modern web services
library and huge user community, all FOR FREE. Those kids are the only
thing that's going to keep Lisp alive, and they're going elsewhere; in
fact, they're pretty much already gone. If all I wanted to do was
build MY house, yeah, I can hire my own contractor and blow the
community, but, well, I guess I'd rather not see the rest of the
neighborhood become a slum, so I'm trying to figure out how to build
my house, and support the neighborhood too. It might not be a very
good idea to try to support the neighborhood in this particular way,
and that's worthy of discussion, but all I'm getting from you is meta-
analysis about why I can't even have that discussion....which is, as
far as I can tell, non sequitur drivel.
The only reason I'm bothering to engage in this noise is that I think
it's really really important to get advice here, but I can't see to
get you all to engage in that analysis, so ... well, fooey! See ya!
From: Don Geddis
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <871wg5qgx7.fsf@geddis.org>
·········@gmail.com" <········@gmail.com> wrote on Wed, 20 Jun 2007:
> the 9000 kids in their parents' basement wanting to start something on
> peanuts CAN'T -- and can't make due with the confusing half-supported free
> Lisps, and so avoid Lisp when they can get fully supported Ruby (etc), not
> to mention a modern web services library and huge user community, all FOR
> FREE.
I agree that those are attractive features of some other languages, and
Common Lisp would be stronger if the CL community could provide them as well.
> I guess I'd rather not see the rest of the neighborhood become a slum, so
> I'm trying to figure out how to build my house, and support the
> neighborhood too.
The discussion got derailed initially, because you didn't explain your
motivation. And the truth is, there really isn't much connection between
you building your own house, and supporting the neighborhood. So it was
a great distraction for you to have mentioned them together, as though they
necessarily were dependent on each other.
> It might not be a very good idea to try to support the neighborhood in this
> particular way, and that's worthy of discussion
OK, if that's your real question, I think that's a reasonable thing for people
here to discuss.
The comments you've gotten back have suggested almost universally that trying
to buy a commercial lisp and force it open source is not productive.
It is actually a strength of Common Lisp that there are multiple
implementations (with different engineering tradeoffs, licenses, and prices)
for the same specification.
That said, you are correct that the community support in CL is not on part
with that of some competing languages.
So if you have a bunch of extra money, and want to improve the community
support, you should simply work on that directly. Make a SRFI-like process
in CL, as a lightweight way to further the ANSI spec, perhaps building on the
CDR effort. Hire some full-time employees to make a Perl CPAN-like system,
with code that has passed some kind of review and is conveniently organized
into a single library, perhaps building on ASDF(-INSTALL). Make it easy for
someone brand new to install and get a working system, perhaps along the
lines of Lisp-In-A-Box. Etc. The "huge user community" is an advantage that
you can't work on directly, but may arise indirectly from working on the
other things.
All the community things can be done cross-implementation. Your focus on
getting the community to rally around a single implementation was misplaced,
as it would have been very expensive (the way you wanted to do it), would have
faced great resistance from the community, and at the end of the day wouldn't
have solved any of the real problems with Lisp vs. the competitors that you
mentioned.
Is that more constructive? Do you want to start this thread over, and just
ask for advice on how money (which you already have?) could be applied to
enhance the CL community?
-- Don
_______________________________________________________________________________
Don Geddis http://don.geddis.org/ ···@geddis.org
I think there should be a standard size of pillow, and if you want to
manufacture a pillow that's larger or smaller, that's fine, but it will not
receive the coveted title of "standard."
-- Deep Thoughts, by Jack Handey [1999]
On Jun 20, 11:15 pm, ·········@gmail.com" <········@gmail.com> wrote:
> No, you (et al) seem to keep missing the point that I'm fully aware of
> this; that I'm not doing it for ME -- I'm doing it for YOU! Yes, I can
> buy a Lisp, but the 9000 kids in their parents' basement wanting to
> start something on peanuts CAN'T -- and can't make due with the
> confusing half-supported free Lisps, and so avoid Lisp when they can
> get fully supported Ruby (etc), not to mention a modern web services
> library and huge user community, all FOR FREE. Those kids are the only
> thing that's going to keep Lisp alive, and they're going elsewhere; in
> fact, they're pretty much already gone.
This is important point that I tried to mention in other post:
there is a set of people who have money to buy commercial lisp, there
is also set of people who have time to study it, but those sets almost
do not intersect.
But a lot of people move from second set to first.
VisualWorks Smalltalk people figured it out and made their product
free for non-commercial use. It would be interesting to find out if it
helped them.
On 2007-06-21 00:15:32 -0400, ·········@gmail.com" <········@gmail.com> said:
> The only reason I'm bothering to engage in this noise is that I think
> it's really really important to get advice here, but I can't see to
> get you all to engage in that analysis, so ... well, fooey! See ya!
Just so you don't think everyone has missed the point, from post 0 of
this thread some of us were already aware from your previous posting
history that you have lots of capital from a successful lisp based
startup.
That being the case, I for one would be very happy if you bought a
commercial lisp and open sourced it, as I'm sure others would be.
However, the important question to ask, imho, is whether this is the
best way to invest your capital given your goal of an excellent open
source lisp to draw new users while simultaneously providing a platform
for your lisp based verticals. I think the only way to answer this is
to calculate the cost to bring an existing open source lisp up to an
acceptable level on the major platforms (windows, linux, mac os x, and
possibly *bsd) and compare this to the cost of buying an existing
commercial implementation.
Of course we would need to specify what we mean by "acceptable level."
The only way to answer this is to ask yourself "what is it about lisp
that could draw new users who would otherwise use (e.g.) python or
ruby?" To my mind, given that these are new users unfamiliar with lisp,
they will have only the vaguest notions of lisp's metaprogramming
facilities. What will jump out at them is the fact that lisp has a
repl, and native code compilation. In other words, to new users
unfamiliar with lisp, it appeals as an interactive, dynamic language
that compiles to fast native executables. They become familiar with the
metaprogramming piece later after they've already devoted some time to
learning and using the language.
To the extent that you think this makes sense, "acceptable level" must
mean a lisp that does native compilation. This eliminates clisp. It
also weighs in favor of implementations, such as sbcl, which focus on
generating very fast native code.
Other things newcomers will be attracted to, in no particular order are:
- the ability to build *native* gui apps for each platform (preferably
from a single code base, a la lispworks capi)
- a killer gui ide
- pre-packaged libraries in common use as a move toward de-facto
standardization
So compute the cost of bringing, say, sbcl up to this standard, and
compare it with the cost of acquiring lispworks or allegro and you'll
have some idea of how to proceed. Do some simple estimates of bringing
sbcl up to the above standards - possibly consulting the sbcl
developers to get an idea of person-hours and costs - and talk to the
boards/owners of lispworks and/or allegro to get a feel for how much
acquiring them would cost and choose the less costly option. My feel
for this is that throwing money at some knowledgeable sbcl developers
would get you where you want to go for significantly less.
Of course you may be more concerned with time than money. If this is
the case the fastest course of action would be to acquire lispworks
since they are the only implementation to my knowledge that can do
native gui apps on all three aforementioned platforms - allegro has no
mac gui. Given the mindshare the Mac OS X platform now has among
developers I think allegro would be the worst of both worlds - paying a
lot for a commercial lisp and then having to wait while developers
implemented a missing piece essential to attracting new users to lisp -
native gui apps.
From: Don Geddis
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <87myysoe2x.fsf@geddis.org>
Raffael Cavallaro <················@pas-d'espam-s'il-vous-plait-mac.com> wrote on Thu, 21 Jun 2007:
> I for one would be very happy if you bought a commercial lisp and open
> sourced it, as I'm sure others would be.
While I generally agreed with the rest of your post, I wanted to make a
comment about this sentence. You expressed it as though "of course" paying
money to make such a thing happen would necessarily be good for the Lisp
community.
I just wanted to point out that there are pros and cons to such an action,
independent of its cost. Kent Pitman, for example, has written often in the
past on this newsgroup about open source in general, and how it embodies
tradeoffs that are not necessarily universally good for society.
I think it's important, if such a thing were being contemplated, to consider
the downsides as well as the benefits of buying and open-sourcing an existing
commercial Common Lisp.
-- Don
_______________________________________________________________________________
Don Geddis http://don.geddis.org/ ···@geddis.org
"If we can put a man on the moon, why can't we put metal in a microwave!"
-- Dr. Frazier Crane, "Cheers"
In article
<····································@pasdespamsilvousplaitmaccom>,
Raffael Cavallaro
<················@pas-d'espam-s'il-vous-plait-mac.com> wrote:
> That being the case, I for one would be very happy if you bought a
> commercial lisp and open sourced it, as I'm sure others would be.
Really? I mean somebody buys a commercial Lisp vendor, open source
it and then it dies or stagnates? This possibility needs to
be prevented and I don't think it is that easy. I would also
not look at a time frame of one or two years of stability.
I would look for longer periods of time.
Looks like the commercial vendors have achieved something that
has not been achieved by open source Lisps, yet:
* Some commercial Lisp are running on a lot of platforms with
very similar feature set and performance.
Open source Lisps with native compilers are much more focused
on a few platforms and even there they offer not the same
features in similar quality.
* There is a extensive native IDE based on a cross-platform toolkit
from commercial vendors.
Open source Lisps don't have that currently.
and so on...
What then? Would the customers using this implementation like it?
Would potential customers like it?
I don't think you can somehow take something that has been
created and has evolved in a completely different context and
transplant it into an open source context just so. Not if the
product is complex like what we are talking about.
The open source Lisps have made lots of progress in the last years.
Why not base new work onto those and use an evolutionary
approach to incrementally improve then? Bring the infrastructure
into place step by step, develop libraries step by step, and so on.
This would also let people grow on these things over time.
I think it is more interesting how to cope with several
different implementations and still have portable
libraries without a maintenance nightmare. Plus the
libraries should be available for all Common Lisp
implementations (even the commercial ones).
But I also think that the range of Common Lisp development
environments is not complete. It should be possible to
develop SBCL and OpenMCL into these directions.
Gaps I see:
* there seems to be no small footprint native Lisp
for ARM processors. Mobile computing will
be even more important in the future.
* there is no successor to MCL
(with a cool Cocoa-based GUI)
Sigh.
* multi-core support is still kind of weak in general (with
only a few exceptions)
* open source application development (especially
with GUI) under Linux with CL is kind of uncommon.
That should be easier.
* Larger environments with integrated Lisp-based applications
that are easily customizable are still missing.
Think in the direction of Squeak or Lisp Machines...
If people find these things interesting, it will happen.
I think it is kind of more interesting to move forward, than
to battle scripting languages/environments on their ground...
--
http://lispm.dyndns.org
On 2007-06-22 05:43:57 -0400, Rainer Joswig <······@lisp.de> said:
> In article
> <····································@pasdespamsilvousplaitmaccom>,
> Raffael Cavallaro
> <················@pas-d'espam-s'il-vous-plait-mac.com> wrote:
>
>> That being the case, I for one would be very happy if you bought a
>> commercial lisp and open sourced it, as I'm sure others would be.
[1]
> Really? I mean somebody buys a commercial Lisp vendor, open source
> it and then it dies or stagnates? This possibility needs to
> be prevented and I don't think it is that easy. I would also
> not look at a time frame of one or two years of stability.
> I would look for longer periods of time.
Clearly this would be a `Bad Thing (TM)' but we're playing "what if
..." here, and Jeff's original scenario has his lisp based verticals
supporting the maintainece of this
to-be-formerly-commercial-now-open-source-common-lisp. Could his
proposed company die and end support for this lisp? Of course, but I'm
inclined to agree with Ron about the health of the commercial lisp
market, so I don't see this as a significantly greater risk than that
of the existing commercial vendor failing on its own due to a shrinking
lisp user base and/or reduction in contract work.
In effect Jeff's plan is really more of an acquisition of an existing
lisp vendor as a subsidiary while providing them with guaranteed
contract work - Jeff's lisp based verticals - to replace existing
for-pay work. There's no reason to believe that such a company would be
unable to transition to other contract work and paid support should
Jeff's business plan fail to pan out. They could always open source it
under a BSD or MIT style license which would allow them to take it
closed source again (cf. scieneer) if necessary.
[1] I could have worded this "as I'm sure some others would be" - but
that scans brokenly to a native speaker since the default meaning of
`others' in English is already `some others.' `Others' can be a bit
misleading in English - it doesn't imply an exhaustive partitioning in
ordinary use - one typically reaches for `everyone else' if that's the
meaning intended. "After the party broke up Bob and Nancy went to a
movie, while others went home," leaves open the distinct possibility
that yet a third group went to a bar or a nightclub. "After the party
broke up Bob and Nancy went to a movie, while everyone else went home,"
does not. In any event, I was only speaking for myself and like minded
people - I didn't say "as I'm sure *everyone else* would be."
Anyone notice that the commercial vendor's reps have not replied or
commented on any of this? It is kind of rude to talk about those
companies blood-sweat-and-tears in the third person like we are doing.
They are flesh and blood people and speculating what they "should be
doing" is nasty. Imagine going for dinner and everyone around the
table is discussing how you should live your life, who you should
marry, decide what kind of car you should drive.... Its their life
and they are free to live it.
If someone wants to be involved with vendors and want to move them in
an open source direction, then get involved with them (and make a
commitment). To have a stranger come in, buy a company and change
everything, would be signing its death sentence.
Wade
On Jun 22, 2:24 pm, Wade Humeniuk <········@telus.net.no.spam> wrote:
> Anyone notice that the commercial vendor's reps have not replied or
> commented on any of this?
I should think that's because they know just how plausible an idea it
is.
In article
<····································@pasdespamsilvousplaitmaccom>,
Raffael Cavallaro
<················@pas-d'espam-s'il-vous-plait-mac.com> wrote:
> On 2007-06-22 05:43:57 -0400, Rainer Joswig <······@lisp.de> said:
>
> > In article
> > <····································@pasdespamsilvousplaitmaccom>,
> > Raffael Cavallaro
> > <················@pas-d'espam-s'il-vous-plait-mac.com> wrote:
> >
> >> That being the case, I for one would be very happy if you bought a
> >> commercial lisp and open sourced it, as I'm sure others would be.
>
> [1]
>
>
> > Really? I mean somebody buys a commercial Lisp vendor, open source
> > it and then it dies or stagnates? This possibility needs to
> > be prevented and I don't think it is that easy. I would also
> > not look at a time frame of one or two years of stability.
> > I would look for longer periods of time.
>
>
> Clearly this would be a `Bad Thing (TM)' but we're playing "what if
> ..." here, and Jeff's original scenario has his lisp based verticals
> supporting the maintainece of this
> to-be-formerly-commercial-now-open-source-common-lisp. Could his
> proposed company die and end support for this lisp? Of course, but I'm
> inclined to agree with Ron about the health of the commercial lisp
> market, so I don't see this as a significantly greater risk than that
> of the existing commercial vendor failing on its own due to a shrinking
> lisp user base and/or reduction in contract work.
I have no idea if the user base is shrinking or not.
I also have no idea how many of the users are willing
to invest money into Lisp-based products.
comp.lang.lisp is still growing and compared to a few years
ago there is much more activity.
But I think a commercial vendor may only partly fail
due to a somehow shrinking user base. The other side is
that the vendor may grow or shrink based on what they
have to offer as products and services. You can stare
on the market and do nothing or you could develop and
sell products and services that people will buy. Only
because someone has a product that was cool twenty years
ago, you will not make many sales today.
The products, in this case related to a Lisp system,
is not the important thing. The important thing is to
look out for problems that customers have that can be
solved with these tools. Or solutions that can be improved.
If the vendor doesn't improve and adapt his 'offerings'
(I hate that word), his business will likely not expand.
Not even stay at the current level.
The question is how active has a vendor to be and where
should the vendor invest (marketing, research, infrastructure,
product development, education, ...)?
If we have a passive vendor, we know where it ends. Just
look at Digitool and MCL.
>
> In effect Jeff's plan is really more of an acquisition of an existing
> lisp vendor as a subsidiary while providing them with guaranteed
> contract work - Jeff's lisp based verticals - to replace existing
> for-pay work. There's no reason to believe that such a company would be
> unable to transition to other contract work and paid support should
> Jeff's business plan fail to pan out. They could always open source it
> under a BSD or MIT style license which would allow them to take it
> closed source again (cf. scieneer) if necessary.
>
>
> [1] I could have worded this "as I'm sure some others would be" - but
> that scans brokenly to a native speaker since the default meaning of
> `others' in English is already `some others.' `Others' can be a bit
> misleading in English - it doesn't imply an exhaustive partitioning in
> ordinary use - one typically reaches for `everyone else' if that's the
> meaning intended. "After the party broke up Bob and Nancy went to a
> movie, while others went home," leaves open the distinct possibility
> that yet a third group went to a bar or a nightclub. "After the party
> broke up Bob and Nancy went to a movie, while everyone else went home,"
> does not. In any event, I was only speaking for myself and like minded
> people - I didn't say "as I'm sure *everyone else* would be."
--
http://lispm.dyndns.org
On 2007-06-22 09:31:05 -0400, Rainer Joswig <······@lisp.de> said:
> If the vendor doesn't improve and adapt his 'offerings'
> (I hate that word), his business will likely not expand.
> Not even stay at the current level.
Well put - for example, your bullet point in a previous post about the
need to improve multi-core support in some existing lisps. This is
growing increasingly urgent if lisp wants to attract users who have
come to expect this in mainstream languages.
On Jun 21, 5:15 am, ·········@gmail.com" <········@gmail.com> wrote:
>
> No, you (et al) seem to keep missing the point that I'm fully aware of
> this; that I'm not doing it for ME -- I'm doing it for YOU! Yes, I can
> buy a Lisp, but the 9000 kids in their parents' basement wanting to
> start something on peanuts CAN'T -- and can't make due with the
> confusing half-supported free Lisps, and so avoid Lisp when they can
> get fully supported Ruby (etc), not to mention a modern web services
> library and huge user community, all FOR FREE.
You know, what's going to motivate those people (who, incidentally,
can afford a commercial Lisp if they care) is seeing someone have a
huge commercial success with Lisp. People don't use ruby because it's
free, they use it because other people do cool stuff and get rich.
> On Jun 21, 5:15 am, ·········@gmail.com" <········@gmail.com> wrote:
>
>> No, you (et al) seem to keep missing the point that I'm fully aware of
>> this; that I'm not doing it for ME -- I'm doing it for YOU! Yes, I can
>> buy a Lisp, but the 9000 kids in their parents' basement wanting to
>> start something on peanuts CAN'T -- and can't make due with the
>> confusing half-supported free Lisps, and so avoid Lisp when they can
>> get fully supported Ruby (etc), not to mention a modern web services
>> library and huge user community, all FOR FREE.
I think you're either forgetting or are too young to remember how few
commercial or non-commercial C++ compilers supported the C++ standard,
for many years. MSVC++ 6.0 had some particularly egregious bugs
(unexplained compiler errors, sometimes silent generation of erroneous
code), though a full commercial version around 1998 would set you back
$1000. MS was basically not interested in fixing the problem because of
an issue with their third-party C++ Standard Library. That didn't stop
us (a semi-academic software house where I worked in 2000-2002) from
producing pretty decent software around MSVC++ 6.0, and it didn't stop
my classmates and I from doing our C++ homeworks when we were in high
school and undergrad.
How is, say, ECL not supported? The main developer always gives me
useful feedback whenever I have a question or report a bug. When I
reported a bug in the Franz ACL 8.1 beta, I got a nice "thank-you"
e-mail, but no patch (maybe they posted one to their home page, but they
didn't notify me directly). Not that Franz isn't good with customer
service, but I'm just saying that commercial doesn't necessarily imply
better service.
The best hackers will always find the right tools for their job -- or
make their own if they want them badly enough (e.g., Perl, Python, ...).
Like Donald Knuth with TeX, they are willing to spend a lot of time on
side projects that have potential to make their main work and the work
of many others much easier. In contrast, the larger commercial compiler
vendors tend to be conservative -- they like to buy up innovators (like
Intel buying KAI) or fund academics to do the research work.
Think about this: who's writing parallel languages and producing
parallel language compilers nowadays? Mostly academics -- because they
are willing to take risks and invest time into forward-looking research
projects. (Think of how many parallel languages never took off! -- HPF
is a major example.) The main vendors then pick up on some of these
(like Cray with UPC), and/or invest in the academics, and then produce
more commercial products.
Cool language stuff happens mostly because academics or hackers create
something new. Then the cool language stuff becomes "mainstream" and
commercial.
mfh
> On Jun 21, 5:15 am, ·········@gmail.com" <········@gmail.com> wrote:
>
>> No, you (et al) seem to keep missing the point that I'm fully aware of
>> this; that I'm not doing it for ME -- I'm doing it for YOU! Yes, I can
>> buy a Lisp, but the 9000 kids in their parents' basement wanting to
>> start something on peanuts CAN'T -- and can't make due with the
>> confusing half-supported free Lisps, and so avoid Lisp when they can
>> get fully supported Ruby (etc), not to mention a modern web services
>> library and huge user community, all FOR FREE.
I think you're either forgetting or are too young to remember how few
commercial or non-commercial C++ compilers supported the C++ standard,
for many years. MSVC++ 6.0 had some particularly egregious bugs
(unexplained compiler errors, sometimes silent generation of erroneous
code), though a full commercial version around 1998 would set you back
$1000. MS was basically not interested in fixing the problem because of
an issue with their third-party C++ Standard Library. That didn't stop
us (a semi-academic software house where I worked in 2000-2002) from
producing pretty decent software around MSVC++ 6.0, and it didn't stop
my classmates and I from doing our C++ homeworks when we were in high
school and undergrad.
How is, say, ECL not supported? The main developer always gives me
useful feedback whenever I have a question or report a bug. When I
reported a bug in the Franz ACL 8.1 beta, I got a nice "thank-you"
e-mail, but no patch (maybe they posted one to their home page, but they
didn't notify me directly). Not that Franz isn't good with customer
service, but I'm just saying that commercial doesn't necessarily imply
better service.
The best hackers will always find the right tools for their job -- or
make their own if they want them badly enough (e.g., Perl, Python, ...).
Like Donald Knuth with TeX, they are willing to spend a lot of time on
side projects that have potential to make their main work and the work
of many others much easier. In contrast, the larger commercial compiler
vendors tend to be conservative -- they like to buy up innovators (like
Intel buying KAI) or fund academics to do the research work.
Think about this: who's writing parallel languages and producing
parallel language compilers nowadays? Mostly academics -- because they
are willing to take risks and invest time into forward-looking research
projects. (Think of how many parallel languages never took off! -- HPF
is a major example.) The main vendors then pick up on some of these
(like Cray with UPC), and/or invest in the academics, and then produce
more commercial products.
mfh
On Jun 18, 8:38 pm, ·········@gmail.com" <········@gmail.com> wrote:
> That's fair enough, but then why would you (for example) engage in the
> conversation at all? Some folks seem to think that it's clear enough,
> at least possibly with subsequent clarifying posts, to engage in a non-
> meta conversation about it.
This is Usenet's Lisp newsgroup. When it diverges from narrow
technical subjects, the conversation becomes notoriously
unconstructive. I'm unaware of any exception.
So I'd think you'd want to broach this topic differently, given that
reality. If this forum is even the right place for it.
Tayssir
|> Appears you just want to generate idle talk; As an
|> investor, are you treating "community" as human
|> "commodity market"? Your ideas on using opensource
|> indicate you do
|
| I have no idea what your last sentence means, but the one before is
| completely wrong. I am not trying to generate idle talk, I'm trying
| to get opinions as to what would happen if ... etc.
|
| I'm sorry but I just don't understand why so many people seem to think
| that there's something wrong with asking a simple "What if" question.
| Am I speaking a foreign language or something? Is "What if" somehow a
| new form in English? If I asked: "What would happen if the sun was to
| explode tomorrow?" you might think it an odd question, but you'd have
| no trouble answering it. You all seem to think that I have some evil
| intent,
You are a smart man (I've been told). You are asking questions which
by their very nature do not admit useful answers. This leads to the
question: Why are you asking these questions in the first place?
What is the "leveraging of community" you have in mind if your What-If
scenario played out by some chance?
Elsewhere you wrote:
> I claim that the optimal strategy is to give away the engine in
> order to garner an open source community for it, and make your money
> where you can.
From which I understood that your ideas for making money had to do
with treating the lisp community that'd gather around your offering as
a commodity.
Still elsewhere you wrote:
> What I wanted (although you'd have to study the thread to get this)
> is a company that supports an excellent CL on- the-side and makes
> their money through verticals, like Google supports Guido to do
> python, and Sun/Java.
google's revenue, I understood, comes from people clicking on it's
webpages. AFAICT This is the advertising industry, not tech industry.
For this they don't need PHDs but people using googlegroups instead of
NNTP etc, more the idle talk, more revenue for google. something like
that.
What sort of new market were you envisioning creating out of the lisp
community?
> Elsewhere you wrote:
> > I claim that the optimal strategy is to give away the engine in
> > order to garner an open source community for it, and make your money
> > where you can.
>
> From which I understood that your ideas for making money had to do
> with treating the lisp community that'd gather around your offering as
> a commodity.
Ah. I see what you are saying. It's possible that you're right, or
that you're misunderstanding me (probably because I wasn't too clear).
I don't mean that I would be making money on the backs of the Lisp
community. The idea was to make money by selling verticals; "using"
the Lisp community as support is, in a sense, the commoditization of
the community -- I see you're point now -- but they are not being
taken advantage of because they get back from that effort an improved
Lisp....at least this is the open source development model. That said,
I'm not personally a proponent of that model; as I said before, I
think that it's confused, and in part precisely the way you point out:
That people contribute more than they get back, and vertical
productizers get the advantage without giving back (the verticals).
But that is the open source development model; I didn't make it up.
> Still elsewhere you wrote:
> > What I wanted (although you'd have to study the thread to get this)
> > is a company that supports an excellent CL on- the-side and makes
> > their money through verticals, like Google supports Guido to do
> > python, and Sun/Java.
>
> google's revenue, I understood, comes from people clicking on it's
> webpages. AFAICT This is the advertising industry, not tech industry.
> For this they don't need PHDs but people using googlegroups instead of
> NNTP etc, more the idle talk, more revenue for google. something like
> that.
Why does it matter what sort of industry it is? It's a vertical built
on a great deal of open source freeware which they, in small part,
support the development of. I think that this is the same model I'm
proposing. (And, again, I didn't make it up.)
> What sort of new market were you envisioning creating out of the lisp
> community?
A focused open source development market -- if you want to call that a
market. Right now it's an unfocused partly open source development
market.
I'm not entirely sure that I've understood you, so it's possible that
we'll have to go around again on this. Sorry if I'm not answering your
questions as precisely as you'd like.
In article <···············@mid.individual.net>,
Pascal Costanza <··@p-cos.net> wrote:
> Ron Garret wrote:
>
> > Problem statement version 2: The Lisp economy is shrinking.
>
> You cannot possibly know that (unless you have done an extensive market
> study, but I don't believe you did).
One needs only survey the landscape of Lisp vendors over the last twenty
years to see that the long-term trend is down. There used to be a dozen
commercial Lisp vendors, now there are two. As for recent trends, when
there are only two vendors, it is not necessary to conduct "extensive
market research" to get a pretty clear picture of what's going on, at
least in the commercial world. Just having a chat with the larger of
the two vendors lets you survey more than 50% of the market.
Now, when companies are in financial trouble they usually do everything
they can to prevent that information from being widely known because
when people know you are desperate it makes it much harder to sell
products (customers wonder if you're going to be around to support them)
and raise capital. Since my goal here is to solve the problem rather
than contribute to it, if I were in possession of such inside
information I would be undermining my own goals (to say nothing of the
vendor's confidences) by revealing it. But if you really don't believe
me when I say I am not engaging in idle speculation you can contact me
privately and we can discuss this further.
> I guess what you actually want to say is this: The Lisp economy is not
> growing as much as you wish it did; it may even be shrinking.
OK, the Lisp economy *may* be shrinking. I have reason to believe that
it is. Is this really worth quibbling over?
rg
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <u645orgiu.fsf@nhplace.com>
Ron Garret <·········@flownet.com> writes:
> In article <···············@mid.individual.net>,
> Pascal Costanza <··@p-cos.net> wrote:
>
> > Ron Garret wrote:
> >
> > > Problem statement version 2: The Lisp economy is shrinking.
> >
> > You cannot possibly know that (unless you have done an extensive market
> > study, but I don't believe you did).
>
> One needs only survey the landscape of Lisp vendors over the last twenty
> years to see that the long-term trend is down. There used to be a dozen
> commercial Lisp vendors, now there are two.
Uh, first of all, might this not equally well be DUE to the
availability of free alternatives? How many vendors do there need to
be? ESPECIALLY given that Lisp is available free from so many
sources. If it weren't, and if the world didn't have free software,
the market might support more Lisp vendors. I'm not trying to make a
political point here (I'll make those elsewhere), just a practical
observation: Since it's not in fashion to buy language software any
more, that takes money out of the market for, specifically, vended
lisps. Consequently, I think it's a remarkably STRONG statement about
Lisp that it can continue to charge and continue to make money.
What is important about Lisp is not that it has waxed and waned at
various times but rather that it has survived the many strong and
focused attempts to define it away as irrelevant, and it keeps on
kicking. It clearly is not the language that is most prominently in
display on the shelves in book stores, but for a language not helped
by that kind of thing, it has remarkable resilience and health.
The market is very well-served by the vendors now, so if you phrase it
as you have there, it sounds bad, but if you phrase it differently, it
doesn't. What has gone away are two things:
* The super-high-end Lisp Machine vendors (not enough Lisp Machine hardware
sales, and many other factors not related to the lisp market)
* The very low-cost Lisp vendors (e.g, Gold Hill) probably somewhat
supplanted by free implementations
* The Lisps that were inevitably tied to markets of a particular machine
that went away (e.g., DEC's VAXLISP).
But other than that, I don't think this really has changed much. The
big ones used to be Lucid, Franz, Symbolics, and later Harlequin.
Symbolics went away, and Lucid sold its assets to Harlequin (renaming
them to Liquid). Through various transformations and hand-offs, the
Harlequin assets became LispWorks assets, and both Liquid CL and
LispWorks are available from LispWorks Ltd. So not that much has
changed. Those products are largely still there ... except Symbolics.
> As for recent trends, when
> there are only two vendors, it is not necessary to conduct "extensive
> market research" to get a pretty clear picture of what's going on, at
> least in the commercial world. Just having a chat with the larger of
> the two vendors lets you survey more than 50% of the market.
The same is probably true for most major commercial languages. I'm not
sure why that is a damning statistic.
> Now, when companies are in financial trouble they usually do everything
> they can to prevent that information from being widely known because
> when people know you are desperate it makes it much harder to sell
> products (customers wonder if you're going to be around to support them)
> and raise capital.
Your scenario is true. But so is the claim that "I am a truthteller."
If I were a liar, I would want to lie about that. So is my claiming
to tell truth a proof that I am a liar? This kind of proof technique
you're citing is weak and needless.
And, as you yourself note, if the case is otherwise and you don't
think it's advisable to talk about in the open (you said something
vague and weird about it that I haven't tried to interpret--I'm just
speaking to your remark. I have no extra data on that..), then don't
allude to it either. Public speculation of that kind is damaging if
it's right and doubly damaging if it's right.
> [...]
> OK, the Lisp economy *may* be shrinking. I have reason to believe that
> it is. Is this really worth quibbling over?
IMO, no. But not because I agree. Just because parts of this
discussion seem strange in the first place.
Let's discuss the ACTUAL issues. They're relevant WHETHER OR NOT the
market is shrinking or growing.
If someone can't get a good or service, if someone can't exchange
something, if someone can't get to market with an end-user product,
those are examples of problems. Worrying about vendor health and
market size seems to me not relevant. Some healthy markets are small
and some unhealthy ones are large.
But most thing anyone wants to complain about sounds like a business
opportunity. And I don't entirely understand why people would discuss
legit business opportunities online in public, if they were really
likely to work out, rather than just go implement them and make money.
But I'm listening just in case.
Kent M Pitman <······@nhplace.com> notes [amongst other things]
> What is important about Lisp is not that it has waxed and waned at
> various times but rather that it has survived the many strong and
> focused attempts to define it away as irrelevant, and it keeps on
> kicking. ...
kent,
was there a particular incident, some series of essays (say) or some
other series of unfortunate events that make you refer to "strong and
focused attempts to define it away as irrelevant?" i am not being
facetious, just curious. underbelly of the programming language
landscape is always fascinating.
oz
---
http://nextbit.blogspot.com | don't count your chickens in glass houses
york u. computer science | until the cows come home. -- david vestal
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <ufy4rmj3k.fsf@nhplace.com>
ozan s yigit <··@red.cs.yorku.ca> writes:
> Kent M Pitman <······@nhplace.com> notes [amongst other things]
>
> > What is important about Lisp is not that it has waxed and waned at
> > various times but rather that it has survived the many strong and
> > focused attempts to define it away as irrelevant, and it keeps on
> > kicking. ...
>
> kent,
> was there a particular incident, some series of essays (say) or some
> other series of unfortunate events that make you refer to "strong and
> focused attempts to define it away as irrelevant?" i am not being
> facetious, just curious. underbelly of the programming language
> landscape is always fascinating.
I mostly meant AI winter.
This, to a large degree, was precipitated by failures similar to
the dot com boom. People have overpromised results and needed a
scapegoat. I think there were actually particular news headlines
where various AI projects announced they had failed and that Lisp
was to blame, but I might be misremembering. There was an AI
conference in which the keynote address had a slide with someone
or something in a noose and the caption was "Lisp is Dead".
My thought, both then and now, was "Sure. And if these AI projects
had been done in C or C++, we'd be knee-deep in AI by now." Various
of those were rescued by people repenting and saying C++ was the way,
and really all that bought was time to regroup organizationally, but
since the project then won, it was seen as confirmation of the claim.
It was sort of like finding a corporate scapegoat to fire when a
firm is found to be in trouble... sometimes it's the person who did it,
but often it's not. The same as true of .com's. You saw companies
claiming they'd developed new models of economics where you didn't have
to make money to succeed, and then, wonder of wonders, they failed.
And what was the rallying cry? "dot com". Not "bad accounting".
Not "overpromising". Not "vaporware". "dot com". Why? Not because
a .org would have done better. But it gave a name to the Devil and that
name was not the people who'd overpromised--it drew attention from the
real cause, and let people survive and move on with only "acceptable
casualties".
But also, even since AI Winter, and that was mostly begun in the late
80's and carried into the 90's, there are the on and off incidents
where it's not enough for another language to succeed--people are
constantly trying to discredit Lisp, which is more proof that Lisp is
still the game to beat than a serious problem with Lisp. It seems to
bother people who claim to be happy in other languages a great deal
that others are not following them--the fact that we're still using
Lisp seems a problem to be rectified.
I happen to not be religious, but I don't try to convert people I know
who are genuinely religious. They have their beliefs and I have mine.
If they try to tell me I'm doing the wrong thing, I might tell them in
return that I don't think so. But I don't try to schedule time in my
day to tell them they're wrong. The world can do worse than having
people be happy for independent, even mutually incompatible,
reasons... as long as they aren't trying to hurt one another. But I
do think there are those who don't think this way, and who won't be
happy succeeding in their language unless they've proven that we're
not happy in ours.
I finally finished reading The Fountainhead, and while I don't
recommend it for either its ending or its philosophy, I count it a
masterwork of presentation and well worth reading for its characters
and situations. The character of Keating has much the attitude I'm
describing as he interacts with Roark--the need not only to succeed
but to know that the success caused others to fail--the utter
inability to conceive success as objectively defined, which I think
was Ayn Rand (the author)'s point. Those who have not read it, but
who enjoy tales that capture a set of characters in thorough detail
that reveals useful things about the nature of man, are well-advised
to have a peek. I've described it as a kind of "A Man for All
Seasons" but without the religion... a quick web search says I'm not
the only person to make comparisons between the two. And the issues
are similar, so if you're familiar with that play, and you can see how
it's not enough for Henry to re-marry, but he must have More's
approval to be happy, you know what I'm getting at. I wouldn't in
that play describe the issues as any single action of Henry's, but
rather a pattern of discontent directed steadily at More over time.
That's a bit stream of consciousness, and I'm afraid also abstract.
But does it help?
> That's a bit stream of consciousness, and I'm afraid also abstract.
> But does it help?
thanks for the summary vis a vis ai winter and its effects on the
language, than and now. i had not thought about that. [i think daniel
crevier, in his "ai: the tumultuous history of the search for artificial
intelligence" discusses lisp as a factor in the failures; i will have to
revisit the book for the details.] as for rand, so far all i can handle
from her is a small book on writing.
oz
--
http://nextbit.blogspot.com || york u. computer science
you take a banana, you get a lunar landscape. -- j. van wijk
On Jun 16, 12:37 am, Kent M Pitman <······@nhplace.com> wrote:
> Ron Garret <·········@flownet.com> writes:
> > One needs only survey the landscape of Lisp vendors over the last twenty
> > years to see that the long-term trend is down. There used to be a dozen
> > commercial Lisp vendors, now there are two.
>
> Uh, first of all, might this not equally well be DUE to the
> availability of free alternatives? How many vendors do there need to
> be? ESPECIALLY given that Lisp is available free from so many
> sources. If it weren't, and if the world didn't have free software,
> the market might support more Lisp vendors.
What the hell? Free software lisps predated commercial ones, and have
existed continuously since. If I'm not screwing anything up here,
Franz Lisp came out in the late 70's as a part of BSD. CMUCL emerged
from SPICE Lisp around 84/85. CLISP came in 1987.
And really, not that much has changed since, aside from Digitool being
paid to open source the core of MCL. The major free lisps you would
use at work are SBCL, CLISP, and OpenMCL.
Ron Garret <·········@flownet.com> wrote:
+---------------
| One needs only survey the landscape of Lisp vendors over the
| last twenty years to see that the long-term trend is down.d
| There used to be a dozen commercial Lisp vendors, now there are two.
+---------------
Hunh?!? I count at least five, without thinking hard:
- Franz (ACL)
- Lispworks (LW)
- Digitool (MCL)
- Corman
- Scieneer
and maybe also:
- Clozure Associates (OpenMCL development/support/consulting)
Why do you say only two?
-Rob
-----
Rob Warnock <····@rpw3.org>
627 26th Avenue <URL:http://rpw3.org/>
San Mateo, CA 94403 (650)572-2607
In article <································@speakeasy.net>,
····@rpw3.org (Rob Warnock) wrote:
> Ron Garret <·········@flownet.com> wrote:
> +---------------
> | One needs only survey the landscape of Lisp vendors over the
> | last twenty years to see that the long-term trend is down.d
> | There used to be a dozen commercial Lisp vendors, now there are two.
> +---------------
>
> Hunh?!? I count at least five, without thinking hard:
>
> - Franz (ACL)
> - Lispworks (LW)
> - Digitool (MCL)
> - Corman
> - Scieneer
>
> and maybe also:
>
> - Clozure Associates (OpenMCL development/support/consulting)
>
> Why do you say only two?
I don't count Digitool because as far as I can tell they are out of
business. I used to not count Corman because it was a one-man
operation, but I see they've expanded. I had completely forgotten about
Scieneer, but at $2000-$15,000 per license they don't do much to help
startups running on a shoestring (which is the context of this thread).
But I stand corrected: there are apparently somewhere in the
neighborhood of three and a half Lisp vendors.
rg
Ron Garret wrote:
> In article <································@speakeasy.net>,
> ····@rpw3.org (Rob Warnock) wrote:
>
>> Ron Garret <·········@flownet.com> wrote:
>> +---------------
>> | One needs only survey the landscape of Lisp vendors over the
>> | last twenty years to see that the long-term trend is down.d
>> | There used to be a dozen commercial Lisp vendors, now there are two.
>> +---------------
>>
>> Hunh?!? I count at least five, without thinking hard:
>>
>> - Franz (ACL)
>> - Lispworks (LW)
>> - Digitool (MCL)
>> - Corman
>> - Scieneer
>>
>> and maybe also:
>>
>> - Clozure Associates (OpenMCL development/support/consulting)
>>
>> Why do you say only two?
>
> I don't count Digitool because as far as I can tell they are out of
> business. I used to not count Corman because it was a one-man
> operation, but I see they've expanded. I had completely forgotten about
> Scieneer, but at $2000-$15,000 per license they don't do much to help
> startups running on a shoestring (which is the context of this thread).
As a matter of curiosity (my Lisp experience doesn't go back that far),
how many of the dozen commercial Lisp vendors from twenty years ago were
selling to "startups running on a shoestring"? And which of them had
offerings substantially better than the current state of the open source
lisps? Adjusting as necessary for the increase in computer power we've
seen since 1987, of course ...
My impression from what others have said is that your 20 year baseline
puts us in the middle of the AI boom, and if the AI boom was anything
like the first dotcom boom the number of vendors then may have been
inflated over whatever we'd expect a sustainable level to be.
-dan
In article <··················@proxy02.news.clara.net>,
Daniel Barlow <···@coruskate.net> wrote:
> Ron Garret wrote:
> > In article <································@speakeasy.net>,
> > ····@rpw3.org (Rob Warnock) wrote:
> >
> >> Ron Garret <·········@flownet.com> wrote:
> >> +---------------
> >> | One needs only survey the landscape of Lisp vendors over the
> >> | last twenty years to see that the long-term trend is down.d
> >> | There used to be a dozen commercial Lisp vendors, now there are two.
> >> +---------------
> >>
> >> Hunh?!? I count at least five, without thinking hard:
> >>
> >> - Franz (ACL)
> >> - Lispworks (LW)
> >> - Digitool (MCL)
> >> - Corman
> >> - Scieneer
> >>
> >> and maybe also:
> >>
> >> - Clozure Associates (OpenMCL development/support/consulting)
> >>
> >> Why do you say only two?
> >
> > I don't count Digitool because as far as I can tell they are out of
> > business. I used to not count Corman because it was a one-man
> > operation, but I see they've expanded. I had completely forgotten about
> > Scieneer, but at $2000-$15,000 per license they don't do much to help
> > startups running on a shoestring (which is the context of this thread).
>
> As a matter of curiosity (my Lisp experience doesn't go back that far),
> how many of the dozen commercial Lisp vendors from twenty years ago were
> selling to "startups running on a shoestring"?
Well, startups were not quite the fashion back then that they are today,
but I remember at least two commercial implementations with price tags
in the low $100s: Coral CL (which became Apple CL which became MCL) and
Golden CL. There may have been others. I was younger then and not
really paying all that much attention.
To be fair, Corman Lisp is in that price range. I've never used it
because it's Windows only. (I just now downloaded a copy and tried to
run it under Parallels. It actually looks fairly promising. I may have
to revise my worldview.)
> And which of them had
> offerings substantially better than the current state of the open source
> lisps?
MCL to my mind, even in its current state (which has not been updated in
years now) is still unsurpassed as an IDE IMHO. (Disclaimer: I have not
done an exhaustive survey of IDE's so it is possible that there's
something better out there. But I doubt it.)
> My impression from what others have said is that your 20 year baseline
> puts us in the middle of the AI boom, and if the AI boom was anything
> like the first dotcom boom the number of vendors then may have been
> inflated over whatever we'd expect a sustainable level to be.
Of course. But the AI crash was a lot longer ago than the dotcom crash,
and the dotcom world seems to be recovering a lot faster than the Lisp
world.
rg
On Sat, 16 Jun 2007 12:44:20 -0700, Ron Garret <·········@flownet.com> said:
| In article <··················@proxy02.news.clara.net>,
| Daniel Barlow <···@coruskate.net> wrote:
| ...
|| My impression from what others have said is that your 20 year baseline
|| puts us in the middle of the AI boom, and if the AI boom was anything
|| like the first dotcom boom the number of vendors then may have been
|| inflated over whatever we'd expect a sustainable level to be.
| Of course. But the AI crash was a lot longer ago than the dotcom crash,
| and the dotcom world seems to be recovering a lot faster than the Lisp
| world.
To compare oranges to oranges, shouldn't the last-but-one word of
the preceding sentence be "AI", rather than "Lisp"? And then it
would be interesting to measure the amount of AI in games, for
example, in order to assess the recovery of the AI world.
---Vassil.
--
The truly good code is the obviously correct code.
Ron Garret <·········@flownet.com> writes:
> In article <··················@proxy02.news.clara.net>,
> Daniel Barlow <···@coruskate.net> wrote:
>
>> Ron Garret wrote:
>> > In article <································@speakeasy.net>,
>> > ····@rpw3.org (Rob Warnock) wrote:
>> >
>> >> Ron Garret <·········@flownet.com> wrote:
>
>> My impression from what others have said is that your 20 year baseline
>> puts us in the middle of the AI boom, and if the AI boom was anything
>> like the first dotcom boom the number of vendors then may have been
>> inflated over whatever we'd expect a sustainable level to be.
>
> Of course. But the AI crash was a lot longer ago than the dotcom crash,
> and the dotcom world seems to be recovering a lot faster than the Lisp
> world.
>
This may have something to do with the idifferent nature of the two crashes. To
some extent, I think the AI crash was largely a result of failure to come up
with the goods - lots of hype, promises and high expectations, but at the end,
fundamental problems that had been there since the 60s were still unsolved. The
dot com crash was a bit different - that was more about over expectation in
return and over speculation regarding demand/uptake. Many of the dot com
failures had good ideas and even good products, but iether were too early (as
in before their time), over estimated their worth or just didn't manage their
cashflow well. The crash brought some sanity back and I think represented a
'correction' in the market. The AI crash was more about investers deciding
after years of high investment that no real outcomes were being produced and
therefore further investment was unlikely to provide any real returns. CL has
suffered more due to its close association and perception as an AI language.
Tim
--
tcross (at) rapttech dot com dot au
Ron Garret wrote:
> In article <···············@mid.individual.net>,
> Pascal Costanza <··@p-cos.net> wrote:
>
>> Ron Garret wrote:
>>
>>> Problem statement version 2: The Lisp economy is shrinking.
>> You cannot possibly know that (unless you have done an extensive market
>> study, but I don't believe you did).
>
> One needs only survey the landscape of Lisp vendors over the last twenty
> years to see that the long-term trend is down. There used to be a dozen
> commercial Lisp vendors, now there are two. As for recent trends, when
> there are only two vendors, it is not necessary to conduct "extensive
> market research" to get a pretty clear picture of what's going on, at
> least in the commercial world. Just having a chat with the larger of
> the two vendors lets you survey more than 50% of the market.
Maybe we don't need that many vendors to have a healthy market. I know
of economically viable languages where there is only one vendor, and I
even of economically viable languages where there is even none.
>> I guess what you actually want to say is this: The Lisp economy is not
>> growing as much as you wish it did; it may even be shrinking.
>
> OK, the Lisp economy *may* be shrinking. I have reason to believe that
> it is. Is this really worth quibbling over?
No, it's not worth quibbling over, but it makes a difference.
There is this whole issue of seeing half-empty glasses vs. half-full
glasses and also that of self-fulfilling prophecies involved as soon as
it comes to subjective assessments.
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
Pascal Costanza <··@p-cos.net> writes:
> Ron Garret wrote:
>> In article <···············@mid.individual.net>,
>> Pascal Costanza <··@p-cos.net> wrote:
>>
>>> Ron Garret wrote:
>>>
>>>> Problem statement version 2: The Lisp economy is shrinking.
>>> You cannot possibly know that (unless you have done an extensive market
>>> study, but I don't believe you did).
>>
>> One needs only survey the landscape of Lisp vendors over the last twenty
>> years to see that the long-term trend is down. There used to be a dozen
>> commercial Lisp vendors, now there are two. As for recent trends, when there
>> are only two vendors, it is not necessary to conduct "extensive market
>> research" to get a pretty clear picture of what's going on, at least in the
>> commercial world. Just having a chat with the larger of the two vendors lets
>> you survey more than 50% of the market.
>
> Maybe we don't need that many vendors to have a healthy market. I know of
> economically viable languages where there is only one vendor, and I even of
> economically viable languages where there is even none.
>
In fact, I cannot think of many commercial language implementations where there
is a lot of vendor competition. I don't see as many ads from different language
vendors as you saw in the 80s and early 90s. I suspect this may be a market
which has consolidated somewhat, which doesn't seem surprising, but more a
reflection of what generally happens in most markets.
--
tcross (at) rapttech dot com dot au
Pascal Costanza <··@p-cos.net> writes:
>> Problem statement version 2: The Lisp economy is shrinking.
> You cannot possibly know that (unless you have done an extensive
> market study, but I don't believe you did).
well, in the absence of an "extensive market study", you could use
the recent o'reilly book market stats as a reasonable substitute.
[+/- some accuracy factor for books vs language economy mismatch]
alas i cannot find recent published reports on actual university
teaching of programming languages to supplement the book data.
[you can use some of the earlier stats and project from them
which would certainly also support his statement]
python and ruby have gutted lisp and scheme, leaving best
wishes behind...
oz
---
http://nextbit.blogspot.com | york u. dept of computer science
the unfortunate and inevitable concomitant of "bring it on" is
"how do you like it now?" -- david mamet
Ron Garret <·········@flownet.com> writes:
> In article <··············@lion.rapttech.com.au>,
> Tim X <····@nospam.dev.null> wrote:
>
>> I think you missed Kent's point. You cannot just say this is the problem and
>> here is the solution. We need to discuss/debate what the problem is until we
>> find agreement on the problem definition before putting forward a solution
>
> At first I thought you were being sarcastic, but upon reflection I
> realize that not only are you being serious, you are actually correct.
> Hence:
>
>> Therefore, I feel we need to
>> start with a much more basic problem definition/question - why is CL not
>> being considered as a viable development platform for startups?
>
> That's a question, not a problem statement. Let me try again:
>
> Problem statement version 2: The Lisp economy is shrinking.
>
> (My proposed solution still applies.)
>
> rg
I don't disagree with what you propose being possibly part of the solution, but
I still think there are more fundamental issues needing to be addressed. The
establishment/creation/existance of a really smacko open source lisp
implementation is certainly beneficial - I just don't know to what extent it
will increase the numbers of lisp users - I think this is a more complex issue.
I do think your problem statement is more accurate - now we need to examine why
it is shrinking. If we can identify the core issues underlying that, then we
can identify possible strategies for addressing it (though I have to admit at
still being skeptical we can clearly identify the reasons and then define
effective strategies to change the situation).
Tim
--
tcross (at) rapttech dot com dot au
BTW...
In article <··············@lion.rapttech.com.au>,
Tim X <····@nospam.dev.null> wrote:
> why is CL not being
> considered as a viable development platform for startups?
I can offer up a data point from my recent experience. For the past
year I have been involved in two startups, both of which could have
benefited from using Lisp, and neither of which used it. In one of
those startups I was a founder and the CTO.
In both cases the limiting factor was ultimately the same: we couldn't
find any Lisp programmers to hire. I actually negotiated a deal with
one of the Lisp vendors where they would provide licenses AND
programming labor at cost, but then an investor came along and poured an
obscene amount of money into the company, and contract-labor-for-equity
was no longer a good deal so the whole thing fell through, and now we're
using C++ instead. (The irony is truly staggering.)
In any case, despite the fact that the plan was thwarted by having too
much money (which is ultimately a nice problem to have I must admit) I
still think it's a winning strategy and someone ought to try it again
while there are still some Lisp vendors out there to try it with.
rg
Ron Garret <·········@flownet.com> writes:
> In both cases the limiting factor was ultimately the same: we couldn't
> find any Lisp programmers to hire.
Well, this could of course mean that they're all employed...
But our experience is that they're easy to find.
--
(espen)
In article <··············@gazonk.netfonds.no>,
Espen Vestre <·····@vestre.net> wrote:
> Ron Garret <·········@flownet.com> writes:
>
> > In both cases the limiting factor was ultimately the same: we couldn't
> > find any Lisp programmers to hire.
>
> Well, this could of course mean that they're all employed...
>
Yes, all six of them ;-)
> But our experience is that they're easy to find.
Maybe I should move to Norway.
rg
From: Johannes Groedem
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <krbqfhb6dt.fsf@mort.netfonds.no>
* Espen Vestre <·····@vestre.net>:
>> In both cases the limiting factor was ultimately the same: we
>> couldn't find any Lisp programmers to hire.
> Well, this could of course mean that they're all employed...
> But our experience is that they're easy to find.
Not everyone has the Gnus/GMANE propaganda machine working for
them. :)
--
johs
Johannes Groedem <·········@netfonds.no> writes:
> Not everyone has the Gnus/GMANE propaganda machine working for
> them. :)
Heh :) But our largest competitor on the lisphead-hunting market in
the city seems to have attracted some brilliant lispheads as well.
--
(espen)
Espen Vestre <·····@vestre.net> writes:
> Ron Garret <·········@flownet.com> writes:
>
>> In both cases the limiting factor was ultimately the same: we couldn't
>> find any Lisp programmers to hire.
>
> Well, this could of course mean that they're all employed...
>
> But our experience is that they're easy to find.
> --
I wonder how much of this is dependent on region/location. In Australia, lisp
is not widely taught. Most of the better institutions will include it in a
course that also looks at other languages, such as prolog etc. In reality,
students get a couple of weeks exposure, but no real time to get a true
feel/taste or the language. Its sort of presented as an interesting, but
largely academic, alternative to procedural languages like C/Pascal or OO
languages like Java, C++ and ruby.
Tim
--
tcross (at) rapttech dot com dot au
Tim X <····@nospam.dev.null> writes:
> I wonder how much of this is dependent on region/location. In Australia, lisp
> is not widely taught.
Neither is it here in Norway, just a few universities offer it, and
not necessarily as part of a CS or engineering study. I've taught lisp
at university myself, but that was in Germany - and at a
comp. linguistics department.
> In reality, students get a couple of weeks exposure, but no real
> time to get a true feel/taste or the language.
These days, my impression is that for many young hackers, emacs spurs
the interest in the lisp family, but also perl, ruby and python helps
making curious young programmers aware of lisp.
--
(espen)
Espen Vestre wrote:
> These days, my impression is that for many young hackers, emacs spurs
> the interest in the lisp family, but also perl, ruby and python helps
> making curious young programmers aware of lisp.
What could be useful would be articles like "Lisp for Ruby programmers",
"Lisp for Pythonistas", etc. pp., on blogs or so, by people who know
both languages respectively.
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
From: Peder O. Klingenberg
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <kssl8tk2fv.fsf@beto.netfonds.no>
Pascal Costanza <··@p-cos.net> writes:
> What could be useful would be articles like "Lisp for Ruby
> programmers", "Lisp for Pythonistas", etc. pp., on blogs or so, by
> people who know both languages respectively.
At a previous employer, I once gave a talk called "Lisp for Perl
hackers" for an internal audience. It was generally well received,
but I don't think I won (m)any converts. And the source for that talk
is probably buried in the source control of said previous employer, so
I doubt I will be publishing it anywhere.
...Peder...
--
Sl�v uten dop.
Espen Vestre wrote:
> Tim X <····@nospam.dev.null> writes:
>
>> I wonder how much of this is dependent on region/location. In Australia, lisp
>> is not widely taught.
> [...]
> Neither is it here in Norway
> These days, my impression is that for many young hackers, emacs spurs
> the interest in the lisp family, but also perl, ruby and python helps
> making curious young programmers aware of lisp.
And how many of Perl, Ruby and Python are formally taught in
universities? My guess is "not much". Curious young programmers will
investigate anything they know about and can get their hands on, they
won't wait for a syllabus to spoonfeed it to them.
-dan
On Jun 15, 7:29 am, Daniel Barlow <····@coruskate.net> wrote:
> And how many of Perl, Ruby and Python are formally taught in
> universities? My guess is "not much".
I think Python is becoming more popular in universities now. But I
think the larger point is that it became popular first, then started
being taught in universities. So popularity leads to being taught in
a university, not the other way around.
-jimbo
>>>>> "DB" == Daniel Barlow <···@coruskate.net> writes:
DB> And how many of Perl, Ruby and Python are formally taught in
DB> universities? My guess is "not much". Curious young
DB> programmers will investigate anything they know about and can
DB> get their hands on, they won't wait for a syllabus to
DB> spoonfeed it to them.
When I was a curious college student programmer, I found that when my
other option was C, Lisp was incredibly useful.
Now that years have passed, I find that most of the advantages that
Lisp had over C can be found in other languages. Not all of them, to
be sure, and the implementations may not be as nice -- but I can get
80%-90% of the bang I need for a lot less politics and advocacy than
it would take to get Lisp adopted.
Charlton
--
Charlton Wilbur
·······@chromatico.net
Daniel Barlow <···@coruskate.net> writes:
> And how many of Perl, Ruby and Python are formally taught in
> universities? My guess is "not much". Curious young programmers will
> investigate anything they know about and can get their hands on, they
> won't wait for a syllabus to spoonfeed it to them.
Exactly. And the lisp programmers you want to employ, are a little of
both, I think: Educated, but also curious and easily
self-taught. Still, I think there are quite a few decent java
programmers by now, that started with java in University and never
looked back. But some (many?) of them may be too much of
inside-the-box thinkers to be usable for a lisp project.
--
(espen)
On Jun 15, 2:27 am, Ron Garret <·········@flownet.com> wrote:
> In both cases the limiting factor was ultimately the same: we couldn't
> find any Lisp programmers to hire.
Did you look? I've been writing commercial Lisp software, using
Lispworks, for six years now. If you advertised your position, and I
saw it, I guess I was either uninterested or unwilling to take the
risk. And, assuming you advertised, I'd be willing to bet other Lisp
developers had similar impressions.
-- david
In article <························@o61g2000hsh.googlegroups.com>,
"David E. Young" <··········@gmail.com> wrote:
> On Jun 15, 2:27 am, Ron Garret <·········@flownet.com> wrote:
> > In both cases the limiting factor was ultimately the same: we couldn't
> > find any Lisp programmers to hire.
>
> Did you look?
Of course. We even advertised in this newsgroup:
http://groups.google.com/group/comp.lang.lisp/msg/f638fa5d322434c8
among other places.
rg
From: Ken Tilton
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <qWIci.73$aq2.71@newsfe12.lga>
Ron Garret wrote:
> In article <························@o61g2000hsh.googlegroups.com>,
> "David E. Young" <··········@gmail.com> wrote:
>
>
>>On Jun 15, 2:27 am, Ron Garret <·········@flownet.com> wrote:
>>
>>>In both cases the limiting factor was ultimately the same: we couldn't
>>>find any Lisp programmers to hire.
>>
>>Did you look?
>
>
> Of course. We even advertised in this newsgroup:
>
> http://groups.google.com/group/comp.lang.lisp/msg/f638fa5d322434c8
>
> among other places.
>
"Senior applications developer: Our advanced applications work is
currently being done in Common Lisp, but that could change depending on
who we find to fill this position."
Oh, yeah, sign me up for a company that has such a tight grasp on
technology that its new hires decide the platform. Reminds me of the
singles ad I saw. "SWM, 36, seeks partner. Current orientation is
heterosexual, but that could change depending on who answers this ad".
Considering that one of the principals is a confirmed Lisp-self-loather,
this ad was DOA. ITA having hired most known Lispniks does not help either.
hth,kt
Ken Tilton <···········@optonline.net> writes:
> Considering that one of the principals is a confirmed
> Lisp-self-loather, this ad was DOA.
The ex-Googler blog, where Ron described how unpleasant he was to work
with, probably wasn't a good recruiting tool either.
Zach
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <uvedoplvk.fsf@nhplace.com>
Ron Garret <·········@flownet.com> writes:
> In article <························@o61g2000hsh.googlegroups.com>,
> "David E. Young" <··········@gmail.com> wrote:
>
> > On Jun 15, 2:27 am, Ron Garret <·········@flownet.com> wrote:
> > > In both cases the limiting factor was ultimately the same: we couldn't
> > > find any Lisp programmers to hire.
> >
> > Did you look?
>
> Of course. We even advertised in this newsgroup:
>
> http://groups.google.com/group/comp.lang.lisp/msg/f638fa5d322434c8
>
> among other places.
Quoting myself from a recent post about what people really mean when they
say there are no available Lisp programmers:
| (e) "I advertised a Lisp job once and no one signed up for the terms
| I offered, so I'm interpreting that as if there are no people out
| there at all"
( http://groups.google.com/group/comp.lang.lisp/msg/c9eb08c610e5f7d3 )
To expand on that, drawing from your job ad, and adding what would have
gone through my head if I had seen this ad and had been looking for a job
at the time...
* Subject = "Lisp/Python programmers (among others) wanted"
"/" often means "and", not just "or", so you may have lost
some people here. Not everyone programs in Python, nor wants to.
* "If you are looking for a stable 9-5 job, look elsewhere."
Nothing wrong with saying this. Nothing wrong with such jobs.
But if you're wanting to say there are no Lisp programmers because
none replied to an offering with this disclaimer, you might want to
consider at least the possibility that they didn't reply because of this.
* "Our advanced applications work is currently being done in Common Lisp,
but that could change depending on who we find to fill this position."
Any experienced person would take this to mean "we're going with what
the strongest voice in our group says", and by implication "if you are
not going to be the strongest voice, expect this to change due to forces
outside of your control". Hence, not only does this job involve Python
but no one doing the hiring is committed to it involving Common Lisp.
* "Smart Charter is located in the Los Angeles area. Telecommuting
is a possibility, but would not be our first choice."
Now your data says you tried only in the LA area. In part because
that's where you are. In part because you said people from farther
away will be treated as "not preferred". You should understand that
a startup, which is often short of money, may have to make harsh
decisions. If one is taking an insecure job it's bad enough already,
but being the one that is "not the first choice of how to work with"
makes it likely you'll be the first to get the axe. As if just being
far away and WANTED wouldn't already be a strike against you, which at
some companies it is. Telecommuting is ALWAYS higher risk than
in-office work. Under other circumstances, some people might move to
take an interesting job, even at their own expense, but rarely for a
startup and rarely for one not committed to the reason for hire. Jobs
in one's own neighborhood last much often less time than they used to.
* "If you are the sort of person who has always wanted an opportunity
to prove that Haskell or OCaml is the Best Way to write software,
this could be your chance."
The ad suggests multiple hires. So it's tipping you off that unless
all your CL friends are hired, it's likely to shape up to be a war
between people about what the best language is, rather than an attempt
to dive in and get things done.
* Coming back to that phrase "is the Best Way to write software" ...
Just that phrase is a strong tip-off that the outfit might not succeed.
Maybe it even did succeed--that's not my point. To someone evaluating
an offer, and deciding whether to apply, the question is "could it", and
what can you tell about the company from the offer.
Any company focused on "language" as "Best Way" rather than
"competent to do" is risking failing, IMO. Although I think it
would be fun to be working on CL, I think the thing you want to
know is that (a) the technical leaders have a plan, (b) the
technical leaders are making good decisions on the basis of sound
information, and (c) the technical leaders are willing to make
hard calls.
I have worked in team situations where a competent technical lead
made a very strict decisions about both language (C++) and coding
style (I believe it was best described as a modified Hungarian
notation), in a context where it was needed in order to maintain a
tight deadline, and they were good decisions that I concurred
with. I've also worked in team situations where team leaders made
decisions on whims just because they had a preference and thought
everyone should agree for no appaernt special reason and it was
largely disastrous (creating extra work, discontent, etc). The
thing that distinguishes the goodness of the two situations is not
that I got my way, nor that Lisp was used. It was that in one
case the decision was motivated by actual business need and
competent in its effect, and in the other case there was no
presented motivation and the effect was not helpful.
So as one reads ads, one looks to see what signals the would-be
leaders are sending by their wording. Certainly I do. And it's
easy to see this wording as sending the cue "this is a place for
us to compete on how we write code" rather than "this is a place
where we'll have lots to do and we'll just care that you get your
part done". Even just wording that said "We're planning to use CL
but if it doesn't get us to where we need to be, we might use
something else." sends the subtly different message that "success
in achieving business need" is the bar, rather than writing style.
Again, if I were writing, I'd have said "If you are the sort of
person who has always wanted to show that you can make Haskell or
OCaml deliver on someone's business needs, this could be your
chance." And once one mention languages in a competitive way, one
might want to add something vaguely along the lines of, "But our
goal is to get out product, not to spar over tools, so if you're a
language or tool evangelist, check it at the door. We'd love to
hear new ideas, but no one wants to be told they're doing things
wrong just because they're not using your favorite tool." Those
sorts of remarks set a good tone that, while strict, says "we are
planning to succeed, not just get distracted".
So let's sum up, shall we?
A very interesting job was offered that might have been attractive
and fun to someone living in the local area who happened to not have
a family to feed, such that they could afford to take a risk. The
offer was generously, and perhaps over-optimistically, but not very
realistically shared with the larger world. The offer indicates
strongly that it's a small group, and that many are not using
Common Lisp, and that in fact it's not committed to that. No problem.
The offer is narrowly tuned to the local area, and even then it's risky.
You're probably paying lots of stock to compensate for all that risk.
That's 100% fine. Sounds like a cool thing. Sounds fair on your part.
But...
Maybe indeed no one applied, or no one who did was who you
wanted. But it just doesn't sound like you can conclude that a
failure to find someone is adequately summed up by "There are no
lisp programmers out there." The top 118 obvious reasons for not
applying don't have anything to do with Lisp.
In article <·············@nhplace.com>,
Kent M Pitman <······@nhplace.com> wrote:
> Ron Garret <·········@flownet.com> writes:
>
> > In article <························@o61g2000hsh.googlegroups.com>,
> > "David E. Young" <··········@gmail.com> wrote:
> >
> > > On Jun 15, 2:27 am, Ron Garret <·········@flownet.com> wrote:
> > > > In both cases the limiting factor was ultimately the same: we couldn't
> > > > find any Lisp programmers to hire.
> > >
> > > Did you look?
> >
> > Of course. We even advertised in this newsgroup:
> >
> > http://groups.google.com/group/comp.lang.lisp/msg/f638fa5d322434c8
> >
> > among other places.
[Much snippage]
> But it just doesn't sound like you can conclude that a
> failure to find someone is adequately summed up by "There are no
> lisp programmers out there."
If you go back and look at what I said you will find that you have
misquoted me and substantially changed the meaning of what I actually
said. Your entire response is a straw man.
rg
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <utzt8dqdn.fsf@nhplace.com>
Ron Garret <·········@flownet.com> writes:
> In article <·············@nhplace.com>,
> Kent M Pitman <······@nhplace.com> wrote:
>
> > Ron Garret <·········@flownet.com> writes:
> >
> > > In article <························@o61g2000hsh.googlegroups.com>,
> > > "David E. Young" <··········@gmail.com> wrote:
> > >
> > > > On Jun 15, 2:27 am, Ron Garret <·········@flownet.com> wrote:
> > > > > In both cases the limiting factor was ultimately the same: we couldn't
> > > > > find any Lisp programmers to hire.
> > > >
> > > > Did you look?
> > >
> > > Of course. We even advertised in this newsgroup:
> > >
> > > http://groups.google.com/group/comp.lang.lisp/msg/f638fa5d322434c8
> > >
> > > among other places.
>
> [Much snippage]
>
> > But it just doesn't sound like you can conclude that a
> > failure to find someone is adequately summed up by "There are no
> > lisp programmers out there."
>
> If you go back and look at what I said you will find that you have
> misquoted me and substantially changed the meaning of what I actually
> said. Your entire response is a straw man.
Ron,
I can see how you would be confused, since I routinely use "..." as a
text-bracketing notation, like parentheses, as a way of grouping
compound concepts in English without having to start a new sentence.
It wasn't my intent to offer that as a quote, only to keep the fact of
embedded sentence nature of that from keeping it from being seen as a
noun. While I intended to characterize what I heard you saying, I
didn't intend it as a quote, and I'm sorry if it seemed that way. My
fault. If you want to summarize what you were saying, I'm happy to
accept that as your actual feeling--I have no desire to
mischaracterize you. I was responding to my impression of the
conversation and what was being offered, just as you have been doing.
The complete text of what I meant to actually quote, and what I was
commenting on, is at the top of my message, and is not misquoted as
far as I can tell. If it was taken out of context, the context-setup
was done by you. I have not read this entire thread, nor do I have
time to. I read isolated messages and respond to them, as everyone
probably does sometimes. Here it is again for easy reference:
> > On Jun 15, 2:27 am, Ron Garret <·········@flownet.com> wrote:
> > > In both cases the limiting factor was ultimately the same: we couldn't
> > > find any Lisp programmers to hire.
> >
> > Did you look?
>
> Of course. We even advertised in this newsgroup:
>
> http://groups.google.com/group/comp.lang.lisp/msg/f638fa5d322434c8
>
> among other places.
While you may not have intended the sense I got from it, it certainly
seems as if you are supporting the innermost quote with the outermost
quote, and I was replying to that. I did take "we couldn't find any
Lisp programmers to hire" in the context of the few messages I'd seen
as close enough to "there weren't any out there", that I didn't think
you were saying "oh, they're there". I'm not saying that's a
necessarily perfect characterization, but most reasonable people could
see how I concluded it, I suspect...
Anyway, as the Supreme Court has suggested, the best answer to bad
speech is more speech. So, speak away. Don't just say what I did
wrong--tell me what you actually meant so we can be back in sync.
Apologies again, though, for any confusion. The floor is yours to
clarify your position.
In article <·············@nhplace.com>,
Kent M Pitman <······@nhplace.com> wrote:
> Don't just say what I did
> wrong--tell me what you actually meant so we can be back in sync.
I meant to say exactly what I said: we looked, and we couldn't find any.
(By way contrast, hiring Python programmers has been relatively easy.)
I'm sorry, but I just don't know how I can be any clearer than that.
rg
>>>>> "RG" == Ron Garret <·········@flownet.com> writes:
RG> I meant to say exactly what I said: we looked, and we couldn't
RG> find any. (By way contrast, hiring Python programmers has
RG> been relatively easy.) I'm sorry, but I just don't know how I
RG> can be any clearer than that.
The important question is *why* you couldn't find any.
I know of a company that's offering $110,000 per year for a senior
Perl architect. I applied, and interviewed; the group is in total
chaos, it's all run on Windows, and the job is basically taking a
bunch of COBOL programmers and getting them to reimplement a content
management system produced by another company that was acquired. Oh,
and none of the management has been in place for longer than a month,
and they're not sure they like Perl but it's what the content
management system the other company wrote was using. In other words:
a nightmare job, in terms of management, personnel, politics, and
technical matters. I interviewed, and I could not get out of there
fast enough.
I know they've interviewed at least a dozen candidates besides me, and
I know they're still looking. I bet their line is "we looked, and we
couldn't find any," which is, to a first approximation, true; but it's
not because there aren't any qualified Perl programmers out there.
Charlton
--
Charlton Wilbur
·······@chromatico.net
In article <··············@mithril.chromatico.net>,
Charlton Wilbur <·······@chromatico.net> wrote:
> >>>>> "RG" == Ron Garret <·········@flownet.com> writes:
>
> RG> I meant to say exactly what I said: we looked, and we couldn't
> RG> find any. (By way contrast, hiring Python programmers has
> RG> been relatively easy.) I'm sorry, but I just don't know how I
> RG> can be any clearer than that.
>
> The important question is *why* you couldn't find any.
>
> I know of a company that's offering $110,000 per year for a senior
> Perl architect. I applied, and interviewed; the group is in total
> chaos, it's all run on Windows, and the job is basically taking a
> bunch of COBOL programmers and getting them to reimplement a content
> management system produced by another company that was acquired. Oh,
> and none of the management has been in place for longer than a month,
> and they're not sure they like Perl but it's what the content
> management system the other company wrote was using. In other words:
> a nightmare job, in terms of management, personnel, politics, and
> technical matters. I interviewed, and I could not get out of there
> fast enough.
>
> I know they've interviewed at least a dozen candidates besides me, and
> I know they're still looking. I bet their line is "we looked, and we
> couldn't find any," which is, to a first approximation, true; but it's
> not because there aren't any qualified Perl programmers out there.
>
> Charlton
Here's some additional data we've had no trouble hiring Python
programmers and C++ programmers. (We've even picked up an OCAML fan.)
We get about 100 resumes for every interview we do. Our offer
acceptance rate is currently 100%.
rg
From: Edi Weitz
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <ur6obu0ap.fsf@agharta.de>
On Sat, 16 Jun 2007 12:56:56 -0700, Ron Garret <·········@flownet.com> wrote:
> Here's some additional data we've had no trouble hiring Python
> programmers and C++ programmers.
The Python and C++ programmers likely didn't know you from c.l.l.
--
Lisp is not dead, it just smells funny.
Real email: (replace (subseq ·········@agharta.de" 5) "edi")
In article <·············@agharta.de>, Edi Weitz <········@agharta.de>
wrote:
> On Sat, 16 Jun 2007 12:56:56 -0700, Ron Garret <·········@flownet.com> wrote:
>
> > Here's some additional data we've had no trouble hiring Python
> > programmers and C++ programmers.
>
> The Python and C++ programmers likely didn't know you from c.l.l.
All but one of our tech hires has known me personally for many years.
rg
On Sat, 16 Jun 2007 12:56:56 -0700, Ron Garret <·········@flownet.com> said:
| ...
| We get about 100 resumes for every interview we do.
On Sat, 16 Jun 2007 14:02:00 -0700, Ron Garret <·········@flownet.com> said:
| ...
| All but one of our tech hires has known me personally for many years.
Hmmm... So, with a ratio of at least about a hundred candidates per
position, and with the exception of one case, you ended up hiring
only people that know you personally, and therefore, presumably,
people you know personally as well? Or am I misunderstanding?
(If I am not, then personal acquaintance is apparently a much better
predictor of hiring in this case than knowledge of a particular
programming language; I am not saying that there is anything wrong
with that, only that it is a rather different kind of criterion.)
---Vassil.
--
The truly good code is the obviously correct code.
In article <··············@localhost.localdomain>,
Vassil Nikolov <···············@pobox.com> wrote:
> On Sat, 16 Jun 2007 12:56:56 -0700, Ron Garret <·········@flownet.com> said:
> | ...
> | We get about 100 resumes for every interview we do.
>
> On Sat, 16 Jun 2007 14:02:00 -0700, Ron Garret <·········@flownet.com> said:
> | ...
> | All but one of our tech hires has known me personally for many years.
>
> Hmmm... So, with a ratio of at least about a hundred candidates per
> position, and with the exception of one case, you ended up hiring
> only people that know you personally, and therefore, presumably,
> people you know personally as well? Or am I misunderstanding?
I would say you are over-extrapolating. At one extreme is a person who
worked for me on my last startup. At the other extreme is a person who
never met me before they interviewed. In between is a broad range of
people, some of whom have worked with me in the past and some of whom
have not but knew me nonetheless. At least two people knew me almost
entirely by reputation; we had never worked together directly and did
not interact much socially. I don't know whether they came aboard
because of me or in spite of me, but there is no basis for believing
that they came aboard because they didn't know what they were getting
into.
rg
Vassil Nikolov <···············@pobox.com> wrote:
+---------------
| ...then personal acquaintance is apparently a much better
| predictor of hiring in this case than knowledge of a particular
| programming language...
+---------------
Actually, if you look at the better job-finding resources
[e.g., the excellent "What Color Is Your Parachute?"],
you will discover that personal acquaintance [even through
a couple of levels of indirection] is the *BEST* predictor
of hiring, bar none!
-Rob
-----
Rob Warnock <····@rpw3.org>
627 26th Avenue <URL:http://rpw3.org/>
San Mateo, CA 94403 (650)572-2607
On Sat, 16 Jun 2007 23:23:03 -0500, ····@rpw3.org (Rob Warnock) said:
| Vassil Nikolov <···············@pobox.com> wrote:
| +---------------
| | ...then personal acquaintance is apparently a much better
| | predictor of hiring in this case than knowledge of a particular
| | programming language...
| +---------------
| Actually, if you look at the better job-finding resources
| [e.g., the excellent "What Color Is Your Parachute?"],
| you will discover that personal acquaintance [even through
| a couple of levels of indirection] is the *BEST* predictor
| of hiring, bar none!
Indeed it is, as you are quite right to point out, and in the
general case that is a long established fact, of course; in my post,
the emphasis was supposed to fall on "much" and "this case", which I
should probably have underlined.
(Besides, at least from a statistical-theoretical point of view, I
should also have asked about the degree of personal acquaintance
among the candidates not hired, but then all that may not be so
important right now.)
---Vassil.
--
The truly good code is the obviously correct code.
Edi Weitz <········@agharta.de> writes:
> On Sat, 16 Jun 2007 12:56:56 -0700, Ron Garret <·········@flownet.com> wrote:
>
> > Here's some additional data we've had no trouble hiring Python
> > programmers and C++ programmers.
>
> The Python and C++ programmers likely didn't know you from c.l.l.
Precisely so. When I saw his job post in cll, my first thought was: which
person here with a clear mind will want to work under that guy?
Nicolas
On 16 Jun., 22:46, Nicolas Neuss <········@mathematik.uni-
karlsruhe.de> wrote:
> Edi Weitz <········@agharta.de> writes:
> > On Sat, 16 Jun 2007 12:56:56 -0700, Ron Garret <·········@flownet.com> wrote:
>
> > > Here's some additional data we've had no trouble hiring Python
> > > programmers and C++ programmers.
>
> > The Python and C++ programmers likely didn't know you from c.l.l.
>
> Precisely so. When I saw his job post in cll, my first thought was: which
> person here with a clear mind will want to work under that guy?
>
> Nicolas
Maybe somebody who has an open mind and finds out that actually Ron
has
done a lot with and for Lisp and that he has lots of experience? I
wouldn't
hire People who judge others only based on Usenet discussions.
·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> writes:
> Maybe somebody who has an open mind and finds out that actually Ron has
> done a lot with and for Lisp and that he has lots of experience? I
> wouldn't hire People who judge others only based on Usenet discussions.
Neither would I. On the other hand, I wouldn't hire people who read Usenet
discussions, but do not learn from them.
Nicolas
From: Ken Tilton
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <A0%ci.22$sV1.6@newsfe12.lga>
Ron Garret wrote:
> I meant to say exactly what I said: we looked, and we couldn't find any.
> (By way contrast, hiring Python programmers has been relatively easy.)
Yeah, but now you are not using Lisp. Thinking that is optional for
serious enterprise application development shows y'all do not know what
you are doing.
If Norvig is right, those Pythonistas would do fine in Lisp right out of
the box, esp. with a couple of mentors on the team. As kent noted, the
biggest issues are not language fluency. An ad saying "Looking for great
Lisp programmers, or Python programmers interested in learning Lisp"
would have produced some terrific developers. Well, maybe, the ad had
other issues. Anyway...
Of course one has to believe in Lisp, which your company does not.
Please return to paragraph one.
kt
On 16 Jun 2007 08:40:20 -0400, Kent M Pitman <······@nhplace.com> said:
| ...
| I routinely use "..." as a
| text-bracketing notation, like parentheses, as a way of grouping
| compound concepts in English without having to start a new sentence.
By the way... The very sentence above illustrates the fact that
quotes are used in several rather distinct cases, and are ambiguous
[1]. The first time I read that sentence, I took it to mean that
its author routinely used an _ellipsis_ for the described purpose,
which, of course, made me stop. Only after re-reading it and
re-reading previous messages in the thread, I realized how I had
misinterpreted it. Obviously, natural intelligence obviates a
pedantic requirement to use nested quotes or something along those
lines in such a case.
[1] This must be redundant, but just in case, here are just three
different cases of using quotes (only one of them being the
literal inclusion of a particular person's utterance):
Caesar wrote to the Senate, "I came, I saw, I conquered."
Caesar's message was, essentially, "I achieved a quick victory."
Caesar ruled Rome, and "Caesar" became one of the names of
rulers after him.
---Vassil.
--
The truly good code is the obviously correct code.
Kent M Pitman <······@nhplace.com> writes:
>
> Any company focused on "language" as "Best Way" rather than
> "competent to do" is risking failing, IMO. Although I think it
> would be fun to be working on CL, I think the thing you want to
> know is that (a) the technical leaders have a plan, (b) the
> technical leaders are making good decisions on the basis of sound
> information, and (c) the technical leaders are willing to make
> hard calls.
>
I remember the ad. I even posted a response asking who the CTO was.
No response. Dead give away. Nor did their web site have the
pertinent info. Fly by night company in LA-LA-Land.
Wade
On Jun 15, 10:27 pm, Ron Garret <·········@flownet.com> wrote:
> Of course. We even advertised in this newsgroup:
>
> http://groups.google.com/group/comp.lang.lisp/msg/f638fa5d322434c8
>
> among other places.
>
> rg
I read this ad and was going to add my thoughts, but decided against
it. Kent and others have done a superb job...
-- david
From: Kent M Pitman
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <u1wgcrb40.fsf@nhplace.com>
"David E. Young" <··········@gmail.com> writes:
> On Jun 15, 2:27 am, Ron Garret <·········@flownet.com> wrote:
> > In both cases the limiting factor was ultimately the same: we couldn't
> > find any Lisp programmers to hire.
>
> Did you look? I've been writing commercial Lisp software, using
> Lispworks, for six years now. If you advertised your position, and I
> saw it, I guess I was either uninterested or unwilling to take the
> risk. And, assuming you advertised, I'd be willing to bet other Lisp
> developers had similar impressions.
I agree, David.
Both from personal experience (looking for jobs now and then myself
over the years, and overlapping into the time period cited) and mail
I've received from others inquiring about whether I had jobs available
at various times over the years, I've generally seen instances of
Common Lisp programmers not being able to find jobs, not vice versa.
What I _do_ see is people advertising consulting gigs in strange
places for brief stints offering not enough money to make it
worthwhile and then being sad no one is a taker (for the stated price
and terms). I don't keep track of that stuff, and I don't doubt
that's been available. I don't mean to say there's been nothing, but
I do mean to say that my impression is that the lack of availability
of Lisp programmers is probably more of a self-fulfilling prophecy
because people advertise jobs that don't request Lisp, and people who
know it don't try to feature it.
Incidentally, I find the entire industry trend of advertising jobs by
saying as the high-order bit "what language you'll use" to be
lamentable. Programming used to be about solving problems, and the
key quality a programmer needed was an open mind, a willing to hear a
hard problem and devise a solution, etc. What tools they would use
was nearly inconsequential. But now, if you can solve something but
not in the right language, that's like saying you can't solve it. Or
so it seems to me so often in the job listings. If the job listings
even said "in a language that can talk to x", that would be good. The
fact that language choice is not a completely personal choice is
silly... Leaving aside Lisp for a moment, there are still a lot of
places where even using Visual Studio and the CLR, where things can
easily and demonstrably interoperate in supposedly industry standard
ways, there's a resistance to having programmers split across
languages. So it's not like that problem is unique to us. It's just
more pronounced in our case.
I _have_ certainly heard people on multiple occasions at multiple
workplaces over the years have someone suggest Lisp use inside a
company and have it shot down by a manager on the spot, with no
further investigation or discussion, citing non-availability of people
to maintain it as if it were an indisputable and widely known lemma
... just as happens with so many other Lisp myths. I always took it
as a coded response similar to a manager saying someone doesn't
have enough experience when really they mean "I don't like them and I
don't know why, so I'll cite the first concrete believable thing that
coms to mind." People often do things in businsses just because they
think it's what they want to do and then apply, retroactively,
rationales that sound good but aren't the real reason. (This is among
the reasons I don't buy the "too many parens" business. I think
people who dislike Lisp do so for reasons much more fundamental and
even if we changed that, they'd just cite some other reason... we'd
just have uselessly changed the language for nothing. That's mostly
what happened with Lisp.)
I have generally always ssumed "there aren't enough lisp programmers"
is secret code for one or more of
(a) "the lisp programmers I'd get would cost too much
because I want to underpay programmers and I can only underpay
programmers who aren't available at commodity rates"
that is, even if you doubled the lisp programmer supply, the
number of java programmers is still so much larger that it would
still mean lisp programmers could command a premium]
(b) "it would displace me"
because the person speaking is not familiar with lisp and doesn't
want to hire people he can't himself manage,
(c) "I haven't checked into it at all but someone said this once and
I tend to believe things I hear"
(except claims to the contrary of positions I've already taken)
(d) "I don't have a good reason, just a strong bias, but I need to cite
an unchallengeable claim that sounds vaguely plausible rather than
reveal my personal arbitrary bias"
(e) "I advertised a Lisp job once and no one signed up for the terms
I offered, so I'm interpreting that as if there are no people out
there at all"
(f) "it would upset the rest of my team, who can't program in lisp.
and i'm not willing to lose them, no matter how good it is."
sometimes i think this claim is made without actually asking the
rest of the team.
(g) "i once had a programmer leave me with code that was hard to maintain,
and I imagine that was the fault of the language the person was
programming in, rather than my fault for not making them document their
work clearly. i also maintain the fantasy that if someone writes obscure
code in C++ or PERL, it will be easy to hire someone to figure it out.
I just don't think the same is true of Lisp, though I can't say why.
probably the parentheses."
I'd probably guess (a) is the most common, but I have no stats to
back that up.
Strange data point: In at least one case, I knew an employer who would
actively search job listings for programmers who cited Lisp on their
resume because he'd decided it was often an earmark of someone who
"thought outside the box" (or some such). Then he would hire them to
do non-Lisp work ... though periodically being perplexed by why they
were always suggesting doing things in Lisp. ;)
This doesn't mean I don't think there's an obstacle to Lisp entry.
What it means is that it's taken us decades to get people to even
admit the possibility that we have a compiler in Lisp, even though
nearly every serious Lisp forever has had one. Prejudices die hard.
And what I'm saying is that seeing someone cite a reason is not proof
that the reason is true, that addressing the reason will help, etc.
We just need to be working on the right things.
In article <·············@nhplace.com>,
Kent M Pitman <······@nhplace.com> wrote:
> Incidentally, I find the entire industry trend of advertising jobs by
> saying as the high-order bit "what language you'll use" to be
> lamentable. Programming used to be about solving problems, and the
> key quality a programmer needed was an open mind, a willing to hear a
> hard problem and devise a solution, etc. What tools they would use
> was nearly inconsequential.
That was in the good old days, before programming became a team sport.
rg
On Jun 15, 7:29 pm, Ron Garret <·········@flownet.com> wrote:
> In article <·············@nhplace.com>,
> Kent M Pitman <······@nhplace.com> wrote:
>
> > Incidentally, I find the entire industry trend of advertising jobs by
> > saying as the high-order bit "what language you'll use" to be
> > lamentable. Programming used to be about solving problems, and the
> > key quality a programmer needed was an open mind, a willing to hear a
> > hard problem and devise a solution, etc. What tools they would use
> > was nearly inconsequential.
>
> That was in the good old days, before programming became a team sport.
And when there were only three programming languages: Lisp, Fortran,
and Assembler, and it was obvious from the problem description which
one was intended.
On Fri, 15 Jun 2007 19:32:48 -0700, ·········@gmail.com" <········@gmail.com> said:
| On Jun 15, 7:29 pm, Ron Garret <·········@flownet.com> wrote:
|| In article <·············@nhplace.com>,
|| Kent M Pitman <······@nhplace.com> wrote:
||
|| > Incidentally, I find the entire industry trend of advertising jobs by
|| > saying as the high-order bit "what language you'll use" to be
|| > lamentable. Programming used to be about solving problems, and the
|| > key quality a programmer needed was an open mind, a willing to hear a
|| > hard problem and devise a solution, etc. What tools they would use
|| > was nearly inconsequential.
||
|| That was in the good old days, before programming became a team sport.
| And when there were only three programming languages: Lisp, Fortran,
| and Assembler, and it was obvious from the problem description which
| one was intended.
There were also COBOL and PL/1, at least (I suppose it would be
debatable whether we should also count, say (in alphabetical order),
Algol, JCL, RPG, etc., so I won't insist on these; but there are
probably other languages I have missed).
Besides, didn't programming become a team sport as early as the
early 50s?
---Vassil.
--
The truly good code is the obviously correct code.
On 15 Jun 2007 20:34:07 -0400, Kent M Pitman <······@nhplace.com> said:
| ...
| Incidentally, I find the entire industry trend of advertising jobs by
| saying as the high-order bit "what language you'll use" to be
| lamentable. Programming used to be about solving problems, and the
| key quality a programmer needed was an open mind, a willing to hear a
| hard problem and devise a solution, etc. What tools they would use
| was nearly inconsequential. But now, if you can solve something but
| not in the right language, that's like saying you can't solve it.
There will be further changes to this state of affairs, though, and
in a sense things may return to the problem-solving "paradise", but
on a different level, as it were. In my opinion (I am not the first
one to have arrived at this, of course), within twenty years
(possibly less, and probably not much more) the number of people
writing programs in today's sense will decrease significantly
(especially if hobbyists are not counted). What will be produced by
human hand, whether in symbolic or graphical form, will be
specifications and designs, the "what" of software (and to some
extent algorithms in abstract form), while programs, the "how" of
software, will be (largely) automatically generated.
One argument in favor of the above is that continuing to write more
and more (procedural, imperative) programs by hand is unsustainable:
this is not an activity for which humans are well-suited, and it
does not seem feasible to double the number of programmers in the
world every decade, say. Another argument is that designs and
specifications are somewhat easier to debug (if produced in
declarative form), especially if they can also be verified by
running them "on the fly", which implies automatic (i.e., fast)
software generation.
Thus, two decades or so from now, arguing about the right
_programming_ language will be irrelevant. There will be languages
to argue about (at least until everything converges onto some subset
of natural language), but they will be different kinds of languages.
(Again, this is opinion; and as any prediction, it should be taken
with an appropriately sized grain of salt.)
Incidentally, I think that it is a pity that the formative years
(centuries ago) of mathematical notation have been forgotten (or
largely forgotten). The language of mathematics today is highly
mature and stable; I am sure that we could relate strongly to an
earlier age when it was growing up.
---Vassil.
--
The truly good code is the obviously correct code.
From: Friedrich Dominicus
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <87d4zrdtys.fsf@flarge.here>
Vassil Nikolov <···············@pobox.com> writes:
> Thus, two decades or so from now, arguing about the right
> _programming_ language will be irrelevant. There will be languages
> to argue about (at least until everything converges onto some subset
> of natural language), but they will be different kinds of
> languages.
I bet against it. The discussion will be why "not Lisp" instead of Bla
III and JAVA NT, C#### (CodeName: RunAWay) ;-)
Regards
Friedrich
--
Please remove just-for-news- to reply via e-mail.
From: Matthias Buelow
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <5eg8rtF388bguU1@mid.dfncis.de>
Kent M Pitman wrote:
> I have generally always ssumed "there aren't enough lisp programmers"
> is secret code for one or more of
[...]
I think these points could be summarized in the observable fact that
people simply don't follow someone or something they perceive to be
weak or losers. Most people follow the herd and make instinctive
decisions. Lisp (and especially Common Lisp) is perceived to be a
loser, no matter if that perception is based on facts or not. You
cannot argue against that perception. If you take Lisp's side and try
to talk your boss or client into adopting the language, all you effect
is that you aswell will be perceived as a loser and no matter what you
say and how well supported it is, it will fall on deaf ears. The first
thing that needs to be done is to rid the language of that loser image
and turn it into a winner image, however I don't think this can be
done with a sulky "holier-than-thou" attitude. The second thing is not
to suck up to the herd animals and try to convince them of anything.
It won't work. You need to be in a stronger position than they are and
they will fall in line automatically.
On Jun 28, 1:06 am, Matthias Buelow <····@incubus.de> wrote:
> Kent M Pitman wrote:
> > I have generally always ssumed "there aren't enough lisp programmers"
> > is secret code for one or more of
>
> [...]
>
> I think these points could be summarized in the observable fact that
> people simply don't follow someone or something they perceive to be
> weak or losers. Most people follow the herd and make instinctive
> decisions. Lisp (and especially Common Lisp) is perceived to be a
> loser, no matter if that perception is based on facts or not. You
> cannot argue against that perception. If you take Lisp's side and try
> to talk your boss or client into adopting the language, all you effect
> is that you aswell will be perceived as a loser and no matter what you
> say and how well supported it is, it will fall on deaf ears. The first
> thing that needs to be done is to rid the language of that loser image
> and turn it into a winner image, however I don't think this can be
> done with a sulky "holier-than-thou" attitude. The second thing is not
> to suck up to the herd animals and try to convince them of anything.
> It won't work. You need to be in a stronger position than they are and
> they will fall in line automatically.
I suport this 100% . We need more applications written in lisp. More
success stories.
Slobodan Blazeski
Slobodan Blazeski wrote:
> I suport this 100% . We need more applications written in lisp. More
> success stories.
I doubt it. What would be far more use is a platform of some kind for
which Lisp is the natural scripting language - in much the same way as
you use Perl to drive Unix, Ruby to drive Rails, PHP to drive the Web
server (the Unix web server, anyway: substitute ASP for MS shops, but
the point is that it's still only one primary language per platform),
and Javascript for the client.
Emacs comes close, but it's not the ideal platform for non-programmers
as non-programmers generally have no interest in running emacs anyway.
What was it JShrager was talking about, "semantic web verticals"?
-dan
From: Don Geddis
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <87y7i32yag.fsf@geddis.org>
Daniel Barlow <···@coruskate.net> wrote on Fri, 29 Jun 2007:
> What would be far more use is a platform of some kind for which Lisp is the
> natural scripting language
> What was it JShrager was talking about, "semantic web verticals"?
Yes, but despite numerous requests he refused to provide any more
information. Just his extreme confidence that they would be greatly
profitable.
-- Don
_______________________________________________________________________________
Don Geddis http://don.geddis.org/ ···@geddis.org
Pig: That boat's about to go over the edge of the world.
Goat: The world's round, Pig.
Pig: So is my kitchen table, but you can still knock off the cereal bowl.
-- Pearls Before Swine, 3/1/2007
From: Matthias Buelow
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <5ej1j4F387jhoU1@mid.dfncis.de>
Daniel Barlow wrote:
> What was it JShrager was talking about, "semantic web verticals"?
You think we should write a really powerful buzzword generator in Lisp?
Something is definately wrong with google groups or safari for windows
it eat my post once again.
> On Jun 29, 1:49 am, Daniel Barlow <····@coruskate.net> wrote:
>
>
>
> > Slobodan Blazeski wrote:
> > > I suport this 100% . We need more applications written in lisp. More
> > > success stories.
>
> > I doubt it. What would be far more use is a platform of some kind for
> > which Lisp is the natural scripting language - in much the same way as
> > you use Perl to drive Unix, Ruby to drive Rails, PHP to drive the Web
> > server (the Unix web server, anyway: substitute ASP for MS shops, but
> > the point is that it's still only one primary language per platform),
> > and Javascript for the client.
>
> > Emacs comes close, but it's not the ideal platform for non-programmers
> > as non-programmers generally have no interest in running emacs anyway.
>
> > What was it JShrager was talking about, "semantic web verticals"?
>
> > -dan
>
Why not hit Java in their own court ? Most software in a certain
domain is written in a language that the other software is written
in? Java is a modern day cobol.If Lisp claimes productivity factor of
3-10x let's show the world what lisp could do.
So why do other shops use Java:
1. Portable , you could run it in windows, linux , freebsd ...
2. Web oriented - whatever that means, but there's only 3 serioues
platforms for web asp,java,php & now ruby
3. Plenty of libraries
4. Plenty of material like books, magazines, tutorials etc to learn
starters how to create business apps with java
5. Plenty of apps to show - it's hard to convince people to use a
technology that nobody else use.
Lisp has 1, 2, and probably good part of 3 but lacks 4 & 5 . If we
decide to make a web business framework we already have hunchentoot,
clsql, cl-mathstat ... . Also a lot of the data mining algorithms are
already implemented in lisp or could be quickly implemented. So we
only need to show the people how to make business apps with lisp,
make some tutorials, create code that other people could use and
implement few libraries that are missing. I alredy started with a
tutorial for a CRM (Customer relationship management ) web
application with lisp, currently down due to formatting issues. Also
I could Implement a financial mathematics libarray and open source my
code for accounting receivables, if there is insteresthing from the
lisp community in going in such direction.
Else I would be very interested what kind of platform are you
proposing ?
Slobodan Blazeski
Slobodan Blazeski wrote:
> Why not hit Java in their own court ?
Because it's been done already. By Java. And, frankly, because it's
boring - Java isn't where the buzz is any longer, it's where the
suit-wearing 9-5 programmers are, and you're not going to get them
enthused about a new language now they've invested all that time and
money to get certified in a perfectly adequate one.
A new language needs a new platform. That may end up involving a little
spin and redefinition of an existing platform (Rails apps are, after
all, still delivered via the web, but all the MVC framework stuff is
enough to set it apart from PHP anyway) but you have to start from the
mindset of identifying new opportunities, not just of doing the same old
same old but faster/better. Covers bands don't fill stadiums.
-dan
On Jun 29, 2:08 pm, Daniel Barlow <····@coruskate.net> wrote:
> Slobodan Blazeski wrote:
> > Why not hit Java in their own court ?
>
> Because it's been done already. By Java. And, frankly, because it's
> boring - Java isn't where the buzz is any longer, it's where the
> suit-wearing 9-5 programmers are, and you're not going to get them
> enthused about a new language now they've invested all that time and
> money to get certified in a perfectly adequate one.
>
Screw suits. Even if the domain is boring we could do it in lisp, and
I don't find data mining boring, though your taste may vary.
> A new language needs a new platform. That may end up involving a little
> spin and redefinition of an existing platform (Rails apps are, after
> all, still delivered via the web, but all the MVC framework stuff is
> enough to set it apart from PHP anyway) but you have to start from the
> mindset of identifying new opportunities, not just of doing the same old
> same old but faster/better. Covers bands don't fill stadiums.
>
> -dan
You're only criticizing. Propose a platform where lisp would
fluorish ? I'm really eager to hear it.
Slobodan Blazeski
Slobodan Blazeski wrote:
> You're only criticizing. Propose a platform where lisp would
> fluorish ? I'm really eager to hear it.
Me too!
--
Dr Jon D Harrop, Flying Frog Consultancy
The OCaml Journal
http://www.ffconsultancy.com/products/ocaml_journal/?usenet
Slobodan Blazeski wrote:
> You're only criticizing. Propose a platform where lisp would
> fluorish ? I'm really eager to hear it.
Seriously, if I knew of one I'd be doing it already. But your proposal
has all the hallmarks of
1) something must be done
2) beating java at its own game is something
3) therefore we must do it
Logicians undoubtedly have some fancy name for the fallacy embodied
here, I only know enough to recognise it as such.
(And don't imagine you're the first person to have this idea: a bit of
rooting around in the free Lisp stuff of the past ten years (approx)
will reveal at least one prior attempt)
-dan
From: Ariel Badichi
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <87bqeyivez.fsf@sneeze.site>
Daniel Barlow <···@coruskate.net> writes:
> Slobodan Blazeski wrote:
>> You're only criticizing. Propose a platform where lisp would
>> fluorish ? I'm really eager to hear it.
>
> Seriously, if I knew of one I'd be doing it already. But your
> proposal has all the hallmarks of
>
> 1) something must be done
> 2) beating java at its own game is something
> 3) therefore we must do it
>
> Logicians undoubtedly have some fancy name for the fallacy embodied
> here, I only know enough to recognise it as such.
>
Undistributed middle term.
> (And don't imagine you're the first person to have this idea: a bit of
> rooting around in the free Lisp stuff of the past ten years (approx)
> will reveal at least one prior attempt)
>
>
> -dan
From: Matthias Buelow
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <5ekvulF38nn36U1@mid.dfncis.de>
Daniel Barlow wrote:
> 1) something must be done
> 2) beating java at its own game is something
> 3) therefore we must do it
Beating java at its own game is nonsense with Lisp; Java programming is
bureaucratic management, both in the language and in the programming
force. Lisp's dynamic, extensible and flexible nature would actually be
a liability rather than an asset here. Hypothetically, Java could only
be beaten by a language than implements the various bureaucratic models
(patterns, uml-modelling, use-cases, whathaveyou) of "modern software
engineering" in a significantly more consistent and thus convenient
package while at the same time making the language even simpler (to
reduce the cost of training the grunts). I don't think Lisp should and
can compete with that, at least not as its main raison d'etre. If you're
aiming for that, take something like Modula, mess it up a bit and give
it a C-like syntax that everyone feels familiar with. Oh, Gosling has
done that already...
The business-programming people can only be won over if they have no
alternative; remember what I wrote about herd animals. They will not
take up Lisp because it's better than their current tools (Java, etc.)
but when everyone else is doing it, so that eventually using a dynamic,
complicated language (probably in a castrated way, with lots of
"patterns", software-engineering "best practices and models", the whole
red tape etc.) is less costly for them than sticking with what they have
been using so far. Which would necessitate a rather thorough "market"
penetration (and here I think mostly of the open-source arena) before
those people would even considering to take the alternative seriously.
Such a thing has been done before with Gnu/Linux, which has more or less
cracked the commercial Unix market but we can all see how difficult it
is to establish it in business sectors where it is competing against
established tools like MS Windows (although I think in the long run, it
will succeed).
>>>>> "MB" == Matthias Buelow <···@incubus.de> writes:
MB> The business-programming people can only be won over if they
MB> have no alternative; remember what I wrote about herd
MB> animals. They will not take up Lisp because it's better than
MB> their current tools (Java, etc.) but when everyone else is
MB> doing it, so that eventually using a dynamic, complicated
MB> language (probably in a castrated way, with lots of
MB> "patterns", software-engineering "best practices and models",
MB> the whole red tape etc.) is less costly for them than sticking
MB> with what they have been using so far.
And beyond that, there's a whole industry that has sprung up around
papering over the deficiencies in C++ and Java. The gurus have
invented things like "patterns," for instance, and made considerable
money from it; the sheep have adopted the doctrine of "patterns,"
which reduces the pain of Java or C++. To get Lisp to have a foothold
there, you have to get the gurus to admit publicly that "patterns" are
simply a papering-over of Java and C++'s flaws *and* you have to get
the sheep to reject the doctrine of patterns when they'd rather just
follow along with the flock. Neither is in the best interests of the
decisionmakers; the gurus make too much money, and the leaders of the
sheep don't mind the pain because they're not the ones suffering it.
Charlton
--
Charlton Wilbur
·······@chromatico.net
Your replies (Wilbur, Buelow, Badichl, Barlow) are mixture of
misunderstanding and not offering any alternatives .
1st I don't give a damn what suits and bureacratic companies use.
Actually I would prefer them to stay with Java or any other language
dumb as that, or even dumber, though that would be hard to imagine.
2nd I don't what you guys do for living, maybe your constructing new
software for the mission to Mars or flight simulators or something as
cool as that but I do business systems as ERP, CRM and some custome
made simple apps. I use Delphi at work and Lisp for my home projects
and I would be far happier and much better lisper if I have a chance
to use Lisp fulltime. So for mere mortals like me using lisp and
getting paid for it is much better than using C#,Java ...
Slobodan Blazeski
My apologies to those who like Java, C#, PHP, Delphi, Visual Basic,
Perl, Python, Ruby, COBOL,F#, or any other language. I know you think
you know a better language than lisp. All I can say is I do, too!
On Jul 1, 4:56 am, Charlton Wilbur <·······@chromatico.net> wrote:
> >>>>> "MB" == Matthias Buelow <····@incubus.de> writes:
>
> MB> The business-programming people can only be won over if they
> MB> have no alternative; remember what I wrote about herd
> MB> animals. They will not take up Lisp because it's better than
> MB> their current tools (Java, etc.) but when everyone else is
> MB> doing it, so that eventually using a dynamic, complicated
> MB> language (probably in a castrated way, with lots of
> MB> "patterns", software-engineering "best practices and models",
> MB> the whole red tape etc.) is less costly for them than sticking
> MB> with what they have been using so far.
>
> And beyond that, there's a whole industry that has sprung up around
> papering over the deficiencies in C++ and Java. The gurus have
> invented things like "patterns," for instance, and made considerable
> money from it; the sheep have adopted the doctrine of "patterns,"
> which reduces the pain of Java or C++.
I'm not completely fallowing about patterns as we have quite flavour
of them in lisp remeber
Norvigs http://norvig.com/design-patt and of Richard Gabriels
http://www.dreamsongs.com/Files/PatternsOfSoftware.pdf and his
opession with Alexander.
> To get Lisp to have a foothold
> there, you have to get the gurus to admit publicly that "patterns" are
> simply a papering-over of Java and C++'s flaws *and* you have to get
> the sheep to reject the doctrine of patterns when they'd rather just
> follow along with the flock.
> Neither is in the best interests of the
> decisionmakers; the gurus make too much money, and the leaders of the
> sheep don't mind the pain because they're not the ones suffering it.
>
> Charlton
>
> --
> Charlton Wilbur
> ·······@chromatico.net
On Jun 29, 5:34 pm, Daniel Barlow <····@coruskate.net> wrote:
> Slobodan Blazeski wrote:
> > You're only criticizing. Propose a platform where lisp would
> > fluorish ? I'm really eager to hear it.
>
> Seriously, if I knew of one I'd be doing it already. But your proposal
> has all the hallmarks of
>
> 1) something must be done
> 2) beating java at its own game is something
> 3) therefore we must do it
>
> Logicians undoubtedly have some fancy name for the fallacy embodied
> here, I only know enough to recognise it as such.
>
> (And don't imagine you're the first person to have this idea: a bit of
> rooting around in the free Lisp stuff of the past ten years (approx)
> will reveal at least one prior attempt)
>
> -dan
1. We don't have an idea what that framework will do
2. Will adopt a fighting style of creating software that is created so
fast, with so few programmers & such quality that all competitors
will be blown off.
3. Profit :)
Slobodan Blazeski
>>>>> "SB" == Slobodan Blazeski <·················@gmail.com> writes:
SB> Why not hit Java in their own court ? Most software in a
SB> certain domain is written in a language that the other
SB> software is written in? Java is a modern day cobol.If Lisp
SB> claimes productivity factor of 3-10x let's show the world what
SB> lisp could do. So why do other shops use Java?
You miss the single major reason: it's hard for morons to screw up
Java.
Lisp gets major productivity boosts, when it does, because it's a
powerful tool in the hands of a good programmer. But I think there
are two major factors here: there's a selection bias in favor of good
programmers in the first place, because bad programmers are never
going to even hear of Lisp beyond some academic curiosity languaage,
let alone try a project in it; and Lisp has almost no bondage-
and-discipline inherent in the language, so it gets out of the way of
the good programmer.
But the audience for Java, or at least the audience that has made it
successful, is the army of corporate drones and their managers. They
*aren't*, by and large, good programmers; if they are, they quickly
get promoted to a management level where they aren't doing any more
programming. This audience isn't going to get a large benefit from a
more powerful tool, because they'll spend as much time fixing the
misuses of it from the less skilled members of the team as they will
benefiting from the proper uses of it. And the bondage and discipline
of Java is, in that environment, a good thing, because it leads to an
emphasis on process.
In other words, to make Lisp palatable for that audience, you'd have
to remove some of the more powerful features and the flexibility, and
you'd have to add structure to the language. Imagine Lisp with no
macros (or with a well-disguised version of them), with a very
well-defined orthodoxy of, say, object orientation approaches, and
with S-expressions replaced by some other syntax. Sounds a lot like
Python, Perl, or Ruby to me....
Charlton
--
Charlton Wilbur
·······@chromatico.net
On Jul 1, 4:46 am, Charlton Wilbur <·······@chromatico.net> wrote:
> >>>>> "SB" == Slobodan Blazeski <·················@gmail.com> writes:
>
> SB> Why not hit Java in their own court ? Most software in a
> SB> certain domain is written in a language that the other
> SB> software is written in? Java is a modern day cobol.If Lisp
> SB> claimes productivity factor of 3-10x let's show the world what
> SB> lisp could do. So why do other shops use Java?
>
> You miss the single major reason: it's hard for morons to screw up
> Java.
Lot of them are very succesfull at.
>
> Lisp gets major productivity boosts, when it does, because it's a
> powerful tool in the hands of a good programmer. But I think there
> are two major factors here: there's a selection bias in favor of good
> programmers in the first place, because bad programmers are never
> going to even hear of Lisp beyond some academic curiosity languaage,
> let alone try a project in it; and Lisp has almost no bondage-
> and-discipline inherent in the language, so it gets out of the way of
> the good programmer.
>
> But the audience for Java, or at least the audience that has made it
> successful, is the army of corporate drones and their managers. They
> *aren't*, by and large, good programmers; if they are, they quickly
> get promoted to a management level where they aren't doing any more
> programming. This audience isn't going to get a large benefit from a
> more powerful tool, because they'll spend as much time fixing the
> misuses of it from the less skilled members of the team as they will
> benefiting from the proper uses of it. And the bondage and discipline
> of Java is, in that environment, a good thing, because it leads to an
> emphasis on process.
>
> In other words, to make Lisp palatable for that audience, you'd have
> to remove some of the more powerful features and the flexibility, and
> you'd have to add structure to the language. Imagine Lisp with no
> macros (or with a well-disguised version of them), with a very
> well-defined orthodoxy of, say, object orientation approaches, and
> with S-expressions replaced by some other syntax. Sounds a lot like
> Python, Perl, or Ruby to me....
>
> Charlton
>
> --
> Charlton Wilbur
> ·······@chromatico.net
Ron Garret <·········@flownet.com> writes:
> BTW...
>
> In article <··············@lion.rapttech.com.au>,
> Tim X <····@nospam.dev.null> wrote:
>
>> why is CL not being
>> considered as a viable development platform for startups?
>
> I can offer up a data point from my recent experience. For the past
> year I have been involved in two startups, both of which could have
> benefited from using Lisp, and neither of which used it. In one of
> those startups I was a founder and the CTO.
>
> In both cases the limiting factor was ultimately the same: we couldn't
> find any Lisp programmers to hire. I actually negotiated a deal with
> one of the Lisp vendors where they would provide licenses AND
> programming labor at cost, but then an investor came along and poured an
> obscene amount of money into the company, and contract-labor-for-equity
> was no longer a good deal so the whole thing fell through, and now we're
> using C++ instead. (The irony is truly staggering.)
>
> In any case, despite the fact that the plan was thwarted by having too
> much money (which is ultimately a nice problem to have I must admit) I
> still think it's a winning strategy and someone ought to try it again
> while there are still some Lisp vendors out there to try it with.
>
> rg
Hmm. That sort of reflects part of my views. The issue isn't as simple as there
not being a good open source implementation. Part of the problem is a lack of
experienced CL programmers, which I suspect is a reflection of the poor way
many of these programmers are introduced to CL. Its certainly not a paradigm
that is actively supported by many institutions and of those that I'm familiar
with, CL is presented as that langauge used by the AI folk and since the AI
winter, anything associated with AI is pretty much discounted. This is
partially why I liked Peter's Practicle Common Lisp as it showed that CL wasn't
just a language for AI research/development etc.
Tim
--
tcross (at) rapttech dot com dot au
> I think you missed Kent's point. You cannot just say this is the problem and
> here is the solution. We need to discuss/debate what the problem is until we
> find agreement on the problem definition before putting forward a solution and
> the problem needs to be presented in a way that encourages
> discussion/contribution.
>
> For example, I find no evidence in your problem definition that indicates the
> small startup even considered lisp, little only discounting it because it was
> too expensive. That seems to be somewhat of a leap. I actually suspect that few
> startups even consider lisp as a possible development platform and never get as
> far as looking at the costs of commercial implementations or even evaluate the
> open source offerings. This is why I don't beleive a high quality open source
> version would make a jot of difference. It is possible that a startup may come
> to this conclusion if they did honestly evaluate/consider lisp, but I don't
> believe that initial step is actually occuring. Therefore, I feel we need to
> start with a much more basic problem definition/question - why is CL not being
> considered as a viable development platform for startups?
>
> Tim
The bottom line is they don't like s-expressions. Implement a modern
IDE (think Eclipse, which implies a new GUI that isn't scary like
CLIM), that can toggle back and forth between something like sweet
expressions (http://www.dwheeler.com/blog/2006/06/17/) and s-
expressions, get SBCL up to speed on Windows................and don't
profit, but you'll get a few converts.
On Jun 15, 3:42 am, ·············@gmail.com" <············@gmail.com>
wrote:
> The bottom line is they don't like s-expressions. Implement a modern
> IDE (think Eclipse, which implies a new GUI that isn't scary like
> CLIM), that can toggle back and forth between something like sweet
> expressions (http://www.dwheeler.com/blog/2006/06/17/) and s-
> expressions, get SBCL up to speed on Windows................and don't
> profit, but you'll get a few converts.
I think the sweet expressions thing was already done. Guido called it
Python.
I'm being a little sarcastic here, but I think s-expressions, and thus
macros are one of the few things that still sets Lisp apart from
"dynamic" languages like Python and Ruby. Most of the other things
that set Lisp apart have been copied by Python, Ruby, even Perl to
some extent.
http://norvig.com/python-lisp.html
The other big win with Lisp is that you get all the productivity of
dynamicly typed, functional programming ready, world's most powerful
object system in a language that's still ridiculously fast. Lisp is
more productive than Python or Ruby, while runtime speed is closer to
C.
Paul Graham has gone on about this at length, but it's the parentheses
that give Lisp a lot of its superpowers. If you're going to get rid
of them, just use Python or Ruby and call it a day (except for the
performance thing).
-jimbo
On Jun 15, 8:54 am, ········@gmail.com wrote:
> On Jun 15, 3:42 am, ·············@gmail.com" <············@gmail.com>
> wrote:
>
> > The bottom line is they don't like s-expressions. Implement a modern
> > IDE (think Eclipse, which implies a new GUI that isn't scary like
> > CLIM), that can toggle back and forth between something like sweet
> > expressions (http://www.dwheeler.com/blog/2006/06/17/) and s-
> > expressions, get SBCL up to speed on Windows................and don't
> > profit, but you'll get a few converts.
>
> I think the sweet expressions thing was already done. Guido called it
> Python.
>
Great, then it's all moot anyway. Python is here now. Javascript is
also considered part of the Lisp family, and widely used. Common Lisp
is all over the place ;)
> I'm being a little sarcastic here, but I think s-expressions, and thus
> macros are one of the few things that still sets Lisp apart from
> "dynamic" languages like Python and Ruby.
> Most of the other things
> that set Lisp apart have been copied by Python, Ruby, even Perl to
> some extent.
>
> http://norvig.com/python-lisp.html
>
> The other big win with Lisp is that you get all the productivity of
> dynamicly typed, functional programming ready, world's most powerful
> object system in a language that's still ridiculously fast. Lisp is
> more productive than Python or Ruby, while runtime speed is closer to
> C.
>
And MOP/CLOS, but ostensibly you don't lose anything with sweet-
expressions and it would just be a toggle of the keyboard between s-
expressions and a "user-friendly" sweet-expressions. I can hear the
cries of "just use OpenDylan"
> Paul Graham has gone on about this at length, but it's the parentheses
> that give Lisp a lot of its superpowers. If you're going to get rid
> of them, just use Python or Ruby and call it a day (except for the
> performance thing).
>
They're not gotten rid of though. Personally, I could care less, but
the question was raised of why CL isn't being chosen, and it's not
just expensive commercial offerings and weak open source offerings.
> -jimbo
> I'm being a little sarcastic here, but I think s-expressions, and thus
> macros are one of the few things that still sets Lisp apart from
> "dynamic" languages like Python and Ruby. Most of the other things
> that set Lisp apart have been copied by Python, Ruby, even Perl to
> some extent.
I think you underestimate the significance of the difference in
performance. To a first appropximation, Python is productive because
its slow. That is to say if you avoided using the inherently slow
features of the language, it wouldn't be particularly productive at
all. So while Python may be closing the gap on overall productivity,
you really couldn't close the gap in performance without either giving
up that productivity, or turning the language into something that
wasn't Python anymore.
PS) It's interesting to note that there has been quite a bit of work
on making Python faster, consisting of a few alternative
implementations (Psyco, IronPython), and maybe half a dozen papers on
type inference. All of them have failed to either provide interesting
speedups (twice as fast as really slow is still really slow), or are
only effective on a uselessly narrow subset of the language.
>>>>> "TX" == Tim X <····@nospam.dev.null> writes:
TX> Therefore, I feel we need to start with a much more basic
TX> problem definition/question - why is CL not being considered
TX> as a viable development platform for startups?
Because if you're starting up, you need to get the venture
capitalists' attention so that you can get funding. They've all seen
Python and Perl and Ruby in business plans; they haven't seen Lisp.
Because startups are all about getting a product to market quickly.
If you choose Lisp, you get a powerful expressive language where you
can write code to do anything you want, but getting a framework where
it integrates nicely with the web costs a lot of money; if you choose
Perl or Ruby, you get a powerful expressive language where a lot of
the code to do what you want is already written, and the framework to
integrate it nicely with the web is also free.
Because startups know that they're going to have to hire programmers
eventually, and while you need to hire good programmers regardless,
it's easier to find a good programmer who's been exposed to Python
than a good programmer who's been exposed to Lisp.
Because startups know that they're going to have to hire mediocre
programmers eventually, and it's easier for a mediocre programmer to
be productive in a more restrictive language.
Because startup founders who are curious about Lisp check out
comp.lang.lisp and see all the whining about how Lisp is the one true
programming language but nobody ever realizes this and accords the
genius programmers who choose Lisp with the adulation and groupies
that they so richly deserve; the founder decides he'd rather work with
a low-maintenance C programmer.
The answer is, I am sure, somewhere in there.
Charlton
--
Charlton Wilbur
·······@chromatico.net
On Jun 15, 12:56 pm, Charlton Wilbur <·······@chromatico.net> wrote:
> Because if you're starting up, you need to get the venture
> capitalists' attention so that you can get funding. They've all seen
> Python and Perl and Ruby in business plans; they haven't seen Lisp.
There was also a time when venture capitalists did not see Python and
Ruby in business plans (if we are discussing web ventures, Perl
probably always has been in there because it was a web development
language almost from the web's beginning). I'm pretty sure both
Python and Ruby were overnight successes over a decade the making.
I think one of the critical elements that allowed Python and Ruby to
go from unknown to programming language rock star was very easy ways
of getting started and quickly becoming productive. For Common Lisp,
my argument is that the necessary pieces exist, but they have not been
put together in a "batteries included", "try before you buy in"
fashion.
And yes, I should stop talking about this and do something about it.
Luckily, I have the excuse of not being experienced enough or
knowledgeable enough to handle the task, yet.
-jimbo
>>>>> "j" == jimbokun <········@gmail.com> writes:
j> On Jun 15, 12:56 pm, Charlton Wilbur <·······@chromatico.net>
j> wrote:
>> Because if you're starting up, you need to get the venture
>> capitalists' attention so that you can get funding. They've
>> all seen Python and Perl and Ruby in business plans; they
>> haven't seen Lisp.
j> There was also a time when venture capitalists did not see
j> Python and Ruby in business plans (if we are discussing web
j> ventures, Perl probably always has been in there because it was
j> a web development language almost from the web's beginning).
j> I'm pretty sure both Python and Ruby were overnight successes
j> over a decade the making.
Oh, I'm not saying it is necessarily a *good* reason.
But honestly, there are two sorts of startups. One is started on a
shoestring, and the eventual goal is to form a business that's
profitable enough to run in the black and stable enough that the
founders don't ever need to work for other people again. This sort is
likely to use whatever the founders are comfortable with.
The other sort is the one started on venture capital money, and the
eventual goal is to pay dividends to the stockholders or to sell to
Google or Amazon for kerjillions of dollars, hopefully at least 10 to
15 times as much as the venture capitalists put in. This sort is
likely to use whatever has the glossiest brochures and whatever the
venture capitalists have heard of.
Lisp isn't set up to do well in either case: in the former, Python,
Perl, and Ruby offer 80% of what Lisp does for about 20% of the pain
of getting started, except when the founders are already familiar with
Lisp (which they aren't, unless they're Ken Tilton or Paul Graham or
Robert Morris); and in the latter, Python, Perl, and Ruby offer 80% of
what Lisp does and names that the VCs are likely to recognize.
Charlton
--
Charlton Wilbur
·······@chromatico.net
Tim X <····@nospam.dev.null> writes:
<snip>
> believe that initial step is actually occuring. Therefore, I feel we need to
> start with a much more basic problem definition/question - why is CL not being
> considered as a viable development platform for startups?
The simple answer is that few established companies use it. Also it
is not heavily used at educational institutions.
Of the three start-ups I have worked for, two have been successful.
The ones that succeeded got going because of support by other
successful entities. (as a side note you will not even get an
agreement as to what successful means). If you need a big example of
this, consider Microsoft's initial support from IBM. To be able to
attract the support of a successful backer like that, you have to look
attractive and interact with them. If you are doing software
development, you have to fit in with what their perception of software
development is and how it works (because you have to work with them
until they are comfortable enough with your abilities).
Wade
Wade Humeniuk <········@telus.net.no.spam> writes:
> Tim X <····@nospam.dev.null> writes:
>
> <snip>
>> believe that initial step is actually occuring. Therefore, I feel we need to
>> start with a much more basic problem definition/question - why is CL not being
>> considered as a viable development platform for startups?
>
> The simple answer is that few established companies use it. Also it
> is not heavily used at educational institutions.
>
> Of the three start-ups I have worked for, two have been successful.
> The ones that succeeded got going because of support by other
> successful entities. (as a side note you will not even get an
> agreement as to what successful means). If you need a big example of
> this, consider Microsoft's initial support from IBM. To be able to
> attract the support of a successful backer like that, you have to look
> attractive and interact with them. If you are doing software
> development, you have to fit in with what their perception of software
> development is and how it works (because you have to work with them
> until they are comfortable enough with your abilities).
So the best move would be for successful lisp companies to help
startups using lisp.
--
__Pascal Bourguignon__ http://www.informatimago.com/
NOTE: The most fundamental particles in this product are held
together by a "gluing" force about which little is currently known
and whose adhesive power can therefore not be permanently
guaranteed.
Pascal Bourguignon <···@informatimago.com> writes:
> So the best move would be for successful lisp companies to help
> startups using lisp.
I think that is true. A culture of success using Lisp would grow
organically. The reason a downturn in the use of Lisp may have
happened is those groups (and those leaders using it) abandoned (and
actively discouraged) its use.
Growth of computer culture is like clumps of solute forming out of a
saturated solution. Matter begins to form around a nucleus. The
bigger the clump the faster matter collects. Its like tapping a glass
of super-saturated salt water.
Wade
> > So the best move would be for successful lisp companies to help
> > startups using lisp.
>
> I think that is true. A culture of success using Lisp would grow
> organically.
Whereas I agree that this would be minorly useful, I don't think it
will be sufficent. Unless kids who are NOT involved in startups get
used to using it, it's not going to get brought into startups by the
kids starting them. And these kids aren't going to pay a dime for a
great Lisp (and won't even think about using a poor one -- i.e., one
which doesn't work on everything, install easily, have full CPAN
support, and something like RAILS is rapidly becoming a part of the
required library set!) because they don't have to pay a dime to use
great Rubies, Pytons, Javas, and C++s. (Startups may pay a dime, but
aren't going to pay $500, and certainly no more than that!) As far as
I can see, free-waring one of the great Lisps, or brigning one of the
moderately good ones up to speed, are our only hope.
(Actually, I think that our real only hope is that everyone converts
to Ruby and JSON, which are approximately Lisp and SExprs, and then
someone in the Ruby community will realize that you can represent you
code elegantly in JSON, and that that gives you Macros...and then Lisp
is reborn!)
·········@gmail.com" <········@gmail.com> writes:
> Unless kids who are NOT involved in startups get used to using it,
> it's not going to get brought into startups by the kids starting them.
> And these kids aren't going to pay a dime for a great Lisp (and won't
> even think about using a poor one -- i.e., one which doesn't work on
> everything, install easily, have full CPAN support, and something like
> RAILS is rapidly becoming a part of the required library set!) because
> they don't have to pay a dime to use great Rubies, Pytons, Javas, and
> C++s.
Well, SBCL runs on Linux, Mac OS, Solaris, *BSD & Windows; it installs
very easily; it support ASDF-INSTALL. Similar things can be said about
other free CLs. There's not yet any RoR-like framework, but the
necessary pieces are there; it's just a Simple Matter of Programming.
I'm working on my own (inspired more by Django than RoR), and it's a
dead cert that some far better Lispers are doing the same.
> (Actually, I think that our real only hope is that everyone converts
> to Ruby and JSON, which are approximately Lisp and SExprs, and then
> someone in the Ruby community will realize that you can represent you
> code elegantly in JSON, and that that gives you Macros...and then Lisp
> is reborn!)
Bleh--Ruby makes me think of Perl.
--
Robert Uhl <http://public.xdi.org/=ruhl>
As a client for MS Exchange, MS Outlook is quite good. As an Internet
e-mail client [e.g, POP3/IMAP], it's roughly equivalent to strapping a
few pounds of plastique to your gonads and painting a day-glo orange
bulls-eye on your knickers. -- Morely Dotes in nan-ae
On Jun 14, 11:16 pm, Ron Garret <·········@flownet.com> wrote:
>
> The current landscape of Lisp implementations seems to have a big hole
> in the middle, in the sense that there are a fairly large number of free
> implementations, all of which are deficient in one way or another [1],
> and a smaller number of commercial implementations which are less
> deficient but all very expensive. There is no Lisp of commercial
> quality which is affordable by the new breed of self-funded web 2.0
> startups running on a shoestring, the result of which is that most such
> startups are using Python, Ruby, or PHP instead.
I think I'm beginning to realise my problem. I keep making the
assumption that people in cll are kind of like me[1]. In particular I
assume that people tend to have jobs and so on. But actually I
suspect everyone who posts here lives in a cardboard box under a
bridge somewhere, surviving on a diet of white cider and fag ends
(including, obviously, the web 2.0 people). I mean, what stops me
buying a license for one (not the most expensive) of the commercial
CLs isn't cost it's that I won't actually have time to use it, and
given that I'd rather spend my money on skiing or something. I think
you guys should dry out and get a job.
--tim
[1] Other than the tentacles, of course.
I am not sure. I have a pretty well paying job, but Lisp is mostly
hobby for me. With two kids, mortgage and older kid approaching
college age I'll have a trouble convincing my wife to fork few
hundreds buck for CL license.
When I was grad student, it would be even more difficult. At that time
there were no slime, ilisp was a bitch, no asdf and cmucl would not
work on Linux yet. These are some reasons why my primary language for
a long time was Python + C/C++ when speed was needed. It is possible
that if I got addicted to lisp at that time, I'd buy couple of
licenses for work later.
Now sbcl+slime is getting pretty close to my ideal enviroment, asdf
made installing libraries way easier. I am starting use lisp more and
more. But I am pretty used to emacs...
I do not think high cost of license is a problem for using Lisp
commercially. It is a problem for college /grad school people who
would be willing to test it for couple of weeks and would drop it with
99% probability. This audience times 0.01 is probably much higher then
current Lisp users base.
I am not sure this is such a problem now, because as I mentioned sbcl
+slime is pretty good environment for me (although getting used to it
for IDE folks is pretty difficult). Visual smalltalk folks decided to
make their environment freely available for noncommercial use and they
are pretty happy. I heard they had some big commercial users who
started with non-commercial edition first and later spend big bucks on
licenses.
Anyway, it is up to Allegro/Lispworks folks to decide how to handle
this situation. If in few years I'll use Lisp commercially, I probably
will have a look at what they have to offer. Then may be not -- I am
not familiar with these reportedly very good implementations and
getting used to them takes time, time that hungry students do have and
people with jobs do not. Meanwhile if I save few hundreds bucks now I
probably spend it getting new harness for my hang glider instead of
hobby Allegro license. Taking care of my tentacles costs money too :-)
On Jun 18, 2:44 pm, Tim Bradshaw <··········@tfeb.org> wrote:
> I think I'm beginning to realise my problem. I keep making the
> assumption that people in cll are kind of like me[1]. In particular I
> assume that people tend to have jobs and so on. But actually I
> suspect everyone who posts here lives in a cardboard box under a
> bridge somewhere, surviving on a diet of white cider and fag ends
> (including, obviously, the web 2.0 people). I mean, what stops me
> buying a license for one (not the most expensive) of the commercial
> CLs isn't cost it's that I won't actually have time to use it, and
> given that I'd rather spend my money on skiing or something. I think
> you guys should dry out and get a job.
>
> --tim
>
> [1] Other than the tentacles, of course.
From: Ken Tilton
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <XyBdi.113$ys7.41@newsfe12.lga>
> On Jun 14, 11:16 pm, Ron Garret <·········@flownet.com> wrote:
>
>
>>The current landscape of Lisp implementations seems to have a big hole
>>in the middle, in the sense that there are a fairly large number of free
>>implementations, all of which are deficient in one way or another [1],
>>and a smaller number of commercial implementations which are less
>>deficient but all very expensive. There is no Lisp of commercial
>>quality which is affordable by the new breed of self-funded web 2.0
>>startups running on a shoestring, the result of which is that most such
>>startups are using Python, Ruby, or PHP instead.
I could have sworn we just had to change the name to make Lisp popular.
Or was it lose the prefix thing? This thread makes Freecell look like an
intelligent use of free time. I just hope we can continue debating how
to make Lisp more popular after it is the most popular language.
kzo
From: Ken Tilton
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <GYBdi.18$ME5.3@newsfe12.lga>
Tim Bradshaw wrote:
> On Jun 14, 11:16 pm, Ron Garret <·········@flownet.com> wrote:
>
>
>>The current landscape of Lisp implementations seems to have a big hole
>>in the middle, in the sense that there are a fairly large number of free
>>implementations, all of which are deficient in one way or another [1],
>>and a smaller number of commercial implementations which are less
>>deficient but all very expensive. There is no Lisp of commercial
>>quality which is affordable by the new breed of self-funded web 2.0
>>startups running on a shoestring, the result of which is that most such
>>startups are using Python, Ruby, or PHP instead.
>
>
> I think I'm beginning to realise my problem. I keep making the
> assumption that people in cll are kind of like me[1]. In particular I
> assume that people tend to have jobs and so on.
In particular my problem is that I assume people here use Lisp, which is
true neither of you nor Ron. In fact, I am starting to think this NG is
just a hangout for Lisp wannabes and burnouts -- the middle ground is
too busy having fun writing Lisp for this BS.
hth,kzo
In article <··············@newsfe12.lga>,
Ken Tilton <···········@optonline.net> wrote:
> Tim Bradshaw wrote:
> > On Jun 14, 11:16 pm, Ron Garret <·········@flownet.com> wrote:
> >
> >
> >>The current landscape of Lisp implementations seems to have a big hole
> >>in the middle, in the sense that there are a fairly large number of free
> >>implementations, all of which are deficient in one way or another [1],
> >>and a smaller number of commercial implementations which are less
> >>deficient but all very expensive. There is no Lisp of commercial
> >>quality which is affordable by the new breed of self-funded web 2.0
> >>startups running on a shoestring, the result of which is that most such
> >>startups are using Python, Ruby, or PHP instead.
> >
> >
> > I think I'm beginning to realise my problem. I keep making the
> > assumption that people in cll are kind of like me[1]. In particular I
> > assume that people tend to have jobs and so on.
>
> In particular my problem is that I assume people here use Lisp, which is
> true neither of you nor Ron. In fact, I am starting to think this NG is
> just a hangout for Lisp wannabes and burnouts -- the middle ground is
> too busy having fun writing Lisp for this BS.
>
> hth,kzo
Give Tim some credit: At least he is not proposing vague schemes (!)
to reach world domination.
--
http://lispm.dyndns.org
From: Holger Schauer
Subject: Re: A problem statement (and a proposed solution)
Date:
Message-ID: <yxzps3tiezb.fsf@gmx.de>
On 5035 September 1993, Ron Garret wrote:
> The current landscape of Lisp implementations seems to have a big hole
> in the middle, in the sense that there are a fairly large number of free
> implementations, all of which are deficient in one way or another [1],
[...]
Sorry to jump in between, but I don't see a footnote explaining this
claim. I'm not saying I disagree but would like to see a list of the
defects that you think are relevant.
> and a smaller number of commercial implementations which are less
> deficient but all very expensive. There is no Lisp of commercial
> quality which is affordable by the new breed of self-funded web 2.0
> startups running on a shoestring, the result of which is that most such
> startups are using Python, Ruby, or PHP instead.
Is this really the 'deficiancies of affordable lisp implementations'
or the lack of an affordable lisp of commercial quality (whatever that
means) that's the reason? Or is it just that Python, PHP and Ruby have
a strong appeal to your self-funded web 2.0 start-up founder because
of the libraries/frameworks available (usually as an add-on) that
ease the pain hacking up some initial functionality? I think this has
much more to do with the funcitonality of those libraries/frameworks
than with the languages tied to them or with the prices paid for the
implementation used for compiling your app in the first place.
Holger
--
--- http://hillview.bugwriter.net/ ---
Fachbegriffe der Informatik - Einfach erkl�rt
276: SMP
Fehlfunktion bei mehr als einer CPU. (nach Holger Veit)
Ron Garret wrote:
> If I were starting a company I would approach the commercial Lisp
> vendor...
Franz did not response to our requests.
Incidentally, what do the commercial Lisp's have that SBCL doesn't have and
that a web developer would want? I was under the impression that the
commercial Lisp's were primarily about legacy code rather than anything
modern...
--
Dr Jon D Harrop, Flying Frog Consultancy
The OCaml Journal
http://www.ffconsultancy.com/products/ocaml_journal/?usenet
Jon Harrop wrote:
> Ron Garret wrote:
>> If I were starting a company I would approach the commercial Lisp
>> vendor...
>
> Franz did not response to our requests.
>
> Incidentally, what do the commercial Lisp's have that SBCL doesn't have and
> that a web developer would want? I was under the impression that the
> commercial Lisp's were primarily about legacy code rather than anything
> modern...
Oh, a stress test for Duane... ;)
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
On Jun 13, 2:13 am, ·········@gmail.com" <········@gmail.com> wrote:
> (long, complex plan for world domination snipped)
I still don't understand why you're proposing this long, complex plan
for buying out a commercial lisp distribution instead of just sprucing
up SBCL with an IDE and include batteries to get mostly where you want
to be. Can you be specific about what's so awful about SBCL besides
the lack of polish?
It would probably be easier to entice current SBCL developers to work
for you anyway, as they could then do for money what they currently do
only for love (although for some of the developers, you may have been
already beaten to it; I think I recall reading a SBCL developer's blog
who said he was hired explicitly to contine working on SBCL for pay).
So write up your secret semantic web business plan, get funded by
YCombinator (they certainly won't have a problem with you writing
software in Lisp), get an SBCL core developer as cofounder, dogfood
your own World Beating Lisp Development Environment while you get
rich, then release to the world.
Much simpler, right?
········@gmail.com wrote:
>>>> This experiment has already been done. OpenMCL is the result.
>>> Interesting. Not being a PPC user I hadn't considered this, but
>>> apparently they've ported it to x86 Linux now as well
>> And OS X as well. I run it on my Macbook Pro. 32-bit processors are so
>> last-millennium.
>>
>>> (although it's only in pre-released)
>> You wanted open-source, this is what you get.
>
> Hmmm. Well, I didn't actually want open source in the sense of not
> professionally supported. What I wanted (although you'd have to study
> the thread to get this) is a company that supports an excellent CL on-
> the-side and makes their money through verticals, like Google supports
> Guido to do python, and Sun/Java. So, MCL probably doesn't actually
> fit the bill afterall.
They exactly fit the bill, AFAICT.
See http://clozure.com/
Pascal
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
(message (Hello ·········@gmail.com)
(you :wrote :on '(Mon, 11 Jun 2007 14:05:49 -0700))
(
??>> This experiment has already been done. OpenMCL is the result.
J> Interesting. Not being a PPC user I hadn't considered this, but
J> apparently they've ported it to x86 Linux now as well
only 64-bit, which is not yet mainstream.
few days ago i've installed some AMD64 Ubuntu systems -- some apps don't
work there, mostly proprietary with 32-bit binaries.
)
(With-best-regards '(Alex Mizrahi) :aka 'killer_storm)
"I am everything you want and I am everything you need")
········@gmail.com wrote:
> Then someone paid to do something like Rails in the foregoing new
> modernized OS Lisp?
If you want a modernized Lisp you should look at some of the modern Lisp
variants, like OCaml, F#, Scala, Nemerle and Haskell.
--
Dr Jon D Harrop, Flying Frog Consultancy
The OCaml Journal
http://www.ffconsultancy.com/products/ocaml_journal/?usenet