I concede (happily) that Naughty Dog is a Lisp success story. So now I
have two in the last decade. Are there any others that I've missed?
By "success story" I mean that 1) someone made a significant amount of
money (more than just a salary) and 2) the principals attribute their
success at least in part to the use of Lisp.
Is ITA turning a profit?
rg
Name: Afferent, Inc.
Domain: Simulated Combinatorial Chemistry and Drug Robotics
Tools: Lisp (ACL), Tk (interface), and some C for DB and robot
interfacing
Size: Probably hundreds of thousands of lines of Lisp, and a few
thousand in other languages.
Chronology: Sellable prototype hacked by 1 engineer in ~2 years.
Total engineers at acquisition: 5. (~4 years after founding)
Acquired in 2000 (I think it was) by MDL, Inc.
ROI (mine and the founder): ~1000x (pretax)
[I'm not sure I can reveal the $ figures bcs it was a private company
acquired by a private company.]
My roles: First employee (after founder), Director of Engineering,
lead DB engineer, lead user support engineer.
lead phone re-wirer, lead janitor...
IMHO: There was no way in hell we could have come anywhere close to
this w/o Lisp. It did simulated chemistry and experiment planning (for
the robots) -- classical Lisp domains. And there's also no way we could
have done the interface w/o macrofying the crap out of Tk. (Tk itself
is a horrific nightmare! The lisp macros hid the horror and enabled us
to do extremely complex things!) In addition to Lisp enabling rapid
development (to prototype and beyond) we were also able to patch
running systems on the fly at user sites, and produce user-specific
mods in near real time. IMHO, as the world moves (rapidly) toward more
toward knowledge-based computing, more and more problems of this sort
will arise, and more and more money will be on the table to be picked
up by folks who what to do what we did in what was at that time a
rarified domain. Let the mountain come to Mohammed.
ps. More about Afferent:
http://www.franz.com/success/customer_apps/bioinformatics/mdl_story.lhtml
http://www.mdl.com/company/news/press_releases/2000/pr_69.jsp
And this >fascinating< paper:
http://nostoc.stanford.edu/jeff/personal/vita/pubs/2001designforscience.pdf
;-)
> Did the buyers port the code to another language?
Unlikley -- they'd never have been able to port the core logic. Maybe
they could have ported the interface to Java or .NET or something
(which would have been a blessing considering how much of a nightmare
Tk was!)
Ron Garret wrote:
> By "success story" I mean that 1) someone made a significant amount of
> money (more than just a salary) and 2) the principals attribute their
> success at least in part to the use of Lisp.
I don't find out the relation between using one or another tool for
making a significant amount of money or not.
Using Lisp is just a matter of taste, has nothing to do with marketing
issues (at least up to the level you are talking).
Take for example Viaweb from PG. It was once written in Lisp, and some
time after rewritten to C++; and it mantained if not incresed its
succes.
It is the idea and being able to develop it what matters.
Ron Garret schrieb:
> I concede (happily) that Naughty Dog is a Lisp success story. So now I
> have two in the last decade. Are there any others that I've missed?
>
> By "success story" I mean that 1) someone made a significant amount of
> money (more than just a salary) and 2) the principals attribute their
> success at least in part to the use of Lisp.
>
> Is ITA turning a profit?
>
> rg
>
How about Gensym's G2? isn't that still Lisp under the hood?
-klaus
> Is ITA turning a profit?
>
> rg
From talking to some friends who work there, I believe ITA is
extremely profitable. They are hiring good LISPers as fast as they
can find them.
rif
Is every Lisp developer infected by this PG virus? Success = $?
Why are so many Lisp people so obsessed about getting rich?
--
Web (en): http://www.no-spoon.de/ -*- Web (de): http://www.frell.de/
Stefan Scholl wrote:
> Is every Lisp developer infected by this PG virus? Success = $?
>
> Why are so many Lisp people so obsessed about getting rich?
I certainly don't want to be poor, but even as a newcomer to the
language, being able to use Lisp rather than .Not, Java, or the
scripting-language-of-the-month is its own reward.
In article <······················@parsec.no-spoon.de>,
Stefan Scholl <······@no-spoon.de> wrote:
> Is every Lisp developer infected by this PG virus? Success = $?
No, but it is not an altogether unreasonable choice for a quality
metric. If nothing else, it's easy to measure.
> Why are so many Lisp people so obsessed about getting rich?
I don't think they are. I think most Lisp people would be content just
to break even (i.e. to be able to earn a living using Lisp). The
problem is that it's hard to find investors who share that attitude.
rg
From: Pascal Costanza
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4lu657F3mh29U1@individual.net>
Ron Garret wrote:
> In article <······················@parsec.no-spoon.de>,
> Stefan Scholl <······@no-spoon.de> wrote:
>
>> Is every Lisp developer infected by this PG virus? Success = $?
>
> No, but it is not an altogether unreasonable choice for a quality
> metric. If nothing else, it's easy to measure.
>
>> Why are so many Lisp people so obsessed about getting rich?
>
> I don't think they are. I think most Lisp people would be content just
> to break even (i.e. to be able to earn a living using Lisp). The
> problem is that it's hard to find investors who share that attitude.
The problem is that the cases where people break even rarely make the
headlines.
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/
In article <··············@individual.net>,
Pascal Costanza <··@p-cos.net> wrote:
> Ron Garret wrote:
> > In article <······················@parsec.no-spoon.de>,
> > Stefan Scholl <······@no-spoon.de> wrote:
> >
> >> Is every Lisp developer infected by this PG virus? Success = $?
> >
> > No, but it is not an altogether unreasonable choice for a quality
> > metric. If nothing else, it's easy to measure.
> >
> >> Why are so many Lisp people so obsessed about getting rich?
> >
> > I don't think they are. I think most Lisp people would be content just
> > to break even (i.e. to be able to earn a living using Lisp). The
> > problem is that it's hard to find investors who share that attitude.
>
> The problem is that the cases where people break even rarely make the
> headlines.
Why is that a problem?
rg
Ron Garret skrev:
> In article <··············@individual.net>,
> Pascal Costanza <··@p-cos.net> wrote:
>
> > Ron Garret wrote:
> > > In article <······················@parsec.no-spoon.de>,
> > > Stefan Scholl <······@no-spoon.de> wrote:
> > >
> > >> Is every Lisp developer infected by this PG virus? Success = $?
> > >
> > > No, but it is not an altogether unreasonable choice for a quality
> > > metric. If nothing else, it's easy to measure.
> > >
> > >> Why are so many Lisp people so obsessed about getting rich?
> > >
> > > I don't think they are. I think most Lisp people would be content just
> > > to break even (i.e. to be able to earn a living using Lisp). The
> > > problem is that it's hard to find investors who share that attitude.
> >
> > The problem is that the cases where people break even rarely make the
> > headlines.
>
> Why is that a problem?
See http://groups.google.no/group/comp.lang.lisp/msg/03b1939f9b676c4c
--
mvh, Lars Rune Nøstdal
http://lars.nostdal.org/
In article <························@b28g2000cwb.googlegroups.com>,
"Lars Rune N�stdal" <···········@gmail.com> wrote:
> Ron Garret skrev:
> > In article <··············@individual.net>,
> > Pascal Costanza <··@p-cos.net> wrote:
> >
> > > Ron Garret wrote:
> > > > In article <······················@parsec.no-spoon.de>,
> > > > Stefan Scholl <······@no-spoon.de> wrote:
> > > >
> > > >> Is every Lisp developer infected by this PG virus? Success = $?
> > > >
> > > > No, but it is not an altogether unreasonable choice for a quality
> > > > metric. If nothing else, it's easy to measure.
> > > >
> > > >> Why are so many Lisp people so obsessed about getting rich?
> > > >
> > > > I don't think they are. I think most Lisp people would be content just
> > > > to break even (i.e. to be able to earn a living using Lisp). The
> > > > problem is that it's hard to find investors who share that attitude.
> > >
> > > The problem is that the cases where people break even rarely make the
> > > headlines.
> >
> > Why is that a problem?
>
> See http://groups.google.no/group/comp.lang.lisp/msg/03b1939f9b676c4c
That doesn't really answer my question.
rg
From: Pascal Costanza
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4lvhhfF3r2rfU1@individual.net>
Ron Garret wrote:
> In article <··············@individual.net>,
> Pascal Costanza <··@p-cos.net> wrote:
>
>> Ron Garret wrote:
>>> In article <······················@parsec.no-spoon.de>,
>>> Stefan Scholl <······@no-spoon.de> wrote:
>>>
>>>> Is every Lisp developer infected by this PG virus? Success = $?
>>> No, but it is not an altogether unreasonable choice for a quality
>>> metric. If nothing else, it's easy to measure.
>>>
>>>> Why are so many Lisp people so obsessed about getting rich?
>>> I don't think they are. I think most Lisp people would be content just
>>> to break even (i.e. to be able to earn a living using Lisp). The
>>> problem is that it's hard to find investors who share that attitude.
>> The problem is that the cases where people break even rarely make the
>> headlines.
>
> Why is that a problem?
You're not looking for successes, but you are looking for the appearance
of success. Things that don't make the headlines won't make you happy in
this regard.
Today's Dilbert strip is related. See
http://www.dilbert.com/comics/dilbert/archive/dilbert-20060903.html
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 wrote:
>
> You're not looking for successes, but you are looking for the appearance
> of success. Things that don't make the headlines won't make you happy in
> this regard.
>
> Today's Dilbert strip is related. See
> http://www.dilbert.com/comics/dilbert/archive/dilbert-20060903.html
Sorry, I don't really believe there are many Lisp guys toiling away in
the bowels of various organizations, providing value add with no
acknowledgement for what they're up to. Sounds like a fantasy to
justify the lack of Lisp in industry. There are of course projects and
programming problems where one could pick "any language" or "any of
several languages." People do that; for instance, a friend of mine was
doing Python code at Microsoft because no one was really paying
attention. And then he jumped ship to Google. But with Python,
there's generally somewhere to jump, and it's far more likely to be
chosen as the "any language."
Cheers,
Brandon Van Every
In article <··············@individual.net>,
Pascal Costanza <··@p-cos.net> wrote:
> Ron Garret wrote:
> > In article <··············@individual.net>,
> > Pascal Costanza <··@p-cos.net> wrote:
> >
> >> Ron Garret wrote:
> >>> In article <······················@parsec.no-spoon.de>,
> >>> Stefan Scholl <······@no-spoon.de> wrote:
> >>>
> >>>> Is every Lisp developer infected by this PG virus? Success = $?
> >>> No, but it is not an altogether unreasonable choice for a quality
> >>> metric. If nothing else, it's easy to measure.
> >>>
> >>>> Why are so many Lisp people so obsessed about getting rich?
> >>> I don't think they are. I think most Lisp people would be content just
> >>> to break even (i.e. to be able to earn a living using Lisp). The
> >>> problem is that it's hard to find investors who share that attitude.
> >> The problem is that the cases where people break even rarely make the
> >> headlines.
> >
> > Why is that a problem?
>
> You're not looking for successes, but you are looking for the appearance
> of success. Things that don't make the headlines won't make you happy in
> this regard.
Not true. I never asked for headlines, I just asked for the particulars
to be documentable.
rg
From: Pascal Costanza
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m4pfjF4canpU1@individual.net>
Ron Garret wrote:
> In article <··············@individual.net>,
> Pascal Costanza <··@p-cos.net> wrote:
>
>> Ron Garret wrote:
>>> In article <··············@individual.net>,
>>> Pascal Costanza <··@p-cos.net> wrote:
>>>
>>>> Ron Garret wrote:
>>>>> In article <······················@parsec.no-spoon.de>,
>>>>> Stefan Scholl <······@no-spoon.de> wrote:
>>>>>
>>>>>> Is every Lisp developer infected by this PG virus? Success = $?
>>>>> No, but it is not an altogether unreasonable choice for a quality
>>>>> metric. If nothing else, it's easy to measure.
>>>>>
>>>>>> Why are so many Lisp people so obsessed about getting rich?
>>>>> I don't think they are. I think most Lisp people would be content just
>>>>> to break even (i.e. to be able to earn a living using Lisp). The
>>>>> problem is that it's hard to find investors who share that attitude.
>>>> The problem is that the cases where people break even rarely make the
>>>> headlines.
>>> Why is that a problem?
>> You're not looking for successes, but you are looking for the appearance
>> of success. Things that don't make the headlines won't make you happy in
>> this regard.
>
> Not true. I never asked for headlines, I just asked for the particulars
> to be documentable.
You won't get the particulars by asking around in Usenet.
We're going in circles again. If you consider projects that "merely" pay
the bills as financial successes, it's unlikely that you will hear about
them because people don't flaunt them. If you want to hear about
projects that go substantially beyond that wrt financial return, it's
unlikely that you will hear about them because Lisp is not a hype.
Partially, that's because there's a chicken-and-egg problem here, and
partially that's because the tools are not necessarily the important
ingredients. [1]
If you consider research projects as valid in your search for success
stories, count me in. (I can definitely pay my bills and I use Common
Lisp almost exclusively.) Also don't forget the various other research
projects that use Common Lisp and the apparently more widely used
Scheme. For example, PLT Scheme is definitely a huge success in this regard.
If you regard some of these projects as invalid for your purposes, you
should be much more specific, and especially conduct a more serious
study than just randomly polling newsgroups.
Pascal
[1] Quiz question: For how many financial successes do you _actually_
know what technology they have used to implement them? Don't forget to
separate claims from actual facts - for example, see the bit about
Oracle in the entry "What database did you use?" at
http://www.paulgraham.com/vwfaq.html
--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
In article <··············@individual.net>,
Pascal Costanza <··@p-cos.net> wrote:
> Ron Garret wrote:
> > In article <··············@individual.net>,
> > Pascal Costanza <··@p-cos.net> wrote:
> >
> >> Ron Garret wrote:
> >>> In article <··············@individual.net>,
> >>> Pascal Costanza <··@p-cos.net> wrote:
> >>>
> >>>> Ron Garret wrote:
> >>>>> In article <······················@parsec.no-spoon.de>,
> >>>>> Stefan Scholl <······@no-spoon.de> wrote:
> >>>>>
> >>>>>> Is every Lisp developer infected by this PG virus? Success = $?
> >>>>> No, but it is not an altogether unreasonable choice for a quality
> >>>>> metric. If nothing else, it's easy to measure.
> >>>>>
> >>>>>> Why are so many Lisp people so obsessed about getting rich?
> >>>>> I don't think they are. I think most Lisp people would be content just
> >>>>> to break even (i.e. to be able to earn a living using Lisp). The
> >>>>> problem is that it's hard to find investors who share that attitude.
> >>>> The problem is that the cases where people break even rarely make the
> >>>> headlines.
> >>> Why is that a problem?
> >> You're not looking for successes, but you are looking for the appearance
> >> of success. Things that don't make the headlines won't make you happy in
> >> this regard.
> >
> > Not true. I never asked for headlines, I just asked for the particulars
> > to be documentable.
>
> You won't get the particulars by asking around in Usenet.
I did not ask for the particulars. I only asked for the particulars to
be documentable. (Why do you have such a penchant for straw-man
arguments?)
> We're going in circles again. If you consider projects that "merely" pay
> the bills as financial successes, it's unlikely that you will hear about
> them because people don't flaunt them.
Not true. They flaunt them a lot, and I am therefore familiar with a
large number of them. But such stories are not useful for my present
purposes.
> If you want to hear about
> projects that go substantially beyond that wrt financial return, it's
> unlikely that you will hear about them because Lisp is not a hype.
Or maybe it's because they don't exist.
> Partially, that's because there's a chicken-and-egg problem here, and
> partially that's because the tools are not necessarily the important
> ingredients. [1]
Well, that is the $64,000 question isn't it. Does Lisp confer a
competitive advantage in business or doesn't it? If it does, where is
the evidence?
> If you consider research projects as valid in your search for success
> stories, count me in.
Did your research produce a product that resulted in a substantial and
documentable ROI for an investor?
> If you regard some of these projects as invalid for your purposes, you
> should be much more specific,
I'm looking for documentable cases where an endeavor produced a
substantial (let's say >15% annualized) and documentable return for an
investor and the participants attribute their success at least in part
to the competitive advantage gained by using Lisp. I don't know how
much more specific I can be.
> and especially conduct a more serious
> study than just randomly polling newsgroups.
You have no way of knowing what other measures I may taking to obtain
this information (to say nothing of the fact that my polling is far from
random), but your defensiveness is to me an indication that the data I
seek do not exist.
> [1] Quiz question: For how many financial successes do you _actually_
> know what technology they have used to implement them? Don't forget to
> separate claims from actual facts - for example, see the bit about
> Oracle in the entry "What database did you use?" at
> http://www.paulgraham.com/vwfaq.html
I can think of at least two right off the top of my head:
http://www.aaronsw.com/weblog/rewritingreddit
http://masshightech.bizjournals.com/masshightech/stories/2006/07/03/story
13.html
And:
http://xooglers.blogspot.com/2005/12/lets-get-real-database.html
Finding documentation for the financial-success part of that second
example is left as an exercise.
rg
From: Don Geddis
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <87zmddudlf.fsf@geddis.org>
Ron Garret <·········@flownet.com> wrote on Tue, 05 Sep 2006:
> Does Lisp confer a competitive advantage in business or doesn't it?
Rarely.
> If it does, where is the evidence?
Let's accept that Lisp is a much better designed language than most
alternatives, and that if we were king, the world would be a better place if
we forced everyone to program in Lisp.
Even given that, here's why Lisp doesn't make a better business:
1. Technology, with rare exceptions, is only a small part of what makes a
business succeed. Your software engineering team might be twice as good as
your competitor's, but that may only give the business itself a 5% advantage
(which will be swamped by other decisions that you make).
2. Lisp is not popular. Hence there is a lot about the third party ecosystem
where Lisp is worse than alternative. Number of trained programmers coming
out of college, conferences, books available, open source software written in
the language, libraries, etc. If someone invents a new kind of public-key
encryption, I doubt Common Lisp will be the first implementation language
released. So it's even questionable whether your Lisp-based software
engineering team IS more efficient than your competitor.
3. Finally, there's the question of whether Lisp really does make for more
productive programmers. It's hard to evaluate this objectively. Most of us
on this newsgroup believe it to be true, but there's precious little evidence.
And we find that fans of other languages are often just as ardent in support
of their favorites.
Summary: Lisp may be better on its own than other languages, but it's hard
to tell; even if it was better in isolation, it may not be better when you
consider the whole programming ecosystem; even if the ecosystem is better, it's
a rare business where that will make-or-break the financial outcome.
I view it much more like English vs. metric measurements in the US. Or the
QWERTY keyboard. There's a better solution out there, but it isn't ENOUGH
better to make a real difference, so inertia wins and nobody changes.
-- Don
_______________________________________________________________________________
Don Geddis http://don.geddis.org/ ···@geddis.org
Cthulhu for President. Don't settle for the lesser evil! -- cthulhu.org
From: Robert Uhl
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m3r6yojqv3.fsf@NOSPAMgmail.com>
Don Geddis <···@geddis.org> writes:
>
> I view it much more like English vs. metric measurements in the US.
Actually, standard units are Lisp and French units are C++. But that's
an argument for a different newsgroup:-)
--
Robert Uhl <http://public.xdi.org/=ruhl>
My Hope is the Father, my Refuge is the Son, my Shelter is the Holy Spirit.
O Holy Trinity, our God, glory to Thee!
Don Geddis <···@geddis.org> writes:
[...]
> Let's accept that Lisp is a much better designed language than most
> alternatives, and that if we were king, the world would be a better
> place if we forced everyone to program in Lisp.
Would that be good? It seems at least plausible that for some
applications (and perhaps for some people) statically typed languages
are better than Lisp (or at least, Common Lisp). (Obviously I'm
thinking more of statically typed languages with type inference than
of languages like Java, C++, where programmers have to spell the
obvious out.)
[...]
> 3. Finally, there's the question of whether Lisp really does make
> for more productive programmers. It's hard to evaluate this
> objectively. Most of us on this newsgroup believe it to be true,
> but there's precious little evidence. And we find that fans of
> other languages are often just as ardent in support of their
> favorites.
There's at least reasonable evidence that powerful languages are more
productive than less powerful ones (by "powerful" I'm just thinking of
functionality per line of source, or something like that), isn't
there? (I'm guessing that's still true: it's certainly quoted
repeatedly that rates of code production tend to be about constant
regardless of language (but vary wildy from individual to individual),
but I guess probably based on research that's years (or decades) old.)
[...]
From: Steven E. Harris
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <831wqomm12.fsf@torus.sehlabs.com>
Bruce Stephens <············@cenderis.demon.co.uk> writes:
> it's certainly quoted repeatedly that rates of code production tend
> to be about constant regardless of language
I find this to be true in terms of code /volume/, but the code I write
in Lisp against, say, Java does a lot more per unit volume. There's
less cruft and greater density of useful, observable "content" being
produced.
--
Steven E. Harris
From: Don Geddis
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <87r6yniiv3.fsf@geddis.org>
> Don Geddis <···@geddis.org> writes:
>> Let's accept that Lisp is a much better designed language than most
>> alternatives, and that if we were king, the world would be a better
>> place if we forced everyone to program in Lisp.
Bruce Stephens <············@cenderis.demon.co.uk> wrote on Wed, 06 Sep 2006:
> Would that be good? It seems at least plausible that for some
> applications (and perhaps for some people) statically typed languages
> are better than Lisp (or at least, Common Lisp). (Obviously I'm
> thinking more of statically typed languages with type inference than
> of languages like Java, C++, where programmers have to spell the
> obvious out.)
There's an argument to be made that Lisp potentially offers optional
static typing, and that this is the best of both worlds. That, just like
optimization, you can choose to statically type portions of your code, without
being forced to statically type all of it.
I'm aware that there are programmers that have different opinions.
> There's at least reasonable evidence that powerful languages are more
> productive than less powerful ones (by "powerful" I'm just thinking of
> functionality per line of source, or something like that)
It's certainly true that programmers are more productive in high-level
languages than low level ones. Nobody is suggesting assembly is as productive
as Lisp.
But comparing Lisp to Java or C++ or Python or Perl or Ruby is much harder.
-- Don
_______________________________________________________________________________
Don Geddis http://don.geddis.org/ ···@geddis.org
Give me the strength to change the things I can, the grace to accept the things
I cannot, and a great big bag of money. -- Imitation Deep Thoughts [Age 13]
From: Pascal Costanza
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m5hmrF4ilo7U1@individual.net>
Ron Garret wrote:
> In article <··············@individual.net>,
> Pascal Costanza <··@p-cos.net> wrote:
>
>> Ron Garret wrote:
>>> In article <··············@individual.net>,
>>> Pascal Costanza <··@p-cos.net> wrote:
>>>
>>>> Ron Garret wrote:
>>>>> In article <··············@individual.net>,
>>>>> Pascal Costanza <··@p-cos.net> wrote:
>>>>>
>>>> You're not looking for successes, but you are looking for the appearance
>>>> of success. Things that don't make the headlines won't make you happy in
>>>> this regard.
>>> Not true. I never asked for headlines, I just asked for the particulars
>>> to be documentable.
>> You won't get the particulars by asking around in Usenet.
>
> I did not ask for the particulars. I only asked for the particulars to
> be documentable. (Why do you have such a penchant for straw-man
> arguments?)
Sorry, I don't understand the difference.
>> If you want to hear about
>> projects that go substantially beyond that wrt financial return, it's
>> unlikely that you will hear about them because Lisp is not a hype.
>
> Or maybe it's because they don't exist.
Maybe. Maybe not.
>> If you consider research projects as valid in your search for success
>> stories, count me in.
>
> Did your research produce a product that resulted in a substantial and
> documentable ROI for an investor?
No, not yet.
>> If you regard some of these projects as invalid for your purposes, you
>> should be much more specific,
>
> I'm looking for documentable cases where an endeavor produced a
> substantial (let's say >15% annualized) and documentable return for an
> investor and the participants attribute their success at least in part
> to the competitive advantage gained by using Lisp. I don't know how
> much more specific I can be.
Thank you. It was indeed not clear to me what you were specifically
asking for.
>> and especially conduct a more serious
>> study than just randomly polling newsgroups.
>
> You have no way of knowing what other measures I may taking to obtain
> this information (to say nothing of the fact that my polling is far from
> random), but your defensiveness is to me an indication that the data I
> seek do not exist.
So where's the study where people can participate to give you the
information you need?
>> [1] Quiz question: For how many financial successes do you _actually_
>> know what technology they have used to implement them? Don't forget to
>> separate claims from actual facts - for example, see the bit about
>> Oracle in the entry "What database did you use?" at
>> http://www.paulgraham.com/vwfaq.html
>
> I can think of at least two right off the top of my head:
>
> http://www.aaronsw.com/weblog/rewritingreddit
> http://masshightech.bizjournals.com/masshightech/stories/2006/07/03/story
> 13.html
I don't see where they claim that Python gave them a competitive
advantage. To the contrary, Marc Battyani has in fact implemented a
clone in virtually no time in Common Lisp.
Yes, they probably made more money than him, but that's unrelated.
Reddit could have sticked to Lisp and probably would have made the same
amount of money. That's because the language is not relevant here.
I basically agree with Tim Bradshaw: Lisp can buy you about six months
under the right circumstances. That's about it.
> And:
>
> http://xooglers.blogspot.com/2005/12/lets-get-real-database.html
>
> Finding documentation for the financial-success part of that second
> example is left as an exercise.
...and where does this example refer to the competitive advantage of
some language? Was the other database they tried not based on SQL or on
a relational model?
Or are you not interested in Lisp as language, but Lisp as a ... what?
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/
In article <··············@individual.net>,
Pascal Costanza <··@p-cos.net> wrote:
> > I did not ask for the particulars. I only asked for the particulars to
> > be documentable. (Why do you have such a penchant for straw-man
> > arguments?)
>
> Sorry, I don't understand the difference.
You don't understand the difference between making a claim for which
supporting documentation exists and actually providing that
documentation? And you're a candidate for a Ph.D.? Standards in German
higher education must be slipping.
> >> and especially conduct a more serious
> >> study than just randomly polling newsgroups.
> >
> > You have no way of knowing what other measures I may taking to obtain
> > this information (to say nothing of the fact that my polling is far from
> > random), but your defensiveness is to me an indication that the data I
> > seek do not exist.
>
> So where's the study where people can participate to give you the
> information you need?
Study? Who said anything about a study?
> >> [1] Quiz question: For how many financial successes do you _actually_
> >> know what technology they have used to implement them? Don't forget to
> >> separate claims from actual facts - for example, see the bit about
> >> Oracle in the entry "What database did you use?" at
> >> http://www.paulgraham.com/vwfaq.html
> >
> > I can think of at least two right off the top of my head:
> >
> > http://www.aaronsw.com/weblog/rewritingreddit
> > http://masshightech.bizjournals.com/masshightech/stories/2006/07/03/story
> > 13.html
>
> I don't see where they claim that Python gave them a competitive
> advantage.
You don't think that having the system work without crashing provides a
competitive advantage?
> > And:
> >
> > http://xooglers.blogspot.com/2005/12/lets-get-real-database.html
> >
> > Finding documentation for the financial-success part of that second
> > example is left as an exercise.
>
> ...and where does this example refer to the competitive advantage of
> some language?
Please refer back to the question that you posed:
"For how many financial successes do you _actually_ know what technology
^^^^^^^^^^
they have used to implement them?"
I chose that particular example simply because it illustrates not only a
financial success for which the implementing technology is known, but
indeed that the success is directly attributable to certain features of
that technology. But even if you had asked about languages, it is also
documented elsewhere in that same blog that AdWords was written in Java.
It is debatable whether AdWords succeeded because it was written in Java
or despite that it was written in Java. But that it succeeded in part
because it used MySQL instead of Sybase is beyond dispute.
rg
From: Pascal Costanza
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m79esF4s69fU1@individual.net>
Ron Garret wrote:
> In article <··············@individual.net>,
> Pascal Costanza <··@p-cos.net> wrote:
>
>>> I did not ask for the particulars. I only asked for the particulars to
>>> be documentable. (Why do you have such a penchant for straw-man
>>> arguments?)
>> Sorry, I don't understand the difference.
>
> You don't understand the difference between making a claim for which
> supporting documentation exists and actually providing that
> documentation? And you're a candidate for a Ph.D.? Standards in German
> higher education must be slipping.
Yeah, fuck you too.
I don't understand what difference it makes to ask for particulars vs.
to ask for particulars to be documentable.
Or to illustrate it further: "Yeah, I know 50 financial successes
because of Lisp, and their particulars are documentable, but I don't
give you any details." Now what? Do you just increase your counter by 50?
>>>> and especially conduct a more serious
>>>> study than just randomly polling newsgroups.
>>> You have no way of knowing what other measures I may taking to obtain
>>> this information (to say nothing of the fact that my polling is far from
>>> random), but your defensiveness is to me an indication that the data I
>>> seek do not exist.
>> So where's the study where people can participate to give you the
>> information you need?
>
> Study? Who said anything about a study?
Me. What other means would you suggest to get some actually objective data?
There are pretty long lists of current Lisp programmers available. If
you could come up with a questionnaire, I think it would be possible to
conduct such a study. _That_ would be more constructive than all the
pure guesswork that's going on in this thread.
>>>> [1] Quiz question: For how many financial successes do you _actually_
>>>> know what technology they have used to implement them? Don't forget to
>>>> separate claims from actual facts - for example, see the bit about
>>>> Oracle in the entry "What database did you use?" at
>>>> http://www.paulgraham.com/vwfaq.html
>>> I can think of at least two right off the top of my head:
>>>
>>> http://www.aaronsw.com/weblog/rewritingreddit
>>> http://masshightech.bizjournals.com/masshightech/stories/2006/07/03/story
>>> 13.html
>> I don't see where they claim that Python gave them a competitive
>> advantage.
>
> You don't think that having the system work without crashing provides a
> competitive advantage?
I don't think that's related to the choice of language.
>>> And:
>>>
>>> http://xooglers.blogspot.com/2005/12/lets-get-real-database.html
>>>
>>> Finding documentation for the financial-success part of that second
>>> example is left as an exercise.
>> ...and where does this example refer to the competitive advantage of
>> some language?
>
> Please refer back to the question that you posed:
>
> "For how many financial successes do you _actually_ know what technology
> ^^^^^^^^^^
> they have used to implement them?"
>
> I chose that particular example simply because it illustrates not only a
> financial success for which the implementing technology is known, but
> indeed that the success is directly attributable to certain features of
> that technology. But even if you had asked about languages, it is also
> documented elsewhere in that same blog that AdWords was written in Java.
> It is debatable whether AdWords succeeded because it was written in Java
> or despite that it was written in Java. But that it succeeded in part
> because it used MySQL instead of Sybase is beyond dispute.
I was expecting you to be able to give an answer that contributes to the
discussion. Switching between MySQL and Sybase is just changing the
backend, I wouldn't consider that a change of technology in a sense
that's relevant for the discussion.
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/
In article <··············@individual.net>,
Pascal Costanza <··@p-cos.net> wrote:
> Ron Garret wrote:
> > In article <··············@individual.net>,
> > Pascal Costanza <··@p-cos.net> wrote:
> >
> >>> I did not ask for the particulars. I only asked for the particulars to
> >>> be documentable. (Why do you have such a penchant for straw-man
> >>> arguments?)
> >> Sorry, I don't understand the difference.
> >
> > You don't understand the difference between making a claim for which
> > supporting documentation exists and actually providing that
> > documentation? And you're a candidate for a Ph.D.? Standards in German
> > higher education must be slipping.
>
> Yeah, fuck you too.
I see that standards of civilized discourse are slipping too.
> I don't understand what difference it makes to ask for particulars vs.
> to ask for particulars to be documentable.
Have you ever read a scientific paper? Have you ever noticed the
section at the end entitled (typically) "References" which lists the
specific sources that contain the supporting evidence for certain claims
made in the paper? Do you understand that the claims made in the paper
and the references used to support those claims are separate and
distinct? Do you see that it is therefore possible to provide one
without providing the other?
> Or to illustrate it further: "Yeah, I know 50 financial successes
> because of Lisp, and their particulars are documentable, but I don't
> give you any details." Now what? Do you just increase your counter by 50?
No because the claim that you know 50 financial successes due to Lisp is
(I'll give you long odds) not documentable. (It is also false, though
that is a largely orthogonal issue. See below.)
> >>>> and especially conduct a more serious
> >>>> study than just randomly polling newsgroups.
> >>> You have no way of knowing what other measures I may taking to obtain
> >>> this information (to say nothing of the fact that my polling is far from
> >>> random), but your defensiveness is to me an indication that the data I
> >>> seek do not exist.
> >> So where's the study where people can participate to give you the
> >> information you need?
> >
> > Study? Who said anything about a study?
>
> Me. What other means would you suggest to get some actually objective data?
Objective data? Who said anything about objective data?
Good grief, I am beginning to think that you genuinely don't understand
this, and you're not just being intentionally obtuse. Consider:
1. Albert Einstein originated the theory of relativity.
This claim is documentable, yet I have provided no documentation for it.
2. The collapse of the World Trade Center was due to a U.S. government
conspiracy.
This claim is also documentable. (A claim does not have to be true to
be documentable.) I still have not provided any actual documentation.
But see for example:
http://reddit.com/goto?id=gklx
Now I have provided some (very minimal) documentation.
Do you see the difference?
Now contrast that with, e.g.:
3. Naughty Dog Software was sold to Sony for $100M.
That claim is not documentable, and it may or may not be true, because:
4. The terms of the ND acquisition are secret.
which is both true and documentable.
> >>>> [1] Quiz question: For how many financial successes do you _actually_
> >>>> know what technology they have used to implement them? Don't forget to
> >>>> separate claims from actual facts - for example, see the bit about
> >>>> Oracle in the entry "What database did you use?" at
> >>>> http://www.paulgraham.com/vwfaq.html
> >>> I can think of at least two right off the top of my head:
> >>>
> >>> http://www.aaronsw.com/weblog/rewritingreddit
> >>> http://masshightech.bizjournals.com/masshightech/stories/2006/07/03/story
> >>> 13.html
> >> I don't see where they claim that Python gave them a competitive
> >> advantage.
> >
> > You don't think that having the system work without crashing provides a
> > competitive advantage?
>
> I don't think that's related to the choice of language.
The Reddit people do, and their opinion in this matter carries more
weight than yours.
> >>> And:
> >>>
> >>> http://xooglers.blogspot.com/2005/12/lets-get-real-database.html
> >>>
> >>> Finding documentation for the financial-success part of that second
> >>> example is left as an exercise.
> >> ...and where does this example refer to the competitive advantage of
> >> some language?
> >
> > Please refer back to the question that you posed:
> >
> > "For how many financial successes do you _actually_ know what technology
> > ^^^^^^^^^^
> > they have used to implement them?"
> >
> > I chose that particular example simply because it illustrates not only a
> > financial success for which the implementing technology is known, but
> > indeed that the success is directly attributable to certain features of
> > that technology. But even if you had asked about languages, it is also
> > documented elsewhere in that same blog that AdWords was written in Java.
> > It is debatable whether AdWords succeeded because it was written in Java
> > or despite that it was written in Java. But that it succeeded in part
> > because it used MySQL instead of Sybase is beyond dispute.
>
> I was expecting you to be able to give an answer that contributes to the
> discussion. Switching between MySQL and Sybase is just changing the
> backend, I wouldn't consider that a change of technology in a sense
> that's relevant for the discussion.
But I do. And given that I am the one who posed the question on which
this discussion is based, my opinion in this matter carries more weight
than yours.
rg
From: Pascal Costanza
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m8h33F51n2bU1@individual.net>
Ron Garret wrote:
> Good grief, I am beginning to think that you genuinely don't understand
> this, and you're not just being intentionally obtuse.
I typically try to say what I mean, and I try to avoid rhetorical
tricks. My impression is that you are currently not interested in a
constructive discussion. That actually seems to be the case since quite
some time. The fact that you haven't reacted to a constructive
suggestion I have made in my previous posting, and instead quibble about
whose "opinion carries more weight" supports this impression. I find
this disappointing, especially because I know you can do better than that.
Maybe you should also think about what you have asked from others not so
long ago wrt condescending attitudes.
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/
In article <··············@individual.net>,
Pascal Costanza <··@p-cos.net> wrote:
> Ron Garret wrote:
>
> > Good grief, I am beginning to think that you genuinely don't understand
> > this, and you're not just being intentionally obtuse.
>
> I typically try to say what I mean, and I try to avoid rhetorical
> tricks. My impression is that you are currently not interested in a
> constructive discussion. That actually seems to be the case since quite
> some time. The fact that you haven't reacted to a constructive
> suggestion I have made in my previous posting,
Are you referring to this?
> Yeah, fuck you too.
Sorry, but I don't consider that to be particularly constructive.
(I find it ironic given all the trouble you had with You-Know-Who that
you are now adopting some of his rhetorical techniques.)
rg
From: Pascal Costanza
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m8mbtF505b3U2@individual.net>
Ron Garret wrote:
> In article <··············@individual.net>,
> Pascal Costanza <··@p-cos.net> wrote:
>
>> Ron Garret wrote:
>>
>>> Good grief, I am beginning to think that you genuinely don't understand
>>> this, and you're not just being intentionally obtuse.
>> I typically try to say what I mean, and I try to avoid rhetorical
>> tricks. My impression is that you are currently not interested in a
>> constructive discussion. That actually seems to be the case since quite
>> some time. The fact that you haven't reacted to a constructive
>> suggestion I have made in my previous posting,
>
> Are you referring to this?
>
>> Yeah, fuck you too.
Make a guess.
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/
Hilarious. Apply this bit of 'reasoning' to other endeavors.
"I want to grow oranges. Not too many mind you, just enough to get
by."
"Invest in my company. Our goal is to make less money than other
investments."
"Give me money so that I can do what I want to without regard to you"
-- This is what you would really be saying. How do you square this
with the categorical imperative?
Ron Garret wrote:
> In article <······················@parsec.no-spoon.de>,
> Stefan Scholl <······@no-spoon.de> wrote:
>
> > Is every Lisp developer infected by this PG virus? Success = $?
>
> No, but it is not an altogether unreasonable choice for a quality
> metric. If nothing else, it's easy to measure.
>
> > Why are so many Lisp people so obsessed about getting rich?
>
> I don't think they are. I think most Lisp people would be content just
> to break even (i.e. to be able to earn a living using Lisp). The
> problem is that it's hard to find investors who share that attitude.
>
> rg
I agree. I think that the poor, unemployed C++ programmer toiling away
into the night for something that is pure and true is an example we
should all strive to emulate.
Stefan Scholl wrote:
> Is every Lisp developer infected by this PG virus? Success = $?
>
> Why are so many Lisp people so obsessed about getting rich?
>
> --
> Web (en): http://www.no-spoon.de/ -*- Web (de): http://www.frell.de/
Stefan Scholl wrote:
> Is every Lisp developer infected by this PG virus? Success = $?
>
> Why are so many Lisp people so obsessed about getting rich?
>
> --
> Web (en): http://www.no-spoon.de/ -*- Web (de): http://www.frell.de/
I think that it is just people in general that are obsessed with
getting rich. Lisp people (or at least me) want to do it with Lisp.
From: Markus Grueneis
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4lu40bF3m382U1@individual.net>
Stefan Scholl schrieb:
> Is every Lisp developer infected by this PG virus? Success = $?
>
> Why are so many Lisp people so obsessed about getting rich?
>
Well, either you program for money (as probably most c.l.l readers do,
to pay their bills), or you do not program for money (as probably some
of the c.l.l. readers so, for the sake of whatever).
For the second part, we already know we are better then the rest of the
world, so we do not need to talk about that any further.
For the first part, success == $$ is inherently true. Here we see
possible improvements. As this seems to be the only not-yet-solved
aspect of Lisp programming, we *must* be obsessed about it.
'-)
-- Markus
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2slj9u20l.fsf@hermes.theari.com>
Stefan Scholl <······@no-spoon.de> writes:
> Is every Lisp developer infected by this PG virus? Success = $?
>
> Why are so many Lisp people so obsessed about getting rich?
I think that the "why hasn't Lisp made me rich yet?" mindset is not
evidence that so many Lisp people are obsessed with getting rich, but
rather is evidence that so many people obsessed with getting rich are
coming to Lisp because of the PG virus. I think it'd be more
interesting to see what percentage of this group are Lisp newbies and
if there are any experienced Lispers who sit around wondering why
they're not rich yet.
Ari Johnson schrieb:
> Stefan Scholl <······@no-spoon.de> writes:
>
...
>
> I think that the "why hasn't Lisp made me rich yet?" mindset is not
> evidence that so many Lisp people are obsessed with getting rich, but
> rather is evidence that so many people obsessed with getting rich are
> coming to Lisp because of the PG virus.
>
You are right, and this mindset is complete bull. The only way to get
rich is to find a market, eventually hire a bunch of cheap coders, and
let them use the language that fits best. Quite a few people got rich
off ABAP this way, or better so say, they got rich off the abap monkeys
labouring for them.
Economically, the alleged superiority of Lisp doesn't really matter.
-klaus
On 2006-09-03 11:08:07 +0100, "klaus" <··········@yahoo.de> said:
> You are right, and this mindset is complete bull. The only way to get
. ^^^^^^^^^^^^^
> rich is to find a market,
So many assumptions ...
To get rich, you can:
* rob a bank (there's risk, comparable to using a language no one
wants yoo to use :-)
* win the lottery
* inherit wealth
* send tons of emails posing as a retired Nigerian official
* become a crook
* invest other people money
* invest/trade your own money
* ...
Most people confuse work with getting money. You have to
chose. Either you work, OR you make money. You probably haven't
enough stamina to pursue both at the same time :)
--
JFB
On 2006-09-03 21:32:51 +0100, verec <·····@mac.com> said:
> * rob a bank (there's risk, comparable to using a language no one
> wants yoo to use :-)
Robing a bank is a pretty bad way of getting rich - they don't really
hold enough cash any more and the risks are absurd
> * win the lottery
Quite good, but low probability
> * inherit wealth
Also rotten, the tax men are too wise to this.
and so on. Best way I think short of the smart idea trick, is to play
the stockmarket, which over the long term does better than most.
From: Christophe Rhodes
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <sq3bb8krvu.fsf@cam.ac.uk>
Tim Bradshaw <···@tfeb.org> writes:
> and so on. Best way I think short of the smart idea trick, is to play
> the stockmarket, which over the long term does better than most.
Well. There are three ways of making money in the stock market: first
is to invest in it, and I'm afraid I think this is largely a mug's
game. A better way (probably what you mean by "play") is to look for
arbitrage opportunities; by far the best way of making money in the
markets, though, is to sell access to them and take a commission of
every trade (see "mug's game", above...). To drag this back towards a
semblance of topicality, Espen Vestre seems to have the right idea :-)
Christophe
Christophe Rhodes wrote:
> Well. There are three ways of making money in the stock market: first
> is to invest in it, and I'm afraid I think this is largely a mug's
> game.
What I specifically meant was that the stock market (in the form of
whatever broadly-based index you want) has been a pretty reasonable
investment over the long term - I think I remember reading that it had
averaged 11% APR over the last 100 years or something. That's much
better than you would have got by many other investments. For a
relatively low-risk strategy (higher risk than a bank unless you pick a
bad bank, but much lower risk than, say, investing in startups), and
one which requires no active involvement at all, it does rather well.
--tim
On 2006-09-04 09:06:14 -0400, "Tim Bradshaw" <··········@tfeb.org> said:
> What I specifically meant was that the stock market (in the form of
> whatever broadly-based index you want) has been a pretty reasonable
> investment over the long term - I think I remember reading that it had
> averaged 11% APR over the last 100 years or something.
Actually, real long term returns (inflation adjusted) are just as good
or better for bonds than Stocks[1] - a whopping 2.7%
1. http://www.internet2.edu/~shalunov/stock-market/
Raffael Cavallaro <················@pas-d'espam-s'il-vous-plait-mac.com>
wrote:
> On 2006-09-04 09:06:14 -0400, "Tim Bradshaw" <··········@tfeb.org> said:
>
> > What I specifically meant was that the stock market (in the form of
> > whatever broadly-based index you want) has been a pretty reasonable
> > investment over the long term - I think I remember reading that it had
> > averaged 11% APR over the last 100 years or something.
>
> Actually, real long term returns (inflation adjusted) are just as good
> or better for bonds than Stocks[1] - a whopping 2.7%
> 1. http://www.internet2.edu/~shalunov/stock-market/
It would be nice if he explained exactly what he means by "adjusted for
dividends" in the methodology section. Or perhaps something about why
his calculation is so radically different from every other one I've
seen. I know survivor bias is significant when tracking back funds
indexes that exist today -- has he done something to eliminate that?
Not clear at all.
The related link at the bottom of the page, indicated that value of
dividends received by holders of the stocks was not considered at all,
and it came up with very similar looking numbers. But dividends were
the source of much of the profit for stockholders over the period
1920-198x. The stock market as a whole paying very low dividend rates
is a relatively recent phenomenon.
I have no confidence whatsover in this guy's conclusions without better
information. It looks like he's ignoring dividends, which would lead
him *far* astray.
Michael
Raffael Cavallaro wrote:
> Actually, real long term returns (inflation adjusted) are just as good
> or better for bonds than Stocks[1] - a whopping 2.7%
>
> 1. http://www.internet2.edu/~shalunov/stock-market/
I think I'll take The Economist's word for it, not some random on the
intarweb, thanks.
--tim
Madhu wrote:
> Helu
> * "Tim Bradshaw" <························@i42g2000cwa.XXXXXXXXXXXX.com> :
>
> > Raffael Cavallaro wrote:
> >
> >> Actually, real long term returns (inflation adjusted) are just as good
> >> or better for bonds than Stocks[1] - a whopping 2.7%
> >>
> >> 1. http://www.internet2.edu/~shalunov/stock-market/
> >
> > I think I'll take The Economist's word for it, not some random on the
> > intarweb, thanks.
>
> I've only read c.l.l on this, what Raffael Cavallaro quoted when he
> first posted this, which might (or not) also explain the Economist's
> position.
>
> |From: Raffael Cavallaro
> |Message-ID: <····································@pasdespamsilvousplaitmaccom|
> [..]
>
> | For example, I'm sure everyone here has seen the rather misleading
> | graphs that purport to show that if you had invested $1000.00 in the
> | "stock market" in 1929 - right before the crash - you'd still be a
> | multimillionaire today. What they don't tell you is that in real dollar
> | terms you'd still only have broken even in the late 1980s - i.e. no
> | long term net gain for 60 years - and that's *before taxes*:
> | <http://www.internet2.edu/~shalunov/stock-market/|
>
> | It turns out that the real return from the DJIA from 1913-2002 is just
> | 2.7% (just 1.3% from 1913-1991 if you want to exclude the tech bubble
> | and crash) - comparable to treasury bonds. And this isn't even the
> | broad market which includes many companies which fail outright and are
> | de-listed (i.e., investment goes to $0.00). Such companies are removed
> | from the DJIA and are not reflected by the relatively rosy picture
> | painted by the DJIA. The losers of major indices are replaced by
> | up-trending issues before the failing companies go into long decline or
> | go belly up entirely.
This last part is very relevant. If the company that fails is frozen
on the stockmarket as they are if they are handed to recievers then you
cannot sell them. You recoup your percentage of whatever the recievers
recoup. An index such as the Dow does not reflect this, the share just
disappears one day and another replaces it. This is one reason why
index trackers are always below the index.
Another issue is dividends and trading charges. Over most periods of
time dividends recieved would pay for charges incurred in dealing in
order to track an index. But there are probably periods of time where
this isn't possible.
Madhu wrote:
> I've only read c.l.l on this, what Raffael Cavallaro quoted when he
> first posted this, which might (or not) also explain the Economist's
> position.
It probably does: in particular the `only broken even till late 80s'
which, very conveniently, is quite close to the time when dividends
became much less common, which was a tech-bubble innovation (basically
coming down to: if your company isn't actually making any money, then
you *can't* pay dividends, I suspect). So sure, you only broke even if
you ignore the main way you made money - I lose a huge amount each year
if I ignore my salary, too.
> | Nevertheless we are given to believe that you can only win by investing
> | in the stock market and that if you lose money they *you* are a loser.
And I certainly have never claimed this: there is obviously risk.
There are also obviously ways of making a lot more money, but they
generally have higher risk. The real issue is that the stockmarket (in
the form of some fund which tracks a broadly based index, there being
many such funds) is not such a bad bet in the long term compared, say,
to interest-bearing accounts & blah.
--tim
From: Espen Vestre
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m18xl01375.fsf@doduo.netfonds.no>
Christophe Rhodes <·····@cam.ac.uk> writes:
> every trade (see "mug's game", above...). To drag this back towards a
> semblance of topicality, Espen Vestre seems to have the right idea :-)
:-) (But don't give me too much credit, I joined this company 5 years
after its startup in 1997)
--
(espen)
Ari Johnson <·········@gmail.com> writes:
> Stefan Scholl <······@no-spoon.de> writes:
>
>> Is every Lisp developer infected by this PG virus? Success = $?
>>
>> Why are so many Lisp people so obsessed about getting rich?
>
> I think that the "why hasn't Lisp made me rich yet?" mindset is not
> evidence that so many Lisp people are obsessed with getting rich, but
> rather is evidence that so many people obsessed with getting rich are
> coming to Lisp because of the PG virus. I think it'd be more
> interesting to see what percentage of this group are Lisp newbies and
> if there are any experienced Lispers who sit around wondering why
> they're not rich yet.
Travaillez, prenez de la peine :
C'est le fonds qui manque le moins.
Un riche Laboureur, sentant sa mort prochaine,
Fit venir ses enfants, leur parla sans t�moins.
Gardez-vous, leur dit-il, de vendre l'h�ritage
Que nous ont laiss� nos parents.
Un tr�sor est cach� dedans.
Je ne sais pas l'endroit ; mais un peu de courage
Vous le fera trouver, vous en viendrez � bout.
Remuez votre champ d�s qu'on aura fait l'O�t.
Creusez, fouiller, b�chez ; ne laissez nulle place
O� la main ne passe et repasse.
Le p�re mort, les fils vous retournent le champ
De��, del�, partout ; si bien qu'au bout de l'an
Il en rapporta davantage.
D'argent, point de cach�. Mais le p�re fut sage
De leur montrer avant sa mort
Que le travail est un tr�sor.
--
__Pascal Bourguignon__ http://www.informatimago.com/
ATTENTION: Despite any other listing of product contents found
herein, the consumer is advised that, in actuality, this product
consists of 99.9999999999% empty space.
On 2006-09-03 05:52:35 +0100, Pascal Bourguignon <···@informatimago.com> said:
> Ari Johnson <·········@gmail.com> writes:
>
>> Stefan Scholl <······@no-spoon.de> writes:
>>
>>> Is every Lisp developer infected by this PG virus? Success = $?
>>>
>>> Why are so many Lisp people so obsessed about getting rich?
>>
>> I think that the "why hasn't Lisp made me rich yet?" mindset is not
>> evidence that so many Lisp people are obsessed with getting rich, but
>> rather is evidence that so many people obsessed with getting rich are
>> coming to Lisp because of the PG virus. I think it'd be more
>> interesting to see what percentage of this group are Lisp newbies and
>> if there are any experienced Lispers who sit around wondering why
>> they're not rich yet.
>
> Travaillez, prenez de la peine :
> C'est le fonds qui manque le moins.
> Un riche Laboureur, sentant sa mort prochaine,
> Fit venir ses enfants, leur parla sans témoins.
> Gardez-vous, leur dit-il, de vendre l'héritage
> Que nous ont laissé nos parents.
> Un trésor est caché dedans.
> Je ne sais pas l'endroit ; mais un peu de courage
> Vous le fera trouver, vous en viendrez à bout.
> Remuez votre champ dès qu'on aura fait l'Oût.
> Creusez, fouiller, bêchez ; ne laissez nulle place
> Où la main ne passe et repasse.
> Le père mort, les fils vous retournent le champ
> Deçà, delà, partout ; si bien qu'au bout de l'an
> Il en rapporta davantage.
> D'argent, point de caché. Mais le père fut sage
> De leur montrer avant sa mort
> Que le travail est un trésor.
Entre Franchouillards (ou francophones):
"Le travail, c'est la santé, ne rien faire, c'est la conserver!"
:-)
--
JFB
Stefan Scholl <······@no-spoon.de> writes:
> Is every Lisp developer infected by this PG virus? Success = $?
>
> Why are so many Lisp people so obsessed about getting rich?
To be able to buy periodically new computers and new hard disks (how
brittle they make them nowadays).
--
__Pascal Bourguignon__ http://www.informatimago.com/
CONSUMER NOTICE: Because of the "uncertainty principle," it is
impossible for the consumer to simultaneously know both the precise
location and velocity of this product.
Stefan Scholl wrote:
> [...]
> Why are so many Lisp people so obsessed about getting rich?
Because, Lisp people want to have the time and resources to create want
_they want_ to create, not what someone is willing to hire them to
create.
It really makes perfect sense.
···············@gmail.com wrote:
> Stefan Scholl wrote:
> > [...]
> > Why are so many Lisp people so obsessed about getting rich?
>
> Because, Lisp people want to have the time and resources to create want
> _they want_ to create, not what someone is willing to hire them to
> create.
>
> It really makes perfect sense.
Yes. And I've front-loaded my personal demands, to a fault. By
pursuing open source and advanced programming languages over the past 2
years, I've found elaborate ways to make no money whatsoever. I've run
out of rope. The business model doesn't work; there is none. And now
I'm facing some very ugly realities about what I'm skilled at vs. what
the world readily pays for.
The only upside is I cannot be prevented from succeeding at my long
term goal, publishing my own games. But it is proving brutal to get
there.
Cheers,
Brandon Van Every
Mallor wrote:
> ···············@gmail.com wrote:
> > Stefan Scholl wrote:
> > > [...]
> > > Why are so many Lisp people so obsessed about getting rich?
> >
> > Because, Lisp people want to have the time and resources to create want
> > _they want_ to create, not what someone is willing to hire them to
> > create.
> >
> > It really makes perfect sense.
>
> Yes. And I've front-loaded my personal demands, to a fault. By
> pursuing open source and advanced programming languages over the past 2
> years, I've found elaborate ways to make no money whatsoever. I've run
> out of rope. The business model doesn't work; there is none. And now
> I'm facing some very ugly realities about what I'm skilled at vs. what
> the world readily pays for.
>
> The only upside is I cannot be prevented from succeeding at my long
> term goal, publishing my own games. But it is proving brutal to get
> there.
You sound waay too depressing, and thats probably because you're
completely missing the point.
For instance, the success of Naughty Dog has less to do with lisp, and
more to do with their creative genius in making characters like Crash
and Jak. Their success is *not* because of lisp, but due to their
brilliant designers. Ofcourse lisp can lend a helping hand, but its no
substitute for a poor product.
So stop obsessing about lisp, and /first/ think of a good product, and
then you can think about how lisp can help you implement it!
sheeesh...
[sreeram;]
K.S.Sreeram wrote:
>
> You sound waay too depressing, and thats probably because you're
> completely missing the point.
>
> For instance, the success of Naughty Dog has less to do with lisp, and
> more to do with their creative genius in making characters like Crash
> and Jak. Their success is *not* because of lisp, but due to their
> brilliant designers. Ofcourse lisp can lend a helping hand, but its no
> substitute for a poor product.
>
> So stop obsessing about lisp, and /first/ think of a good product, and
> then you can think about how lisp can help you implement it!
No, *WE AGREE* on the point. The point is that Naughty Dog is not an
example of Lisp success. It's an example of a success where Lisp was
along for the ride. Recognizing this, Sony acquired Naughty Dog and
blew off Lisp. Naughty Dog has done nothing to spread high level
languages in the game industry. People still use C++ for the heavy
lifting, and more popular scripting languages like Python, Ruby, and
Lua when flexibility is required but performance isn't.
The best that can be said about Naughty Dog is Lisp didn't *harm* their
development. They did a postmortem where they said GOAL was strictly
50/50 benefit/liability. This would have been true of any custom
scripting language. Postmortems over the years are full of tales of
guys who got paid to write their own silly language, usually to the
detriment of the product's production pipeline. But, people like to
get paid to do things that benefit themselves. It's called careerism.
Cheers,
Brandon Van Every
"K.S.Sreeram" <·········@gmail.com> writes:
> Mallor wrote:
>> ···············@gmail.com wrote:
>> > Stefan Scholl wrote:
>> > > [...]
>> > > Why are so many Lisp people so obsessed about getting rich?
>> >
>> > Because, Lisp people want to have the time and resources to create want
>> > _they want_ to create, not what someone is willing to hire them to
>> > create.
>> >
>> > It really makes perfect sense.
>>
>> Yes. And I've front-loaded my personal demands, to a fault. By
>> pursuing open source and advanced programming languages over the past 2
>> years, I've found elaborate ways to make no money whatsoever. I've run
>> out of rope. The business model doesn't work; there is none. And now
>> I'm facing some very ugly realities about what I'm skilled at vs. what
>> the world readily pays for.
>>
>> The only upside is I cannot be prevented from succeeding at my long
>> term goal, publishing my own games. But it is proving brutal to get
>> there.
>
> You sound waay too depressing, and thats probably because you're
> completely missing the point.
>
> For instance, the success of Naughty Dog has less to do with lisp, and
> more to do with their creative genius in making characters like Crash
> and Jak. Their success is *not* because of lisp, but due to their
> brilliant designers. Ofcourse lisp can lend a helping hand, but its no
> substitute for a poor product.
>
> So stop obsessing about lisp, and /first/ think of a good product, and
> then you can think about how lisp can help you implement it!
>
> sheeesh...
> [sreeram;]
>
I agree. I've seen quite a few posts recently where people blame open
source, the lack of acceptance for lisp, jobs going off-shore and sun
spots for why they haven't become rich.
Putting aside the pointless obsession with wealth, it seems many over
look the importance of a good idea being developed by someone with
creativity and imagination. While a good idea is no guarantee of
success, you are far less likely to succeed without one, regardless of
implementation platform, marketing strategy and industry trends.
Tim
--
tcross (at) rapttech dot com dot au
Stefan Scholl <······@no-spoon.de> writes:
> Why are so many Lisp people so obsessed about getting rich?
We often say that Lisp is so much more productive than other
languages. In a commercial context, where we're getting paid for
developing products or enabling services, this should translate into
significantly greater returns than on systems developed with more
common languages.
PG's ViaWeb story is a good example of what should be generally taking
place among Lispers if we really are that much better.
The fact is that even using computers is about economics, speeding
things up, doing things more cheaply than by hand or with some other
machines.
--
Matt Curtin, author of Brute Force: Cracking the Data Encryption Standard
Founder of Interhack Corporation +1 614 545 4225 http://web.interhack.com/
From: Patrick May
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2y7t086j4.fsf@Dagney.local>
Matt Curtin <········@interhack.net> writes:
> We often say that Lisp is so much more productive than other
> languages. In a commercial context, where we're getting paid for
> developing products or enabling services, this should translate into
> significantly greater returns than on systems developed with more
> common languages.
The actual coding and unit testing of a system is only a small
percentage of the total effort required to put it into production.
Even if Lisp allowed us to reduce the development time by a factor of
ten to one hundred, requirements gathering, analysis, design,
acceptance testing, regression testing, deployment configuration,
operations support, and numerous other tasks make up a far larger
resource requirement. There is therefore only a small increase in ROI
for using even the most powerful language.
Focusing on how Lisp reduces the time and resources required to
perform those other tasks would be interesting, as would providing
support for the idea that Lisp-based systems are easier to extend. I
suspect both are valid claims.
Regards,
Patrick
------------------------------------------------------------------------
S P Engineering, Inc. | Large scale, mission-critical, distributed OO
| systems design and implementation.
···@spe.com | (C++, Java, Common Lisp, Jini, middleware, SOA)
Patrick May wrote:
> Matt Curtin <········@interhack.net> writes:
> > We often say that Lisp is so much more productive than other
> > languages. In a commercial context, where we're getting paid for
> > developing products or enabling services, this should translate into
> > significantly greater returns than on systems developed with more
> > common languages.
>
> The actual coding and unit testing of a system is only a small
> percentage of the total effort required to put it into production.
> Even if Lisp allowed us to reduce the development time by a factor of
> ten to one hundred, requirements gathering, analysis, design,
> acceptance testing, regression testing, deployment configuration,
> operations support, and numerous other tasks make up a far larger
> resource requirement. There is therefore only a small increase in ROI
> for using even the most powerful language.
This depends on the program though, sometimes it's simple. Sometimes
as simple as someone pointing to another program and saying "I'll have
one exactly like that"! though that is a rather extreme case.
Matt Curtin <········@interhack.net> writes:
> Stefan Scholl <······@no-spoon.de> writes:
>
>> Why are so many Lisp people so obsessed about getting rich?
>
> We often say that Lisp is so much more productive than other
> languages. In a commercial context, where we're getting paid for
> developing products or enabling services, this should translate into
> significantly greater returns than on systems developed with more
> common languages.
>
> PG's ViaWeb story is a good example of what should be generally taking
> place among Lispers if we really are that much better.
>
> The fact is that even using computers is about economics, speeding
> things up, doing things more cheaply than by hand or with some other
> machines.
>
Although I don't disagree, I think the often overlooked critical
ingrediant is actually having the right idea - the implementation
language can make it easier to get that idea up and profitable, but
you have to have the idea.
Putting it another way, I suspect PG would have been successful with
viaweb even if he had used another implementation language. certainly,
he might have had to do some things a little differently and maybe it
would have been a bit harder to have the flexibility he feels was a
big part of the secret to the success, but I believe a talented
development team with a good idea would probably have pulled it off
using another language just as successfully.
Another factor which I think is relevant is that lisp, while one of
the easiest languages I've tried to learn, is also one of the hardest
languages I've tried to master (and one of the most enjoyable because
I seem to learn something new every day). Of course, this progress is
artificially delayed by the fact I can only practice at home in my own
time - unlike many other languages I've learnt, where I was paid and
able to spend full-time getting to know the language.
finally, although many get upset at the following claim, I still think
it holds - most of the currently popular languages have lots of
libraries available for doing what is common in many applications
these days, making it fast to develop at least the prototype. CL is
lacking in some of these - while it is true that for most things you
can find a CL library if you search enough, you have to do the
searching and often you have to evaluate different solutions to find
the one you want. for example, parsing XML, something which is often
remarked upon as being natural for lisp, but there is no standard
library for it like there is in perl, java or python. Worse yet, when
you go searching, you find quite a few and have to spend some time
working out which one has the right balance of features and complexity
for what you need. then you try it out and find some minor issues and
have to go back and reevaluate.
Some will argue the solution is to use one of the commercial
implementations and maybe they are right. However, compared to perl,
java and python, this is likely to be quite expensive (and since you
are unlikely to get buckets of venture capital to use because you are
implementing using lisp, money is an issue right now).
Tim
--
tcross (at) rapttech dot com dot au
From: Ken Tilton
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <cjVKg.1$Bu3.0@newsfe08.lga>
Tim X wrote:
> Matt Curtin <········@interhack.net> writes:
>
>
>>Stefan Scholl <······@no-spoon.de> writes:
>>
>>
>>>Why are so many Lisp people so obsessed about getting rich?
>>
>>We often say that Lisp is so much more productive than other
>>languages. In a commercial context, where we're getting paid for
>>developing products or enabling services, this should translate into
>>significantly greater returns than on systems developed with more
>>common languages.
>>
>>PG's ViaWeb story is a good example of what should be generally taking
>>place among Lispers if we really are that much better.
>>
>>The fact is that even using computers is about economics, speeding
>>things up, doing things more cheaply than by hand or with some other
>>machines.
>>
>
>
> Although I don't disagree, I think the often overlooked critical
> ingrediant is actually having the right idea - the implementation
> language can make it easier to get that idea up and profitable, but
> you have to have the idea.
>
> Putting it another way, I suspect PG would have been successful with
> viaweb even if he had used another implementation language. certainly,
> he might have had to do some things a little differently and maybe it
> would have been a bit harder to have the flexibility he feels was a
> big part of the secret to the success, but I believe a talented
> development team with a good idea would probably have pulled it off
> using another language just as successfully.
You have forgotten the main point of PG's write-up on ViaWeb: yes, you
are right (and this thread is forgetting) that the "good idea" was
necessary. But the whole point of what PG wrote is that more than a few
others had had the same good idea. He extolled Lisp for the competitive
edge it gave them.
>
> Another factor which I think is relevant is that lisp, while one of
> the easiest languages I've tried to learn, is also one of the hardest
> languages I've tried to master (and one of the most enjoyable because
> I seem to learn something new every day). Of course, this progress is
> artificially delayed by the fact I can only practice at home in my own
> time - unlike many other languages I've learnt, where I was paid and
> able to spend full-time getting to know the language.
I agree about the mastery vs learn diff, but not the implication that
this hurts someone using Lisp. I was more productive immediately with
Lisp because just writing C in Lisp (which one can do in a few days) is
a big win over writing C in C.
Mind you, we see reports here from people who have read three books on
lisp and want to know what other three they should read before writing
any code. Uh, yeah, that approach could slow one down.
>
> finally, although many get upset at the following claim, I still think
> it holds - most of the currently popular languages have lots of
> libraries available for doing what is common in many applications
> these days, making it fast to develop at least the prototype. CL is
> lacking in some of these - while it is true that for most things you
> can find a CL library if you search enough, you have to do the
> searching and often you have to evaluate different solutions to find
> the one you want. for example, parsing XML, something which is often
> remarked upon as being natural for lisp, but there is no standard
> library for it like there is in perl, java or python. Worse yet, when
> you go searching, you find quite a few and have to spend some time
> working out which one has the right balance of features and complexity
> for what you need. then you try it out and find some minor issues and
> have to go back and reevaluate.
The above matters IFF one is not adding much value in one's own code.
That may be the case for a lot of Good Ideas, but then with any success
one can expect a whole lot of competition in short order.
>
> Some will argue the solution is to use one of the commercial
> implementations and maybe they are right. However, compared to perl,
> java and python, this is likely to be quite expensive (and since you
> are unlikely to get buckets of venture capital to use because you are
> implementing using lisp, money is an issue right now).
Buckets of venture capital to buy LW Pro? That simply makes no sense.
You are confusing the silly hobbyists who hang around c.l.l with a
serious entrepreneur, who will take a second mortgage, borrow from
family, use $$$ from a severance package, sell some stuff on eBay, etc
etc to get reasonably small amounts of initial capital.
The big problem is the subject header, which confuses financial success
with language quality. As a measure of that silliness, it would be fun
to turn the question around: Any XXX financial success stories such as
ViaWeb in which the entrepreneur (to any interesting degree) credits the
language XXX with their success?
One could argue, ah, but proponents of those languages are not claiming
those languages are especially good!
To which I would respond, "Exactly."
kt
--
Cells: http://common-lisp.net/project/cells/
"I'll say I'm losing my grip, and it feels terrific."
-- Smiling husband to scowling wife, New Yorker cartoon
Ken Tilton wrote:
>
> You have forgotten the main point of PG's write-up on ViaWeb: yes, you
> are right (and this thread is forgetting) that the "good idea" was
> necessary. But the whole point of what PG wrote is that more than a few
> others had had the same good idea. He extolled Lisp for the competitive
> edge it gave them.
Yes! That was my point in some other thread: Lisp buys you about six
months in the right circumstances. But about six months is what you
need.
--tim
Ken Tilton ha escrito:
> Buckets of venture capital to buy LW Pro? That simply makes no sense.
> You are confusing the silly hobbyists who hang around c.l.l with a
> serious entrepreneur, who will take a second mortgage, borrow from
> family, use $$$ from a severance package, sell some stuff on eBay, etc
> etc to get reasonably small amounts of initial capital.
A "serious entrepreneur" will exactly think about that, and if he is
serious enoght will avoid as much as possible to borrow money from the
family, use $$$ from a severance package, sell some stuff on eBay, and
take a second mortage.
Ah! And I am pretty sure that there are people here hanging aroung
c.l.l which already does what you claim that only "seious" people do,
so your claim is nonsense (if you allow me to say you).
From: Andreas
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m2v3dF45m95U1@uni-berlin.de>
Javier wrote:
> Ken Tilton ha escrito:
>
>>Buckets of venture capital to buy LW Pro? That simply makes no sense.
>>You are confusing the silly hobbyists who hang around c.l.l with a
>>serious entrepreneur, who will take a second mortgage, borrow from
>>family, use $$$ from a severance package, sell some stuff on eBay, etc
>>etc to get reasonably small amounts of initial capital.
>
>
> A "serious entrepreneur" will exactly think about that, and if he is
> serious enoght will avoid as much as possible to borrow money from the
> family, use $$$ from a severance package, sell some stuff on eBay, and
> take a second mortage.
> Ah! And I am pretty sure that there are people here hanging aroung
> c.l.l which already does what you claim that only "seious" people do,
> so your claim is nonsense (if you allow me to say you).
>
Sorry Javier but Kenny is right.
If you want to create a bussiness and $1100 for LWPro becomes a problem
then there is something very wrong with you bussiness plan.
Regards
AHz
Andreas ha escrito:
> Sorry Javier but Kenny is right.
> If you want to create a bussiness and $1100 for LWPro becomes a problem
> then there is something very wrong with you bussiness plan.
If you are going to make a bussiness plan, you are going to include a
lot of things which will cost you a lot more. It is the sum of every
one which might get it impossible to reach, so if you are serious you
are going to consider buying or using free solutions of every part.
And it is not $1100 for LW, it is $3900 if you need the enterprise, and
it is *per seat*, so if you need to hire somebody to help you (and you
may), add a lot of money for the total.
Also add to the total that if your program needs to run on several
plattforms, you also need to pay for all of them, per seat.
If you are not going to form a society, then you are limited to write
only very simple applications, because you've got not the time for
doing everything.
So, or you've got a lot of money for starting your project, or you need
to use open source solutions.
Or you may go to Python, Perl, Java, or Ruby, which aren't really so
bad, are open source, and are complete and ready to use out-of-the-box.
;) This is actualy what most people do.
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2slj74lwp.fsf@hermes.theari.com>
"Javier" <·······@gmail.com> writes:
> If you are going to make a bussiness plan, you are going to include a
> lot of things which will cost you a lot more. It is the sum of every
> one which might get it impossible to reach, so if you are serious you
> are going to consider buying or using free solutions of every part.
> And it is not $1100 for LW, it is $3900 if you need the enterprise, and
> it is *per seat*, so if you need to hire somebody to help you (and you
> may), add a lot of money for the total.
If you need to hire someone and it breaks the bank to spend $3,900 on
his tools, I would hate to know what kind of wages you are paying him.
The point is this: the topic of this thread refers to "financial
success stories." Someone who can't beat $3,900 of gross revenues is
not a financial success story.
And if you really do need to cut corners at the outset, you can more
than make do with the free alternatives. What you have to ask
yourself is this very simple question of economics, which demands a
rational answer: Does the developer time (including opportunity cost
of delayed access to the market) spent in getting your development
tools up to the point that they would be if you had purchased
LispWorks for $3,900 per platform-seat cost more or less than $3,900
per developer-seat? That is the only question you need to answer to
decide which tool to use.
> Also add to the total that if your program needs to run on several
> plattforms, you also need to pay for all of them, per seat.
> If you are not going to form a society, then you are limited to write
> only very simple applications, because you've got not the time for
> doing everything.
If you sell even 100 copies per platform per developer, at $39 per
copy, then you have broken even on your development tools. If you
aren't selling at least that many copies, then you're not a success
story.
From: Andreas
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m33c8F3t6ciU1@uni-berlin.de>
Javier wrote:
> Andreas ha escrito:
>
>
>>Sorry Javier but Kenny is right.
>>If you want to create a bussiness and $1100 for LWPro becomes a problem
>>then there is something very wrong with you bussiness plan.
>
>
> If you are going to make a bussiness plan, you are going to include a
> lot of things which will cost you a lot more. It is the sum of every
> one which might get it impossible to reach, so if you are serious you
> are going to consider buying or using free solutions of every part.
> And it is not $1100 for LW, it is $3900 if you need the enterprise, and
> it is *per seat*, so if you need to hire somebody to help you (and you
> may), add a lot of money for the total.
> Also add to the total that if your program needs to run on several
> plattforms, you also need to pay for all of them, per seat.
> If you are not going to form a society, then you are limited to write
> only very simple applications, because you've got not the time for
> doing everything.
> So, or you've got a lot of money for starting your project, or you need
> to use open source solutions.
> Or you may go to Python, Perl, Java, or Ruby, which aren't really so
> bad, are open source, and are complete and ready to use out-of-the-box.
> ;) This is actualy what most people do.
>
cute :-) Some rough (and somewhat old) numbers:
A software developer here (germany) is calculated with around
�100k/year. This includes salary, employer's contribution and rents
but no license costs.
This rough number may change, dependend on the region, by some 10%.
He is used to work on around 200days/year (due to weekends, vacation,
illness etc).
So we end up with costs of �500/working day (!).
And that's what you need to state in your bussiness plan (excepting
you create a one-person-working-at-home-as-consulting company. Thats
a different story. And you will often have no bussiness plan at all).
Finally lets assume 3 developer (300k) and one sales person (100k too -
because im too lazy to look it up).
Then we are talking about 400k/year and you still have no secretary
for the telephone nor support ppl or anything else.
And you really care about $1100/seat for the software ?
If you can speed up development by only 3 days because they uses LWPro
then you have earned �400 (3*�500-�1100) per seat.
I think that was it what Kenny wants you to say.
Regards
AHz
Andreas ha escrito:
> A software developer here (germany) is calculated with around
> 100k/year. This includes salary, employer's contribution and rents
> but no license costs.
Oh my god!!! Please, look for a job in Germany for me, here in Spain we
earn 15.000/year!! (And only after working for 800/month for
several years until you've got enought experience!!!)
So you're right. In Germany it is possible to buy it. In Spain, it's
almost impossible if you are a normal guy (and I'm pretty sure that
this is the same in almost any country).
So now I understand why there are so much open source developers in
Spain, and why you paid us so much to enter in the European Union.
Thanks!! but I feel that we need much more time to get your level of
earnings!.
Unfortunaly, LW cost the same for us than for you. That's life.
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2veo3blwb.fsf@hermes.theari.com>
"Javier" <·······@gmail.com> writes:
> Andreas ha escrito:
>> A software developer here (germany) is calculated with around
>> 100k/year. This includes salary, employer's contribution and rents
>> but no license costs.
>
> Oh my god!!! Please, look for a job in Germany for me, here in Spain we
> earn 15.000/year!! (And only after working for 800/month for
> several years until you've got enought experience!!!)
> So you're right. In Germany it is possible to buy it. In Spain, it's
> almost impossible if you are a normal guy (and I'm pretty sure that
> this is the same in almost any country).
That's not what he meant. He meant that the cost of a developer is
that amount, not the salary. There are three general categories of
expenses in having an employee (salary/wages, benefits such as group
health insurance, and tools/workspace), and salary is generally the
smallest factor. For the 15,000 euros you are paid in a year, how
much more do you think it costs your employer to provide you with a
desk and either an office or cubicle, plus a computer and electricity
to run it, plus lavatory facilities, plus a parking space, plus
lighting, plus air conditioning, etc.? It is almost certainly more
than 15,000 euros.
Ari Johnson ha escrito:
> That's not what he meant. He meant that the cost of a developer is
> that amount, not the salary. There are three general categories of
> expenses in having an employee (salary/wages, benefits such as group
> health insurance, and tools/workspace), and salary is generally the
> smallest factor. For the 15,000 euros you are paid in a year, how
> much more do you think it costs your employer to provide you with a
> desk and either an office or cubicle, plus a computer and electricity
> to run it, plus lavatory facilities, plus a parking space, plus
> lighting, plus air conditioning, etc.? It is almost certainly more
> than 15,000 euros.
Sure, but I think that 100K is too far from 15k (almost 7x more),
and don't think that having to pay everything you say will sum so much!
At the most, perhaps 30k for me. Here we pay 200 for insurance,
and 10% for wages every month. The rest are colateral costs, which are
shared from all other employers in the same office, so this might be
not so much in comparation.
And the worst thing is that we ussually work for 12h every day and some
saturdays, and this is considered normal (althought not legal). They
say, more or less, "if you don't agree, you are invited to go".
And prices are almost at the same level as in Germany. So imagine how
bad we reach the end of the month. Now, imagne too that I have decided
to mount my own bussiness. With that salary, it would take months only
for buying a single LW seat, which is not admisible. So the only way
for me would be to use open source implementations.
And don't think that my situation is exceptional. In Spain and a lot of
countries of the EU are in the same situation. I can't even imagine how
bad there could be in Africa, Asia, South America, if they earn even
much much less than us.
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2r6yrbkni.fsf@hermes.theari.com>
"Javier" <·······@gmail.com> writes:
> Ari Johnson ha escrito:
>
>> That's not what he meant. He meant that the cost of a developer is
>> that amount, not the salary. There are three general categories of
>> expenses in having an employee (salary/wages, benefits such as group
>> health insurance, and tools/workspace), and salary is generally the
>> smallest factor. For the 15,000 euros you are paid in a year, how
>> much more do you think it costs your employer to provide you with a
>> desk and either an office or cubicle, plus a computer and electricity
>> to run it, plus lavatory facilities, plus a parking space, plus
>> lighting, plus air conditioning, etc.? It is almost certainly more
>> than 15,000 euros.
>
> Sure, but I think that 100K is too far from 15k (almost 7x more),
> and don't think that having to pay everything you say will sum so much!
> At the most, perhaps 30k for me. Here we pay 200 for insurance,
> and 10% for wages every month. The rest are colateral costs, which are
> shared from all other employers in the same office, so this might be
> not so much in comparation.
> And the worst thing is that we ussually work for 12h every day and some
> saturdays, and this is considered normal (althought not legal). They
> say, more or less, "if you don't agree, you are invited to go".
> And prices are almost at the same level as in Germany. So imagine how
> bad we reach the end of the month. Now, imagne too that I have decided
> to mount my own bussiness. With that salary, it would take months only
> for buying a single LW seat, which is not admisible. So the only way
> for me would be to use open source implementations.
> And don't think that my situation is exceptional. In Spain and a lot of
> countries of the EU are in the same situation. I can't even imagine how
> bad there could be in Africa, Asia, South America, if they earn even
> much much less than us.
You are writing under the assumption that people become financial
successes by working for a salary and then spending their salary on
development tools. This is not the case.
"Javier" <·······@gmail.com> writes:
> Sure, but I think that <somecost> is too far from <somecost> (almost
> 7x more), and don't think that having to pay everything you say will
> sum so much! ...etc...
Don't take this the wrong way, but it is painfully clear you hve never
been in business...
/Jon
--
'j' - a n t h o n y at romeo/charley/november com
jayessay ha escrito:
> "Javier" <·······@gmail.com> writes:
>
>
> > Sure, but I think that <somecost> is too far from <somecost> (almost
> > 7x more), and don't think that having to pay everything you say will
> > sum so much! ...etc...
>
> Don't take this the wrong way, but it is painfully clear you hve never
> been in business...
I've been not in software business, but in other areas and not of a
such technical level.
Elsewhere, I still think that if you are a freelance worker, or even
associating to another person, $7800 (for 2 seats) is big amount of
money, taking in consideration that there are free tools and that it is
not the only thing you have to buy. If you are starting with little
money (and most of starting business does), you are going to take care
of all the money you are going to spend.
"Javier" <·······@gmail.com> writes:
> I've been not in software business, but in other areas and not of a
> such technical level.
> Elsewhere, I still think that if you are a freelance worker, or even
> associating to another person, $7800 (for 2 seats) is big amount of
> money, taking in consideration that there are free tools and that it is
> not the only thing you have to buy. If you are starting with little
> money (and most of starting business does), you are going to take care
> of all the money you are going to spend.
What other area of business have you been in that the tools of the
trade were free?
Ari Johnson ha escrito:
> "Javier" <·······@gmail.com> writes:
> > I've been not in software business, but in other areas and not of a
> > such technical level.
> > Elsewhere, I still think that if you are a freelance worker, or even
> > associating to another person, $7800 (for 2 seats) is big amount of
> > money, taking in consideration that there are free tools and that it is
> > not the only thing you have to buy. If you are starting with little
> > money (and most of starting business does), you are going to take care
> > of all the money you are going to spend.
>
> What other area of business have you been in that the tools of the
> trade were free?
I owned a cybercafe and the operating system was Linux. It had 16 seats
and saved me a good amount of money in licensing, and a lot of
headaches with virus, reconfigurations and more. I give up the
business because of some problematic clients and to have a better job,
but not because it was not success economically speaking. I know it is
not at the technical level of a software company, but it allowed me to
realised what people want, what the costs are, and how to maximize it
while not being tied to a software company.
I know some examples of other cibercafes that wasted a lot of money for
external companies doing the control software, tunning Windows, and
having to do periodical revisions and reinstallations (more often that
you can believe, some days there were more than 5 computers not working
at the same time, even using programs like RestoreIt and similars. 5
seats not working costs a lot of money.).
I wrote my own control software (in TCL and C), and configured the
clients and the server myself, using only free software, and some
commercial games using Cedega (paying their license, of course).
Curiusly, that licensed software where the most expensive ones, both
because of their bugs (games does have a lot, and they run on top of
Cedega in my case) and the big amount of money I gave them.
Also, I've been working for years in two different companies as a
programer (mainly Java, PHP and VB, against my pleasure) which both
used Linux and Windows, and I can assure that the Windous machines
where the most problematic and expensive ones.
I'm actually using OSX, and I agree that it is the best of the OS I've
been working all of these years. But my opinion is that it is because
OSX is Darwin (open source based in BSD) with Aqua at the user
interface.
I do not consider myself as an open source guru, not even a defender. I
prefer, from my own experience, to use open source, and buy programs
only when there is not alternative for what I require.
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2psebxmqb.fsf@hermes.theari.com>
"Javier" <·······@gmail.com> writes:
> Ari Johnson ha escrito:
>
>> "Javier" <·······@gmail.com> writes:
>> > I've been not in software business, but in other areas and not of a
>> > such technical level.
>> > Elsewhere, I still think that if you are a freelance worker, or even
>> > associating to another person, $7800 (for 2 seats) is big amount of
>> > money, taking in consideration that there are free tools and that it is
>> > not the only thing you have to buy. If you are starting with little
>> > money (and most of starting business does), you are going to take care
>> > of all the money you are going to spend.
>>
>> What other area of business have you been in that the tools of the
>> trade were free?
>
> I owned a cybercafe and the operating system was Linux. It had 16 seats
> and saved me a good amount of money in licensing, and a lot of
> headaches with virus, reconfigurations and more. I give up the
> business because of some problematic clients and to have a better job,
> but not because it was not success economically speaking. I know it is
> not at the technical level of a software company, but it allowed me to
> realised what people want, what the costs are, and how to maximize it
> while not being tied to a software company.
The operating system was free and did everything that any other
operating system would have in that application with not a lot of
extra work on your part. How about all the other equipment? Were the
chairs, tables, coffee equipment, building, and staff all free, as
well?
The point is simply this: if your business plan is made or broken by
$3,900 of software and you cannot cost-justify the man-hours of
implementing what you need using free software, then your problem is
not the software prices but instead is your business plan as a whole.
Ari Johnson ha escrito:
> The operating system was free and did everything that any other
> operating system would have in that application with not a lot of
> extra work on your part. How about all the other equipment? Were the
> chairs, tables, coffee equipment, building, and staff all free, as
> well?
They were not free, but I took care of selecting the ones with the best
compromise between price and quality, and not directly selected the
most expensives ones because I was cheap nor broken, but because I
precissely had a busines plan and a budget to fit into.
> The point is simply this: if your business plan is made or broken by
> $3,900 of software and you cannot cost-justify the man-hours of
> implementing what you need using free software, then your problem is
> not the software prices but instead is your business plan as a whole.
Not at all. You can start successfully working business with just $1.
Probably not the one you want to, but you can.
Take for example Microsoft and Google, they started out working in
their garages. They had no money but a great ambition and later a
fantastic business plan. Now they are where they are.
Elsewhere, and following up my original answer for Ken, most of the
entrepreneurs start their bussiness with a ridiculus amount of money,
probably with $60k, $30k, and even $10k. There are even people who just
start with a computer, a telephon and a plug with electricity to
connect it. For them, $3900 or even $1100 is a considerable amount of
money and certainly deserve to be carefully considered.
And I forget to talk about the discriminating opensourcered programmer.
S/He is not going to depend from a commercial compiler unless s/he
isn't afraid of his/er work being ignored by the open source comunity.
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2bqpv7zzs.fsf@hermes.theari.com>
"Javier" <·······@gmail.com> writes:
> Not at all. You can start successfully working business with just $1.
> Probably not the one you want to, but you can.
> Take for example Microsoft and Google, they started out working in
> their garages. They had no money but a great ambition and later a
> fantastic business plan. Now they are where they are.
I assure you that Microsoft and Google both started in well-appointed
garages.
> Elsewhere, and following up my original answer for Ken, most of the
> entrepreneurs start their bussiness with a ridiculus amount of money,
> probably with $60k, $30k, and even $10k. There are even people who just
> start with a computer, a telephon and a plug with electricity to
> connect it. For them, $3900 or even $1100 is a considerable amount of
> money and certainly deserve to be carefully considered.
> And I forget to talk about the discriminating opensourcered programmer.
> S/He is not going to depend from a commercial compiler unless s/he
> isn't afraid of his/er work being ignored by the open source comunity.
And all those people can use free software to do their work. What's
the problem with a company identifying a market for premium
development tools and charging money to that market?
Ari Johnson wrote:
>
> And all those people can use free software to do their work. What's
> the problem with a company identifying a market for premium
> development tools and charging money to that market?
Well it's like the 3D workstation vendors of yesteryear. Those who
resisted commodification eventually died. I mean, there are an awful
lot of competing languages out there, and Lisp doesn't have to retain
its advantages forever.
Cheers,
Brandon Van Every
Ari Johnson <·········@gmail.com> writes:
> "Javier" <·······@gmail.com> writes:
>
>> Not at all. You can start successfully working business with just $1.
>> Probably not the one you want to, but you can.
>> Take for example Microsoft and Google, they started out working in
>> their garages. They had no money but a great ambition and later a
>> fantastic business plan. Now they are where they are.
>
> I assure you that Microsoft and Google both started in well-appointed
> garages.
Indeed. Bill Gates never had to sell anything to go on with his
business. Steve Jobs and Steven Wozniak on the other hand, they did
sell their VW van to buy components.
--
__Pascal Bourguignon__ http://www.informatimago.com/
This is a signature virus. Add me to your signature and help me to live.
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m23bb77ym3.fsf@hermes.theari.com>
Pascal Bourguignon <···@informatimago.com> writes:
> Ari Johnson <·········@gmail.com> writes:
>
>> "Javier" <·······@gmail.com> writes:
>>
>>> Not at all. You can start successfully working business with just $1.
>>> Probably not the one you want to, but you can.
>>> Take for example Microsoft and Google, they started out working in
>>> their garages. They had no money but a great ambition and later a
>>> fantastic business plan. Now they are where they are.
>>
>> I assure you that Microsoft and Google both started in well-appointed
>> garages.
>
> Indeed. Bill Gates never had to sell anything to go on with his
> business. Steve Jobs and Steven Wozniak on the other hand, they did
> sell their VW van to buy components.
And probably to make space in the garage to get some work done. ;)
Javier wrote:
>
> And I forget to talk about the discriminating opensourcered programmer.
> S/He is not going to depend from a commercial compiler unless s/he
> isn't afraid of his/er work being ignored by the open source comunity.
Indeed, I made a decision about this once upon a time with my pursuit
of high level languages. The goal isn't just to develop my own stuff.
The goal, in the long run, is to get the game industry to do
development my way. Because C++ sucks, frankly. Now, we'll see once
I'm rich and famous whether I care anymore. Possibly I will still
care. I do have a strong technology streak in me.
Now I'm at the murky precipice of "to 3D engine or not to 3D engine?"
I've skirted around that chasm before. On the one hand, it could save
me a lot of low-level 3D graphical gruntwork. On the other hand, I may
not need that much for my specific game, and I may need to generate it
procedurally in a way that off-the-shelf 3D engines can't do anyways.
Definitely there can be maintenance nightmares using other people's
code. Chicken has taken me 10 months, just so I could have MinGW
working nicely. Now what if it's the same story again with 3D engines?
I've been looking at G3D, which is a 3D library not a full engine.
http://g3d-cpp.sourceforge.net/ It's a mature open source project,
it's been proven in the field, but it doesn't appear that anyone's
currently building with MinGW at all. That spells trouble. On the
other hand, they're trying to move to a CMake build system, so that
could solve the problem. Still, I haven't figured out whether this
thing really works on Windows. I guess I could finish that mulling
now, since I'm thinking about it. Got interrupted the other day.
Cheers,
Brandon Van Every
Mallor ha escrito:
> Now I'm at the murky precipice of "to 3D engine or not to 3D engine?"
> I've skirted around that chasm before. On the one hand, it could save
> me a lot of low-level 3D graphical gruntwork. On the other hand, I may
> not need that much for my specific game, and I may need to generate it
> procedurally in a way that off-the-shelf 3D engines can't do anyways.
> Definitely there can be maintenance nightmares using other people's
> code. Chicken has taken me 10 months, just so I could have MinGW
> working nicely. Now what if it's the same story again with 3D engines?
I would recommend doing the 3D engine by yourself, taking in
consideration what you're going to need when developing the game.
I will take longer, but you'll end up with a 3D engine that anybody
could use, and that yourself can use several times.
And you should not worry too much about the time required, unless your
life will depend on this (and I doubt it will). You should worry only
on having a lot of fun! :)
> I've been looking at G3D, which is a 3D library not a full engine.
> http://g3d-cpp.sourceforge.net/ It's a mature open source project,
> it's been proven in the field, but it doesn't appear that anyone's
> currently building with MinGW at all. That spells trouble. On the
> other hand, they're trying to move to a CMake build system, so that
> could solve the problem. Still, I haven't figured out whether this
> thing really works on Windows. I guess I could finish that mulling
> now, since I'm thinking about it. Got interrupted the other day.
Windows is not a good plattform for developing open source programs.
You might better base your work in core OpenGL. It is a standar, and it
is cross-plattform. SBCL will be avaiable soon for windows, perhaps
before you can publish any serius work. Meanwhile you can use CLISP or
buy Cornan.
I don't know much about Chicken (have been using for some days only),
but I think is not as adecuate and convenient as SBCL/CMUCL because its
speed, its debugging capabilities, and the standar library.
Javier wrote:
>
> Windows is not a good plattform for developing open source programs.
Truly it is a PITA. I'm pretty much in the vanguard on this. CMake is
the right infrastructural tool to get things to happen. My new
homemade bizcards say 2 notable things on them. I pitch "Open Source
on Windows" as one of the things I do, and at the bottom of my card it
says, "Use CMake. No one need suffer GNU Autoconf."
> You might better base your work in core OpenGL. It is a standar, and it
> is cross-plattform.
That's pretty much why I'm evaluating G3D. My initial reaction is that
even though G3D is only a library, it does stuff that as a Scheme
programmer I simply don't want. Like helper functions that do things
better than the C++ STL and whatnot.
> SBCL will be avaiable soon for windows, perhaps
> before you can publish any serius work. Meanwhile you can use CLISP or
> buy Cornan.
> I don't know much about Chicken (have been using for some days only),
> but I think is not as adecuate and convenient as SBCL/CMUCL because its
> speed, its debugging capabilities, and the standar library.
Chicken doesn't have any speed problem. It's a Scheme-to-C compiler,
and all compilers are in the same rough equivalence class as far as
performance goes. Plus it is BSD licensed, and I am an optimization
jock, so speeding it up doesn't trouble me.
Debugging, dunno... I'm not a fan of debuggers anyways. Generally I
use source control and make micro-incremental changes. Debugging is
typically only necessary when I'm porting someone else's code, or when
a problem is exceptionally thorny. Yes, sometimes I need a debugger.
Don't know that it'll happen often enough for me to forswear languages
without a debugger; we'll see.
Standard library, not important. I need OpenGL + core language
capabilities.
Chicken runs on Windows 'cuz I just spent 10 months making it run. It
ain't broke so I ain't fixin' it. We'll see how far I get in Stage II.
Cheers,
Brandon Van Every
Mallor wrote:
<snipped>
> the right infrastructural tool to get things to happen. My new
> homemade bizcards say 2 notable things on them. I pitch "Open Source
> on Windows" as one of the things I do, and at the bottom of my card it
> says, "Use CMake. No one need suffer GNU Autoconf."
>
Don't put evangelism on business cards; it's a stupid
business decision and smart businessmen like to do business
with other smart businessmen, not smart techies.
<snipped>
>
> Chicken runs on Windows 'cuz I just spent 10 months making it run. It
> ain't broke so I ain't fixin' it. We'll see how far I get in Stage II.
>
What *is* stage 2? Writing an application with it?
Writing a few libraries for it? Wrapping up win32 API?
To be perfectly honest, you might just have wasted
your 10 months. It certainly doesn't take 10 months
to write a simple language implementation, especially
if you plan on emitting C code instead native opcodes.
goose,
techies, OTOH, don't really read business cards
anyway so won't see your message, and even if they
*did* read, it's doubtful that a single sentence
would urge them any further than their own
opinions would do.
goose <····@webmail.co.za> wrote:
> To be perfectly honest, you might just have wasted
> your 10 months. It certainly doesn't take 10 months
> to write a simple language implementation, especially
> if you plan on emitting C code instead native opcodes.
I don't think he is that incompetent. I'm sure these 10 months
aren't full-time.
Stefan Scholl wrote:
> goose <····@webmail.co.za> wrote:
> > To be perfectly honest, you might just have wasted
> > your 10 months. It certainly doesn't take 10 months
> > to write a simple language implementation, especially
> > if you plan on emitting C code instead native opcodes.
>
> I don't think he is that incompetent. I'm sure these 10 months
> aren't full-time.
Whoops! I never meant to imply that he is incompetent
("that" or otherwise). I apologise for any negative
implication about his (BvE) technical abilities.
Maybe I should clarify: If he had known *in advance*
that it would take 10 months to get a windows platform
build going (whether part-time or full-time), I doubt
that he would have proceeded with it.
I have often embarked on a porting (and/or) fixing spree
which was, in hindsight, a non-optimal decision. Doesn't
make me incompetent, just (in my case) naive about the
competence of the developers who wrote the original
code. Also means that I had wasted time on fixing
bugs *that should not have been there in the first place*
instead of solving the problem at hand.
IME a few (or even a single) bad programmers can
drag even the most productive programmer down to
their level. This is why badly written code is
almost always more expensive in the long run that
well-written code, no matter how little the original
coder charged.
goose,
to BvE, sorry.
goose wrote:
> Stefan Scholl wrote:
> > goose <····@webmail.co.za> wrote:
> > > To be perfectly honest, you might just have wasted
> > > your 10 months. It certainly doesn't take 10 months
> > > to write a simple language implementation, especially
> > > if you plan on emitting C code instead native opcodes.
> >
> > I don't think he is that incompetent. I'm sure these 10 months
> > aren't full-time.
>
> Whoops! I never meant to imply that he is incompetent
> ("that" or otherwise). I apologise for any negative
> implication about his (BvE) technical abilities.
Why thank you. I'm surprised you underestimated the difficulty of
kicking a build into shape on 3 different Windows compiler targets,
while simultaneously maintaining the stability of 3 completely
different legacy build systems, and also implementing support for Linux
and MacOS X, but thanks nevertheless.
> Maybe I should clarify: If he had known *in advance*
> that it would take 10 months to get a windows platform
> build going (whether part-time or full-time), I doubt
> that he would have proceeded with it.
I would have thought very, very carefully about whether the Corman Lisp
community was healthy enough to risk the future on. It didn't look
very healthy a year ago, but then, I had other options available it
seemed. I was sick of Bigloo build problems and wanted to move on, but
viewed from hindsight, I could have plowed through those issues in much
less time than it took me to set Chicken aright. I do think that *at
present* I am better off with Chicken than Bigloo, however.
Another factor I'm often remiss to bring up, is that 1 year ago I was
thinking about writing my own compiler. I did think of ditching Lisp
and Scheme altogether, and just bootstrapping a new language from C and
Assembly code. Given that I had leanings in this direction, Chicken's
BSD license was much more attractive than Bigloo's GPL license or
Corman Lisp's proprietary license. And that is still true today. If I
choose to optimize the hell out of Chicken, and develop some kind of
special "game domain language," there's nothing stopping me from going
commercial with it someday. I have no fear of losing my investments, I
fully own the code and can do what I want with it.
> I have often embarked on a porting (and/or) fixing spree
> which was, in hindsight, a non-optimal decision. Doesn't
> make me incompetent, just (in my case) naive about the
> competence of the developers who wrote the original
> code.
Felix and the Chicken contributors are very competent. They are not,
however, Windows-centric. That is the new ingredient I brought to
Chicken development. I've caused a lot more testing and bugfixing to
happen on MinGW and MSVC. Did a lot of that bugfixing and refactoring
myself, that's part of those 10 months. Nowadays we also have John
Cowan slinging test builds in that halfway house known as Cygwin. His
testing has caused me to improve the Cygwin CMake support. I had no
basic motivation to do it otherwise. I would say that at present,
MinGW, Cygwin, and MSVC are all supported equally well.
The value of the working relationships I've established with Felix, the
Chicken community, Kitware, and the CMake community, are yet more
reasons I'm not inclined to switch horses. The dirty little secret of
open source is it's not about the code, it's about the people who keep
things running. These are good acts, they get things done.
Cheers,
Brandon Van Every
Stefan Scholl wrote:
> goose <····@webmail.co.za> wrote:
> > To be perfectly honest, you might just have wasted
> > your 10 months. It certainly doesn't take 10 months
> > to write a simple language implementation, especially
> > if you plan on emitting C code instead native opcodes.
What crack are you smoking? Why don't you go bug Felix Winkelmann
about how much work goes into Chicken Scheme?
> I don't think he is that incompetent.
For the record, neither does Felix. Or else how do you think we're
going with my build as the standard distribution mechanism from now on?
> I'm sure these 10 months aren't full-time.
You'd be wrong. I've had months where I was hardly doing anything for
Chicken Scheme at all, and months where I was working double-time on
it. It averages out to full time, defined as 40 hour work weeks.
I tend to have contempt for people who think build engineering problems
are easy. Actually they are incredibly laborious.
Cheers,
Brandon Van Every
"Mallor" <···········@gmail.com> writes:
> I tend to have contempt for people who think build engineering problems
> are easy. Actually they are incredibly laborious.
And I tend to have contempt (as I think I've made clear) for people
who think everyone on comp.lang.lisp *cares* about their build
engineering problems.
Bill Atkins wrote:
> "Mallor" <···········@gmail.com> writes:
> > I tend to have contempt for people who think build engineering problems
> > are easy. Actually they are incredibly laborious.
> And I tend to have contempt (as I think I've made clear) for people
> who think everyone on comp.lang.lisp *cares* about their build
> engineering problems.
And I tend to have contempt for people who tend to have contempt for
people. (I thought I'd sneak this in before Kenny did!)
········@gmail.com wrote:
>
> And I tend to have contempt for people who tend to have contempt for
> people. (I thought I'd sneak this in before Kenny did!)
So whose language implementation will tail call optimize that better?
Cheers,
Brandon Van Every
Bill Atkins wrote:
> "Mallor" <···········@gmail.com> writes:
>
> > I tend to have contempt for people who think build engineering problems
> > are easy. Actually they are incredibly laborious.
>
> And I tend to have contempt (as I think I've made clear) for people
> who think everyone on comp.lang.lisp *cares* about their build
> engineering problems.
Google lists your post as #124. You have no one to blame but yourself
for what you read and respond to.
Cheers,
Brandon Van Every
Mallor wrote:
> Stefan Scholl wrote:
> > goose <····@webmail.co.za> wrote:
> > > To be perfectly honest, you might just have wasted
> > > your 10 months. It certainly doesn't take 10 months
> > > to write a simple language implementation, especially
> > > if you plan on emitting C code instead native opcodes.
>
> What crack are you smoking? Why don't you go bug Felix Winkelmann
> about how much work goes into Chicken Scheme?
Err. Slow down there :-) No need to fly off the handle,
I clarified my position in elsethread.
>
> I tend to have contempt for people who think build engineering problems
> are easy. Actually they are incredibly laborious.
>
I agree, which is why I tend to favour C for small things:
build on all systems known to man (hosted, of course),
natively. It does rather limit the solution to "only
solutions that can be implemented in C", though.
goose,
Mallor <···········@gmail.com> wrote:
> Chicken doesn't have any speed problem. It's a Scheme-to-C compiler,
> and all compilers are in the same rough equivalence class as far as
> performance goes. Plus it is BSD licensed, and I am an optimization
Slower than Java in the shootout.
Javier wrote:
> And I forget to talk about the discriminating opensourcered programmer.
> S/He is not going to depend from a commercial compiler unless s/he
> isn't afraid of his/er work being ignored by the open source comunity.
Why do you have to commit to one or the other?
What about using open source and selling a discounted 'beta' program to
a few select customers. Then you might be able to fund a commercial
development suite. After that, release the product built with the
commercial product and sell it at the target price.
Or, as some developers have done, write an open-source core, and sell
proprietary plug-ins or other enhancements, even documentation and/or
support?
Javier <·······@gmail.com> wrote:
> Take for example Microsoft and Google, they started out working in
> their garages. They had no money but a great ambition and later a
Microsoft invented the Internet and the GUI, too!
Stefan Scholl <······@no-spoon.de> writes:
> Microsoft invented the Internet and the GUI, too!
Ha, I did not know Al Gore worked for Microsoft! When?
Benjamin
--
"Master your instrument, master the music, and then forget all that
bullshit and just play." -- Charlie Parker
Ari Johnson <·········@gmail.com> writes:
> "Javier" <·······@gmail.com> writes:
>
>> Ari Johnson ha escrito:
>>
>>> "Javier" <·······@gmail.com> writes:
>>> > I've been not in software business, but in other areas and not of a
>>> > such technical level.
>>> > Elsewhere, I still think that if you are a freelance worker, or even
>>> > associating to another person, $7800 (for 2 seats) is big amount of
>>> > money, taking in consideration that there are free tools and that it is
>>> > not the only thing you have to buy. If you are starting with little
>>> > money (and most of starting business does), you are going to take care
>>> > of all the money you are going to spend.
>>>
>>> What other area of business have you been in that the tools of the
>>> trade were free?
>>
>> I owned a cybercafe and the operating system was Linux. It had 16 seats
>> and saved me a good amount of money in licensing, and a lot of
>> headaches with virus, reconfigurations and more. I give up the
>> business because of some problematic clients and to have a better job,
>> but not because it was not success economically speaking. I know it is
>> not at the technical level of a software company, but it allowed me to
>> realised what people want, what the costs are, and how to maximize it
>> while not being tied to a software company.
>
> The operating system was free and did everything that any other
> operating system would have in that application with not a lot of
> extra work on your part. How about all the other equipment? Were the
> chairs, tables, coffee equipment, building, and staff all free, as
> well?
>
> The point is simply this: if your business plan is made or broken by
> $3,900 of software and you cannot cost-justify the man-hours of
> implementing what you need using free software, then your problem is
> not the software prices but instead is your business plan as a whole.
You're missing the bootstrapping process. $3,900 can kill the
bootstrapping. Of course, it doesn't matter for businessmen who can
get VC to finance them at a minimum of $1,000,000, but if you're not
so fortunate, you just avoid any unnecessary expense, until you don't
need the VC anymore.
--
__Pascal Bourguignon__ http://www.informatimago.com/
"Remember, Information is not knowledge; Knowledge is not Wisdom;
Wisdom is not truth; Truth is not beauty; Beauty is not love;
Love is not music; Music is the best." -- Frank Zappa
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2k64j89et.fsf@hermes.theari.com>
Pascal Bourguignon <···@informatimago.com> writes:
> Ari Johnson <·········@gmail.com> writes:
>
>> "Javier" <·······@gmail.com> writes:
>>
>>> Ari Johnson ha escrito:
>>>
>>>> "Javier" <·······@gmail.com> writes:
>>>> > I've been not in software business, but in other areas and not of a
>>>> > such technical level.
>>>> > Elsewhere, I still think that if you are a freelance worker, or even
>>>> > associating to another person, $7800 (for 2 seats) is big amount of
>>>> > money, taking in consideration that there are free tools and that it is
>>>> > not the only thing you have to buy. If you are starting with little
>>>> > money (and most of starting business does), you are going to take care
>>>> > of all the money you are going to spend.
>>>>
>>>> What other area of business have you been in that the tools of the
>>>> trade were free?
>>>
>>> I owned a cybercafe and the operating system was Linux. It had 16 seats
>>> and saved me a good amount of money in licensing, and a lot of
>>> headaches with virus, reconfigurations and more. I give up the
>>> business because of some problematic clients and to have a better job,
>>> but not because it was not success economically speaking. I know it is
>>> not at the technical level of a software company, but it allowed me to
>>> realised what people want, what the costs are, and how to maximize it
>>> while not being tied to a software company.
>>
>> The operating system was free and did everything that any other
>> operating system would have in that application with not a lot of
>> extra work on your part. How about all the other equipment? Were the
>> chairs, tables, coffee equipment, building, and staff all free, as
>> well?
>>
>> The point is simply this: if your business plan is made or broken by
>> $3,900 of software and you cannot cost-justify the man-hours of
>> implementing what you need using free software, then your problem is
>> not the software prices but instead is your business plan as a whole.
>
> You're missing the bootstrapping process. $3,900 can kill the
> bootstrapping. Of course, it doesn't matter for businessmen who can
> get VC to finance them at a minimum of $1,000,000, but if you're not
> so fortunate, you just avoid any unnecessary expense, until you don't
> need the VC anymore.
And if you are bootstrapping, what is wrong with SBCL, OpenMCL, or
clisp? Plus, if the idea is sufficiently good, you can find capital.
There are actually projects that I don't have the time to do
personally but would gladly put up some capital to get done - there is
a definite market for software development if you spot the gaps or can
get financing from someone who has spotted the gaps.
Ari Johnson wrote:
> Pascal Bourguignon <···@informatimago.com> writes:
> >
> > You're missing the bootstrapping process. $3,900 can kill the
> > bootstrapping. Of course, it doesn't matter for businessmen who can
> > get VC to finance them at a minimum of $1,000,000, but if you're not
> > so fortunate, you just avoid any unnecessary expense, until you don't
> > need the VC anymore.
>
> And if you are bootstrapping, what is wrong with SBCL, OpenMCL, or
> clisp?
For Windows game development, the first 2 do not work on Windows. It
is important to use stuff that actually works on your platform,
particularly open source stuff. Nothing open source is going to work
on Windows unless it is regularly stress tested. You simply can't take
for granted that open source gives a hoot about Windows. Generally
they don't, so if you're not constantly vetting your tools, you're
going to have trouble. I have this debate with a friend of mine, who
advocates developing on Linux and "porting once later," rather than all
this blood spilling I've done with Chicken Scheme. I say he's not
considering the realities of vetting tools on Windows. If I hadn't
stepped into the breach 10 months ago, Chicken Scheme simply wouldn't
be viable for Windows game development.
CLisp works on Windows, but it is only an interpreter, and it has a C
FFI specific to CLisp. Performance code is core to the kinds of games
I wish to write; that is, 4X Turn Based Strategy in the mold of
Civilization or Sid Meier's Alpha Centauri. Lotta AI number crunching
in those. My original project vision was strongly 3D graphical as
well, although I'm less interested in that lately. Anyways there are
performance requirements. I could have gotten started with CLisp, but
I would need C interfaces to OpenGL libaries, or C++ interfaces to
off-the-shelf 3D engines, almost immediately. I didn't want to waste
development effort on C/C++ FFIs that I was just going to throw out.
I've said a number of times, there's nothing "Common" about Common Lisp
with respect to C at all, let alone C++.
Corman Lisp would have been a reasonable choice. A Corman-specific C
FFI yet again, but at least for a Windows version of the game, I
wouldn't have to drop Corman to ship it. I did feel, however, with C
FFIs each being "a law unto themselves," that I was no better off than
in the Scheme world and I should go back to it. So I did. Also at the
time, the Corman community looked like it was on its last legs. But
possibly the imminent death of Corman has been exaggerated on a number
of occasions.
I had some peripheral adoption considerations a year ago. I was trying
to build an Eclipse based toolchain. That was to be the "slick for the
biz types" world domination strategy. There were no Lisp plugins;
there was the SchemeWay plugin. But it never worked for me, and
eventually I gave up and went to XEmacs. Also I wanted to poison the
minds of other game developers. They could try Corman for free, but I
don't know that they'd stick with something they had to pay for. That
probably wasn't the biggest issue though. They probably wouldn't stick
with anything less than a fullblown development system that I had
shipped a game with, that was proven in the marketplace.
> Plus, if the idea is sufficiently good, you can find capital.
Not really, in the creative industries.
> There are actually projects that I don't have the time to do
> personally but would gladly put up some capital to get done - there is
> a definite market for software development if you spot the gaps or can
> get financing from someone who has spotted the gaps.
You know, I don't have time for all this mess. I don't know if
everyone else in software just gets rich hanging out in big companies,
sipping martinis with people who have got piles of money they want to
give away, but I'm busy coding. I stay in the trenches trying to solve
real problems. And unfortunately, it's easy for this to be profitless,
I've found out. Even for the most simple survival jobhunting right
now, I am constantly torn between the need to survive vs. my strong
will / determination to kick the Chicken Scheme build out the door.
The latter is certainly a more rewarding pursuit than the former,
obviously. Now why is jobhunting so difficult? Because everyone wants
C++, Java, C#, or SQL in this town, to judge by the ads. Not Lisp,
Scheme, or CMake build systems.
I did start a local organization, partly to try to foster business
interests that are lacking in the region. That aspect of SeaFunc is a
failure. It could succeed in the future, but at present, the failure
is the complete lack of interest on anybody's part to lift so much as a
finger for business issues. And the last guy who wanted to re-launch
LispSea, with more of a business promotion focus, gave up in a huff
when people weren't immediately jumping at the chance to pursue that
agenda. For my part, I'll start making the rounds with the Washington
Software Alliance when I'm no longer concerned with my most immediate
survival. The enterpreneurial moneybags aren't coming to us, so I'll
have to go to them.
Cheers,
Brandon Van Every
On 5 Sep 2006 03:08:06 -0700, "Mallor" <···········@gmail.com> wrote:
>
> Ari Johnson wrote:
> > Pascal Bourguignon <···@informatimago.com> writes:
> > >
> > > You're missing the bootstrapping process. $3,900 can kill the
> > > bootstrapping. Of course, it doesn't matter for businessmen who can
> > > get VC to finance them at a minimum of $1,000,000, but if you're not
> > > so fortunate, you just avoid any unnecessary expense, until you don't
> > > need the VC anymore.
> >
> > And if you are bootstrapping, what is wrong with SBCL, OpenMCL, or
> > clisp?
>
> For Windows game development, the first 2 do not work on Windows.
I'm actually getting a little tired of reading this false claim
that SBCL doesn't work on Windows. It does work, well enough to
get started with; I was able to port Graphic-Forms and thus add
SBCL to my list of supported CL implementations. YMMV and maybe
it's current state doesn't qualify as "working" for you, but it
does for me.
Now, if you want to talk about some annoying rough edges that
remain, OK let's talk. But such a discussion has to start with
the premise that in fact a lot of work went into getting SBCL
running as well as it does on Windows today.
--
Jack Unrue
Jack Unrue wrote:
> On 5 Sep 2006 03:08:06 -0700, "Mallor" <···········@gmail.com> wrote:
> >
> > Ari Johnson wrote:
> > > Pascal Bourguignon <···@informatimago.com> writes:
> > > >
> > > > You're missing the bootstrapping process. $3,900 can kill the
> > > > bootstrapping. Of course, it doesn't matter for businessmen who can
> > > > get VC to finance them at a minimum of $1,000,000, but if you're not
> > > > so fortunate, you just avoid any unnecessary expense, until you don't
> > > > need the VC anymore.
> > >
> > > And if you are bootstrapping, what is wrong with SBCL, OpenMCL, or
> > > clisp?
> >
> > For Windows game development, the first 2 do not work on Windows.
>
> I'm actually getting a little tired of reading this false claim
> that SBCL doesn't work on Windows. It does work, well enough to
> get started with; I was able to port Graphic-Forms and thus add
> SBCL to my list of supported CL implementations. YMMV and maybe
> it's current state doesn't qualify as "working" for you, but it
> does for me.
>
> Now, if you want to talk about some annoying rough edges that
> remain, OK let's talk. But such a discussion has to start with
> the premise that in fact a lot of work went into getting SBCL
> running as well as it does on Windows today.
The last Windows SBCL released is 0.9.12 it seems to have lot of rough
edges to me. It often crashes or produces bizarre error messages.
It's not something you could do development with.
I understand this will change soon though, and I look forward to it.
On 5 Sep 2006 09:04:28 -0700, "Rob Thorpe" <·············@antenova.com> wrote:
>
> The last Windows SBCL released is 0.9.12 it seems to have lot of rough
> edges to me. It often crashes or produces bizarre error messages.
> It's not something you could do development with.
The platform page lists 0.9.12, but versions tagged since then have
continued to build and run (0.9.16 requires a trivial workaround).
I agree, it does crash. I did (and still do) development with it.
> I understand this will change soon though, and I look forward to it.
Yep. I'm not sure yet how to help get the 'GC invariant lost' problem
fixed, it's hard to reproduce in a predictable manner.
--
Jack Unrue
Jack Unrue wrote:
> On 5 Sep 2006 09:04:28 -0700, "Rob Thorpe" <·············@antenova.com> wrote:
> >
> > The last Windows SBCL released is 0.9.12 it seems to have lot of rough
> > edges to me. It often crashes or produces bizarre error messages.
> > It's not something you could do development with.
>
> The platform page lists 0.9.12, but versions tagged since then have
> continued to build and run (0.9.16 requires a trivial workaround).
I've heard this, but I don't know how to build it from source on
Windows without a working binary.
> I agree, it does crash. I did (and still do) development with it.
>
> > I understand this will change soon though, and I look forward to it.
>
> Yep. I'm not sure yet how to help get the 'GC invariant lost' problem
> fixed, it's hard to reproduce in a predictable manner.
Sorry, neither do I :)
On 5 Sep 2006 10:30:23 -0700, "Rob Thorpe" <·············@antenova.com> wrote:
>
> Jack Unrue wrote:
> > On 5 Sep 2006 09:04:28 -0700, "Rob Thorpe" <·············@antenova.com> wrote:
> > >
> > > The last Windows SBCL released is 0.9.12 it seems to have lot of rough
> > > edges to me. It often crashes or produces bizarre error messages.
> > > It's not something you could do development with.
> >
> > The platform page lists 0.9.12, but versions tagged since then have
> > continued to build and run (0.9.16 requires a trivial workaround).
>
> I've heard this, but I don't know how to build it from source on
> Windows without a working binary.
I used the 0.9.12 binary to get .15 built, then later .15 to get .16
built.
Some folks (including myself) have encountered the following problem
at startup:
VirtualAlloc: No error
ensure_space failed to validate ...
This is another rough edge in the Windows port. There is a workaround
for this problem that involves finding out which DLLs have their base
addresses in the memory ranges that SBCL wants to reserve. There is a
dynamic relocation patch floating around, but I don't know much about
the details of that patch at this point.
--
Jack Unrue
Jack Unrue wrote:
> On 5 Sep 2006 03:08:06 -0700, "Mallor" <···········@gmail.com> wrote:
> >
> > Ari Johnson wrote:
> > >
> > > And if you are bootstrapping, what is wrong with SBCL, OpenMCL, or
> > > clisp?
> >
> > For Windows game development, the first 2 do not work on Windows.
>
> I'm actually getting a little tired of reading this false claim
> that SBCL doesn't work on Windows. It does work, well enough to
> get started with; I was able to port Graphic-Forms and thus add
> SBCL to my list of supported CL implementations. YMMV and maybe
> it's current state doesn't qualify as "working" for you, but it
> does for me.
>
> Now, if you want to talk about some annoying rough edges that
> remain, OK let's talk. But such a discussion has to start with
> the premise that in fact a lot of work went into getting SBCL
> running as well as it does on Windows today.
A year ago, when I made my committment to Chicken Scheme, SBCL most
definitely didn't work on Windows. If it's actually ready for people
to do real work with it now, it's very recent, like past 2 months.
This is not "ready for prime time" stuff, you don't risk a business on
such things. Microsoft engineers have an epithet for people who say
things like you do. "Works on my box." Here's the matrix of what SBCL
currently claims to support:
http://sbcl.sourceforge.net/platform-table.html If you look at the
file release page
http://sourceforge.net/project/showfiles.php?group_id=1373 you will see
binaries for Linux, Darwin, and FreeBSD, but no Windows. At a glance,
why should I believe you? But I will download it and see for myself,
just to rub it in that it's really not there yet.
Cheers,
Brandon Van Every
Mallor wrote:
> Jack Unrue wrote:
> > On 5 Sep 2006 03:08:06 -0700, "Mallor" <···········@gmail.com> wrote:
> > >
> > > Ari Johnson wrote:
> > > >
> > > > And if you are bootstrapping, what is wrong with SBCL, OpenMCL, or
> > > > clisp?
> > >
> > > For Windows game development, the first 2 do not work on Windows.
> >
> > I'm actually getting a little tired of reading this false claim
> > that SBCL doesn't work on Windows. It does work, well enough to
> > get started with; I was able to port Graphic-Forms and thus add
> > SBCL to my list of supported CL implementations. YMMV and maybe
> > it's current state doesn't qualify as "working" for you, but it
> > does for me.
> >
> > Now, if you want to talk about some annoying rough edges that
> > remain, OK let's talk. But such a discussion has to start with
> > the premise that in fact a lot of work went into getting SBCL
> > running as well as it does on Windows today.
>
> A year ago, when I made my committment to Chicken Scheme, SBCL most
> definitely didn't work on Windows. If it's actually ready for people
> to do real work with it now, it's very recent, like past 2 months.
> This is not "ready for prime time" stuff, you don't risk a business on
> such things. Microsoft engineers have an epithet for people who say
> things like you do. "Works on my box." Here's the matrix of what SBCL
> currently claims to support:
> http://sbcl.sourceforge.net/platform-table.html If you look at the
> file release page
> http://sourceforge.net/project/showfiles.php?group_id=1373 you will see
> binaries for Linux, Darwin, and FreeBSD, but no Windows. At a glance,
> why should I believe you? But I will download it and see for myself,
> just to rub it in that it's really not there yet.
So, I'm looking at the 0.9.16 release, dated August 25, 2006. The
README, INSTALL, etc. documentation don't mention Windows at all. It
is not listed as a supported platform, and there's not a peep about how
to build it, other than following the generalized build instructions
for any old platform. Grepping for "Windows" doesn't yield much
either, other than some indications that a few souls are working on
something. install.sh yields "src/runtime/sbcl.exe not found, aborting
installation." So at this point alone, anyone would conclude that
Windows support is alpha quality.
Someone earlier in this thread mentioned that earlier SBCLs had win32
binaries, and that they've used these for bootstrap. I'm currently
downloading 0.9.12, the last version that offered a win32 binary. In
my next installment, I'll report on whether it's even vaguely useful
for building 0.9.16. A 4-release gap in proferred binaries does not
instill great expectations.
Cheers,
Brandon Van Every
On 5 Sep 2006 14:24:16 -0700, "Mallor" <···········@gmail.com> wrote:
>
> So, I'm looking at the 0.9.16 release, dated August 25, 2006. The
> README, INSTALL, etc. documentation don't mention Windows at all. It
> is not listed as a supported platform, and there's not a peep about how
> to build it, other than following the generalized build instructions
> for any old platform. Grepping for "Windows" doesn't yield much
> either, other than some indications that a few souls are working on
> something. install.sh yields "src/runtime/sbcl.exe not found, aborting
> installation." So at this point alone, anyone would conclude that
> Windows support is alpha quality.
Yes, it's alpha quality. That's a lot different than saying it
doesn't run.
> Someone earlier in this thread mentioned that earlier SBCLs had win32
> binaries, and that they've used these for bootstrap. I'm currently
> downloading 0.9.12, the last version that offered a win32 binary. In
> my next installment, I'll report on whether it's even vaguely useful
> for building 0.9.16. A 4-release gap in proferred binaries does not
> instill great expectations.
Since I don't speak for the SBCL team, I can't provide justification
for why they don't position the Win32 port more prominently than they
do.
If you have problems building SBCL, there have already been articles
in this group about common problems and solutions. The only other
comment I will make (and then I'm done) is that you may need to
execute
sh make.sh "sbcl" ; assuming your PATH is set up correctly
because just doing this
sh make.sh
tickles a buglet in the command-line option processing code,
which (after I poked around in the CVS logs) I believe has
been fixed since .16 was tagged.
--
Jack Unrue
Jack Unrue wrote:
>
> Yes, it's alpha quality. That's a lot different than saying it
> doesn't run.
To whom? Especially when 0.9.12 runs and has a win32 binary, and the
last 4 releases up through 0.9.16 don't? This is an indication that
SBCL is not ready to be taken seriously on Windows. The build is a
failure, see my log at the end of this post.
> Since I don't speak for the SBCL team, I can't provide justification
> for why they don't position the Win32 port more prominently than they
> do.
Well I can. The current SBCL Windows build sucks. They either
couldn't get the Windows stuff working for 0.9.16, or someone went on
hiatus.
> If you have problems building SBCL, there have already been articles
> in this group about common problems and solutions.
I'm not interested. You guys wake me when SBCL actually works on
Windows.
Meanwhile, a beta-quality release of the CMake build is imminent for
Chicken Scheme. It will be a unified build; either Autoconf or CMake
can be employed. No Chicken compiler will be required for bootstrap,
all you'll need is a C compiler. It will build under MinGW, Cygwin,
and Visual Studio; CMake produces native VS .sln files for the latter.
I will show you what real Windows support is. I've been at this 10
months, and it is indeed much easier to get a Scheme working properly
than a Lisp working properly. At this rate we're going to be ready for
prime time a year sooner than SBCL will be, even having started in a
similar timeframe.
Cheers,
Brandon Van Every
Brandon J. Van ·····@KAHLO ~/src/sbcl-0.9.16
$ sh make.sh
//starting build: Tue Sep 5 14:54:28 PDT 2006
//SBCL_XC_HOST="sbcl --disable-debugger --userinit NUL --sysinit NUL"
//entering make-config.sh
//ensuring the existence of output/ directory
//initializing /e/devel/src/sbcl-0.9.16/local-target-features.lisp-expr
//guessing default target CPU architecture from host architecture
//setting up CPU-architecture-dependent information
sbcl_arch="x86"
//setting up symlink src/compiler/target
//setting up symlink src/assembly/target
//setting up symlink src/compiler/assembly
//setting up OS-dependent information
//finishing /e/devel/src/sbcl-0.9.16/local-target-features.lisp-expr
Microsoft (R) File Expansion Utility Version 5.00.2134.1
Copyright (C) Microsoft Corp 1990-1999. All rights reserved.
Unrecognized switch --.
//entering make-host-1.sh
//building cross-compiler, and doing first genesis
This is SBCL 0.9.12, an implementation of ANSI Common Lisp.
More information about SBCL is available at <http://www.sbcl.org/>.
SBCL is free software, provided as is, with absolutely no warranty.
It is mostly in the public domain; some portions are provided under
BSD-style licenses. See the CREDITS and COPYING files in the
distribution for more information.
os_map: 3, 0x1000, 02000000, 0x1000.
os_map: 3, 0x2000, 05000000, 0x1000.
os_map: 3, 0x3000, 09000000, 0x15bb000.
This is experimental prerelease support for the Windows platform: use
at your own risk. "Your Kitten of Death awaits!"
fatal error before reaching READ-EVAL-PRINT loop:
The file "NUL" was not found.
Brandon J. Van ·····@KAHLO ~/src/sbcl-0.9.16
$
On 5 Sep 2006 15:09:21 -0700, "Mallor" <···········@gmail.com> wrote:
>
> Jack Unrue wrote:
> >
> > Yes, it's alpha quality. That's a lot different than saying it
> > doesn't run.
>
> To whom? Especially when 0.9.12 runs and has a win32 binary, and the
> last 4 releases up through 0.9.16 don't? This is an indication that
> SBCL is not ready to be taken seriously on Windows. The build is a
> failure, see my log at the end of this post.
So you didn't even bother reading the info I included, which
was meant to help you past that exact problem. Oh by the way:
DRIVEN% uname -a
MINGW32_NT-5.1 DRIVEN 1.0.11(0.46/3/2) 2004-04-30 18:55 i686 unknown
DRIVEN% sbcl
This is SBCL 0.9.16, an implementation of ANSI Common Lisp.
More information about SBCL is available at <http://www.sbcl.org/>.
SBCL is free software, provided as is, with absolutely no warranty.
It is mostly in the public domain; some portions are provided under
BSD-style licenses. See the CREDITS and COPYING files in the
distribution for more information.
os_map: 3, 0x1000, 02000000, 0x1000.
os_map: 3, 0x2000, 05000000, 0x1000.
os_map: 3, 0x3000, 09000000, 0x1562000.
This is experimental prerelease support for the Windows platform: use
at your own risk. "Your Kitten of Death awaits!"
*
> > Since I don't speak for the SBCL team, I can't provide justification
> > for why they don't position the Win32 port more prominently than they
> > do.
>
> Well I can. The current SBCL Windows build sucks. They either
> couldn't get the Windows stuff working for 0.9.16, or someone went on
> hiatus.
>
> > If you have problems building SBCL, there have already been articles
> > in this group about common problems and solutions.
>
> I'm not interested. You guys wake me when SBCL actually works on
> Windows.
There you go again. But I'll let it pass.
> Meanwhile, a beta-quality release of the CMake build is imminent for
> Chicken Scheme. It will be a unified build; either Autoconf or CMake
> can be employed. No Chicken compiler will be required for bootstrap,
> all you'll need is a C compiler. It will build under MinGW, Cygwin,
> and Visual Studio; CMake produces native VS .sln files for the latter.
> I will show you what real Windows support is. I've been at this 10
> months, and it is indeed much easier to get a Scheme working properly
> than a Lisp working properly. At this rate we're going to be ready for
> prime time a year sooner than SBCL will be, even having started in a
> similar timeframe.
Your work on Chicken is to be commended. Brandon, I'm happy to leave
it at that.
--
Jack Unrue
Jack Unrue wrote:
> >
> > I'm not interested. You guys wake me when SBCL actually works on
> > Windows.
>
> There you go again. But I'll let it pass.
Look, getting people to hunt through a newsgroup for a bunch of missing
steps is not build engineering. That's alpha testing and using the
people in the newsgroup as your guinea pigs. We don't even do that on
the Chicken mailing list really. Sure I rely on people to catch errors
in my code that I miss, but I make every effort to get it working
before I show it to anybody. I certainly don't put it out in front of
the world claiming it works, when I haven't provided anything
resembling a documented and reproducible build process.
"Works on my box" is not good enough. It has to work on most boxes
without spilling blood or incanting to the dead.
Cheers,
Brandon Van Every
From: Robert Uhl
Subject: Re: SBCL on Win32, was Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m3veo0jrgg.fsf@NOSPAMgmail.com>
Jack Unrue <·······@example.tld> writes:
>
> If you have problems building SBCL, there have already been articles
> in this group about common problems and solutions.
And in this case he has extensive experience getting another personal
computer Lisp to build on Windows...
--
Robert Uhl <http://public.xdi.org/=ruhl>
Try travelling from state to state in America without a driver's license
and proof of insurance, to be yielded up to the first uniformed
road-thug who demands it. --L. Neil Smith
On 5 Sep 2006 13:46:17 -0700, "Mallor" <···········@gmail.com> wrote:
>
> Jack Unrue wrote:
> > On 5 Sep 2006 03:08:06 -0700, "Mallor" <···········@gmail.com> wrote:
> > >
> > > For Windows game development, the first 2 do not work on Windows.
> >
> > I'm actually getting a little tired of reading this false claim
> > that SBCL doesn't work on Windows. It does work, well enough to
> > get started with; I was able to port Graphic-Forms and thus add
> > SBCL to my list of supported CL implementations. YMMV and maybe
> > it's current state doesn't qualify as "working" for you, but it
> > does for me.
> >
> > Now, if you want to talk about some annoying rough edges that
> > remain, OK let's talk. But such a discussion has to start with
> > the premise that in fact a lot of work went into getting SBCL
> > running as well as it does on Windows today.
>
> A year ago, when I made my committment to Chicken Scheme, SBCL most
> definitely didn't work on Windows. If it's actually ready for people
> to do real work with it now, it's very recent, like past 2 months.
> This is not "ready for prime time" stuff, you don't risk a business on
> such things. Microsoft engineers have an epithet for people who say
> things like you do. "Works on my box." Here's the matrix of what SBCL
> currently claims to support:
> http://sbcl.sourceforge.net/platform-table.html If you look at the
> file release page
> http://sourceforge.net/project/showfiles.php?group_id=1373 you will see
> binaries for Linux, Darwin, and FreeBSD, but no Windows. At a glance,
> why should I believe you? But I will download it and see for myself,
> just to rub it in that it's really not there yet.
I don't care if you believe me or not, I'm worried about others
who are getting the wrong impression based on misinformation.
--
Jack Unrue
Jack Unrue wrote:
> On 5 Sep 2006 13:46:17 -0700, "Mallor" <···········@gmail.com> wrote:
> >
> > Jack Unrue wrote:
> > > On 5 Sep 2006 03:08:06 -0700, "Mallor" <···········@gmail.com> wrote:
> > > >
> > > > For Windows game development, the first 2 do not work on Windows.
> > >
> > > I'm actually getting a little tired of reading this false claim
> > > that SBCL doesn't work on Windows. It does work, well enough to
> > > get started with; I was able to port Graphic-Forms and thus add
> > > SBCL to my list of supported CL implementations. YMMV and maybe
> > > it's current state doesn't qualify as "working" for you, but it
> > > does for me.
> > >
> > > Now, if you want to talk about some annoying rough edges that
> > > remain, OK let's talk. But such a discussion has to start with
> > > the premise that in fact a lot of work went into getting SBCL
> > > running as well as it does on Windows today.
> >
> > A year ago, when I made my committment to Chicken Scheme, SBCL most
> > definitely didn't work on Windows. If it's actually ready for people
> > to do real work with it now, it's very recent, like past 2 months.
> > This is not "ready for prime time" stuff, you don't risk a business on
> > such things. Microsoft engineers have an epithet for people who say
> > things like you do. "Works on my box." Here's the matrix of what SBCL
> > currently claims to support:
> > http://sbcl.sourceforge.net/platform-table.html If you look at the
> > file release page
> > http://sourceforge.net/project/showfiles.php?group_id=1373 you will see
> > binaries for Linux, Darwin, and FreeBSD, but no Windows. At a glance,
> > why should I believe you? But I will download it and see for myself,
> > just to rub it in that it's really not there yet.
>
> I don't care if you believe me or not, I'm worried about others
> who are getting the wrong impression based on misinformation.
The misinformation that you're espousing, is that SBCL is even beta
quality yet. It's not, it's alpha. Alphas can be said to be "working"
in some sense, but to bill them as good enough for everyone else to use
is irresponsible. Let alone in the original entrepreneurial context of
this thread, "Why isn't SBCL good enough for a business bootstrap?"
The SBCL authors aren't trying to pull any wool over anyone's eyes, so
you shouldn't either. I've now cranked up the last win32 binary,
released in 0.9.12. The installation wasn't automagical, but the
documentation on what to do was sufficient.
Brandon J. Van ·····@KAHLO /e/sbcl/0.9.12/bin
$ sbcl
fatal error encountered in SBCL pid 2232:
can't find core file
Brandon J. Van ·····@KAHLO /e/sbcl/0.9.12/bin
$ sbcl --core /e/sbcl/0.9.12/lib/sbcl/sbcl.core
This is SBCL 0.9.12, an implementation of ANSI Common Lisp.
More information about SBCL is available at <http://www.sbcl.org/>.
SBCL is free software, provided as is, with absolutely no warranty.
It is mostly in the public domain; some portions are provided under
BSD-style licenses. See the CREDITS and COPYING files in the
distribution for more information.
os_map: 3, 0x1000, 02000000, 0x1000.
os_map: 3, 0x2000, 05000000, 0x1000.
os_map: 3, 0x3000, 09000000, 0x15bb000.
This is experimental prerelease support for the Windows platform: use
at your own risk. "Your Kitten of Death awaits!"
*
Cheers,
Brandon Van Every
On 5 Sep 2006 14:41:55 -0700, "Mallor" <···········@gmail.com> wrote:
> The misinformation that you're espousing, is that SBCL is even beta
> quality yet. It's not, it's alpha.
I never said it was beta quality. I corrected your misinformation
that "it doesn't run."
--
Jack Unrue
"Mallor" <···········@gmail.com> writes:
> Ari Johnson wrote:
>> Pascal Bourguignon <···@informatimago.com> writes:
>> >
>> > You're missing the bootstrapping process. $3,900 can kill the
>> > bootstrapping. Of course, it doesn't matter for businessmen who can
>> > get VC to finance them at a minimum of $1,000,000, but if you're not
>> > so fortunate, you just avoid any unnecessary expense, until you don't
>> > need the VC anymore.
>>
>> And if you are bootstrapping, what is wrong with SBCL, OpenMCL, or
>> clisp?
>
> For Windows game development, the first 2 do not work on Windows. It
> is important to use stuff that actually works on your platform,
> particularly open source stuff. Nothing open source is going to work
> on Windows unless it is regularly stress tested. You simply can't take
> for granted that open source gives a hoot about Windows. Generally
> they don't, so if you're not constantly vetting your tools, you're
> going to have trouble. I have this debate with a friend of mine, who
> advocates developing on Linux and "porting once later," rather than all
> this blood spilling I've done with Chicken Scheme. I say he's not
> considering the realities of vetting tools on Windows. If I hadn't
> stepped into the breach 10 months ago, Chicken Scheme simply wouldn't
> be viable for Windows game development.
>
> CLisp works on Windows, but it is only an interpreter, and it has a C
> FFI specific to CLisp. Performance code is core to the kinds of games
> I wish to write; that is, 4X Turn Based Strategy in the mold of
> Civilization or Sid Meier's Alpha Centauri. Lotta AI number crunching
> in those. My original project vision was strongly 3D graphical as
> well, although I'm less interested in that lately. Anyways there are
> performance requirements. I could have gotten started with CLisp, but
> I would need C interfaces to OpenGL libaries, or C++ interfaces to
> off-the-shelf 3D engines, almost immediately. I didn't want to waste
> development effort on C/C++ FFIs that I was just going to throw out.
> I've said a number of times, there's nothing "Common" about Common Lisp
> with respect to C at all, let alone C++.
>
> Corman Lisp would have been a reasonable choice. A Corman-specific C
> FFI yet again, but at least for a Windows version of the game, I
> wouldn't have to drop Corman to ship it. I did feel, however, with C
> FFIs each being "a law unto themselves," that I was no better off than
> in the Scheme world and I should go back to it. So I did. Also at the
> time, the Corman community looked like it was on its last legs. But
> possibly the imminent death of Corman has been exaggerated on a number
> of occasions.
>
> I had some peripheral adoption considerations a year ago. I was trying
> to build an Eclipse based toolchain. That was to be the "slick for the
> biz types" world domination strategy. There were no Lisp plugins;
> there was the SchemeWay plugin. But it never worked for me, and
> eventually I gave up and went to XEmacs. Also I wanted to poison the
> minds of other game developers. They could try Corman for free, but I
> don't know that they'd stick with something they had to pay for. That
> probably wasn't the biggest issue though. They probably wouldn't stick
> with anything less than a fullblown development system that I had
> shipped a game with, that was proven in the marketplace.
>
>> Plus, if the idea is sufficiently good, you can find capital.
>
> Not really, in the creative industries.
>
>> There are actually projects that I don't have the time to do
>> personally but would gladly put up some capital to get done - there is
>> a definite market for software development if you spot the gaps or can
>> get financing from someone who has spotted the gaps.
>
> You know, I don't have time for all this mess. I don't know if
> everyone else in software just gets rich hanging out in big companies,
> sipping martinis with people who have got piles of money they want to
> give away, but I'm busy coding. I stay in the trenches trying to solve
> real problems. And unfortunately, it's easy for this to be profitless,
> I've found out. Even for the most simple survival jobhunting right
> now, I am constantly torn between the need to survive vs. my strong
> will / determination to kick the Chicken Scheme build out the door.
> The latter is certainly a more rewarding pursuit than the former,
> obviously. Now why is jobhunting so difficult? Because everyone wants
> C++, Java, C#, or SQL in this town, to judge by the ads. Not Lisp,
> Scheme, or CMake build systems.
>
> I did start a local organization, partly to try to foster business
> interests that are lacking in the region. That aspect of SeaFunc is a
> failure. It could succeed in the future, but at present, the failure
> is the complete lack of interest on anybody's part to lift so much as a
> finger for business issues. And the last guy who wanted to re-launch
> LispSea, with more of a business promotion focus, gave up in a huff
> when people weren't immediately jumping at the chance to pursue that
> agenda. For my part, I'll start making the rounds with the Washington
> Software Alliance when I'm no longer concerned with my most immediate
> survival. The enterpreneurial moneybags aren't coming to us, so I'll
> have to go to them.
I have to confess that I don't really know what your game project is
about (I'm sure you've explained it several thousand times in your
posts, so maybe I should have paid more attention), but I don't see
what the problem is.
Is this game for commercial sale? Are you writing it solo? Why are
commercial Lisps not an option (LispWorks and Allegro both have OpenGL
interfaces)?
Why is the game's language so important? As you note, performance is
crucial in game development. I'm sure you know that most of what
makes a game successful has to do with things like art, gameplay
design, music, etc.. It seems like taking care of these things ought
to be a higher priority than whether or not you use Lisp or Scheme.
Believe it or not, productive work can be done in straight C - indeed,
most games are in C (well, maybe C++). If performance is critical and
you agree that the choice of programming language is not really
central to the quality of a game, just get it done.
Alternatively, use Lisp for development only. Take advantage of
macros, interactive development, SLIME/your IDE of choice and get
something working. Then mutilate it and crush it until it fits into
the C model. OR - buy a commercial Lisp, get performance, a working
implementation, executable delivery. Are you against commercial
programs in general? Then write your game with commercial tools
(assuming this game is meant to be sold), make some money, and invest
that into improving open-source software.
Like Lars, I'm having trouble understanding what you want here. If
you truly want to write a game, please do it. Use C if you have to
(and use Lisp for initial development if you want). Buy a commercial
Lisp if you have to. What's holding you back and what's keeping you
on USENET posting about this game instead of just sitting down and
writing it?
On Tue, 05 Sep 2006 03:08:06 -0700, Mallor wrote:
>> Ari Johnson wrote:
>> And if you are bootstrapping, what is wrong with SBCL, OpenMCL, or
>> clisp?
>
> For Windows game development, the first 2 do not work on Windows. It is
> important to use stuff that actually works on your platform, particularly
> open source stuff. Nothing open source is going to work on Windows unless
> it is regularly stress tested. You simply can't take for granted that
> open source gives a hoot about Windows. Generally they don't, so if
> you're not constantly vetting your tools, you're going to have trouble. I
> have this debate with a friend of mine, who advocates developing on Linux
> and "porting once later," rather than all this blood spilling I've done
> with Chicken Scheme. I say he's not considering the realities of vetting
> tools on Windows. If I hadn't stepped into the breach 10 months ago,
> Chicken Scheme simply wouldn't be viable for Windows game development.
>
> CLisp works on Windows, but it is only an interpreter, and it has a C FFI
> specific to CLisp.
Someone has already pointed out that SBCL works on Windows. Even if it is
in alpha, it will not be when you are done with your stuff.
You have CFFI; a FFI specific to multiple implementations. What does
Scheme, or Chicken Scheme have that is not "Chicken Scheme-specific"? CFFI
is easy to use and has a great manual.
> Performance code is core to the kinds of games I wish
> to write; that is, 4X Turn Based Strategy in the mold of Civilization or
> Sid Meier's Alpha Centauri. Lotta AI number crunching in those. My
> original project vision was strongly 3D graphical as well, although I'm
> less interested in that lately. Anyways there are performance
> requirements.
*shrug* .. performance? go with SBCL then, or start with CLisp then port
later (remember CFFI?) if you want to do work on Windows right away. But
I think your friend is right; Linux is in many ways the best place to
do development IMHO.
> I could have gotten started with CLisp, but I would need C interfaces
> to OpenGL libaries,
Yeah, and you will not need this with Chicken Scheme?
Check out cl-opengl (portable between Lisps because of CFFI). By the time
you're done, you can port stuff over to SBCL when the win32-port is less
alpha and get performance that beats both CLisp, Chicken Scheme and even C
in some cases. Yes, I've seen multiple threads about SBCL beating C.
> or C++ interfaces to off-the-shelf 3D
> engines, almost immediately. I didn't want to waste development effort on
> C/C++ FFIs that I was just going to throw out.
I do not understand; would you not need to do this now also? +10 month
offset - but still the same? Skipping one step ahead and starting "almost
immediately" on what in the first case would be the second step (+10
months after first step) does not seem bad, but I might be missing
something ...
> I've said a number of times, there's nothing "Common" about Common Lisp
> with respect to C at all, let alone C++.
Scheme nor Chicken Scheme does not have a "common" FFI either; does Scheme
in general have something like CFFI?
Heck; I do not care what you do or use - but the things you say seem
strange and sometimes even false.
--
Lars Rune Nøstdal
http://lars.nostdal.org/
I agree. It seems like a waste of time to be messing around with
Chicken Scheme when SBCL will get there and he can start off with CLisp
now. The benefits of SBCL and CL(in general) seem self evident. He
could have divided his 10 months on working in CLisp and helping out
with the win32 port of SBCL.
Lars Rune Nøstdal wrote:
> On Tue, 05 Sep 2006 03:08:06 -0700, Mallor wrote:
> >> Ari Johnson wrote:
> >> And if you are bootstrapping, what is wrong with SBCL, OpenMCL, or
> >> clisp?
> >
> > For Windows game development, the first 2 do not work on Windows. It is
> > important to use stuff that actually works on your platform, particularly
> > open source stuff. Nothing open source is going to work on Windows unless
> > it is regularly stress tested. You simply can't take for granted that
> > open source gives a hoot about Windows. Generally they don't, so if
> > you're not constantly vetting your tools, you're going to have trouble. I
> > have this debate with a friend of mine, who advocates developing on Linux
> > and "porting once later," rather than all this blood spilling I've done
> > with Chicken Scheme. I say he's not considering the realities of vetting
> > tools on Windows. If I hadn't stepped into the breach 10 months ago,
> > Chicken Scheme simply wouldn't be viable for Windows game development.
> >
> > CLisp works on Windows, but it is only an interpreter, and it has a C FFI
> > specific to CLisp.
>
> Someone has already pointed out that SBCL works on Windows. Even if it is
> in alpha, it will not be when you are done with your stuff.
>
> You have CFFI; a FFI specific to multiple implementations. What does
> Scheme, or Chicken Scheme have that is not "Chicken Scheme-specific"? CFFI
> is easy to use and has a great manual.
>
> > Performance code is core to the kinds of games I wish
> > to write; that is, 4X Turn Based Strategy in the mold of Civilization or
> > Sid Meier's Alpha Centauri. Lotta AI number crunching in those. My
> > original project vision was strongly 3D graphical as well, although I'm
> > less interested in that lately. Anyways there are performance
> > requirements.
>
> *shrug* .. performance? go with SBCL then, or start with CLisp then port
> later (remember CFFI?) if you want to do work on Windows right away. But
> I think your friend is right; Linux is in many ways the best place to
> do development IMHO.
>
> > I could have gotten started with CLisp, but I would need C interfaces
> > to OpenGL libaries,
>
> Yeah, and you will not need this with Chicken Scheme?
>
> Check out cl-opengl (portable between Lisps because of CFFI). By the time
> you're done, you can port stuff over to SBCL when the win32-port is less
> alpha and get performance that beats both CLisp, Chicken Scheme and even C
> in some cases. Yes, I've seen multiple threads about SBCL beating C.
>
> > or C++ interfaces to off-the-shelf 3D
> > engines, almost immediately. I didn't want to waste development effort on
> > C/C++ FFIs that I was just going to throw out.
>
> I do not understand; would you not need to do this now also? +10 month
> offset - but still the same? Skipping one step ahead and starting "almost
> immediately" on what in the first case would be the second step (+10
> months after first step) does not seem bad, but I might be missing
> something ...
>
> > I've said a number of times, there's nothing "Common" about Common Lisp
> > with respect to C at all, let alone C++.
>
> Scheme nor Chicken Scheme does not have a "common" FFI either; does Scheme
> in general have something like CFFI?
>
> Heck; I do not care what you do or use - but the things you say seem
> strange and sometimes even false.
>
> --
> Lars Rune Nøstdal
> http://lars.nostdal.org/
Lars Rune Nøstdal wrote:
>
> Someone has already pointed out that SBCL works on Windows.
We will forever disagree on the definition of "works." As far as I'm
concerned, "alpha" means "not working well enough for public
distribution." And judging by the SBCL readmes, installation
instructions, and binary releases, they agree with me.
> Even if it is
> in alpha, it will not be when you are done with your stuff.
I am not entirely sure I believe that. I definitely wouldn't have
believed it 1 year ago.
> You have CFFI; a FFI specific to multiple implementations.
http://common-lisp.net/project/cffi/
[quote]
Implementation Status
CFFI is under heavy development, and not as mature as UFFI in terms of
API stability. See chapter 2 of the CFFI User Manual: "Implementation
Support."
[/quote]
In other words, "that way there be dragons." I'm not interested in
determining whether it's alpha quality or beta quality; it's definitely
not production quality.
> What does
> Scheme, or Chicken Scheme have that is not "Chicken Scheme-specific"?
Nothing. In the Scheme universe, you live and die by your
implementation. Chicken Scheme's C FFI is production quality.
> > I could have gotten started with CLisp, but I would need C interfaces
> > to OpenGL libaries, or C++ interfaces to off-the-shelf 3D
> > engines, almost immediately. I didn't want to waste development effort on
> > C/C++ FFIs that I was just going to throw out.
>
> I do not understand; would you not need to do this now also? +10 month
> offset - but still the same?
If I live and die by Chicken Scheme, I get to keep my code. No porting
required. Remember, I don't share your confidence in the portability
of CFFI.
Cheers,
Brandon Van Every
On Thu, 07 Sep 2006 15:51:54 -0700, Mallor wrote:
>
> Lars Rune Nøstdal wrote:
>>
>> Someone has already pointed out that SBCL works on Windows.
>
> We will forever disagree on the definition of "works." As far as I'm
> concerned, "alpha" means "not working well enough for public
> distribution."
I see; maybe you have been tainted by the "Release Candidate" of Vista. :D
>> You have CFFI; a FFI specific to multiple implementations.
>
> http://common-lisp.net/project/cffi/
> [quote]
> Implementation Status
> CFFI is under heavy development, and not as mature as UFFI in terms of API
> stability. See chapter 2 of the CFFI User Manual: "Implementation
> Support."
> [/quote]
>
> In other words, "that way there be dragons." I'm not interested in
> determining whether it's alpha quality or beta quality; it's definitely
> not production quality.
Naaaw; it seems very much API-stable. I do not know how many projects
currently use CFFI, but I think there are quite a few.
>> What does Scheme, or Chicken Scheme have that is not "Chicken
>> Scheme-specific"?
>
> Nothing. In the Scheme universe, you live and die by your
> implementation. Chicken Scheme's C FFI is production quality.
Right - you suddenly "turn around" and no longer care about the
portability of FFIs and things in general; what about SBCL's FFI then? Or
any Lisp?
>> > I could have gotten started with CLisp, but I would need C interfaces
>> > to OpenGL libaries, or C++ interfaces to off-the-shelf 3D engines,
>> > almost immediately. I didn't want to waste development effort on
>> > C/C++ FFIs that I was just going to throw out.
>>
>> I do not understand; would you not need to do this now also? +10 month
>> offset - but still the same?
>
> If I live and die by Chicken Scheme, I get to keep my code.
You sure do.
--
Lars Rune Nøstdal
http://lars.nostdal.org/
Lars Rune Nøstdal wrote:
> On Thu, 07 Sep 2006 15:51:54 -0700, Mallor wrote:
> >
> > Lars Rune Nøstdal wrote:
>
> >> What does Scheme, or Chicken Scheme have that is not "Chicken
> >> Scheme-specific"?
> >
> > Nothing. In the Scheme universe, you live and die by your
> > implementation. Chicken Scheme's C FFI is production quality.
>
> Right - you suddenly "turn around" and no longer care about the
> portability of FFIs and things in general; what about SBCL's FFI then? Or
> any Lisp?
There's nothing sudden about it. If you want to look up my personal
history of trying things out, it was:
- got fed up with C++
- boosted Python, got fed up with lack of business culture at the time
- boosted OCaml, got fed up with its C FFI
- went to Bigloo, got fed up with build issues and Emacs
- almost went to Common Lisp, realized not much worked on Windows that
I wanted to pay for. Note that Corman Lisp's support of CFFI is only
listed as partial today http://common-lisp.net/project/cffi/ and I made
my decision 1 year ago. Corman didn't and doesn't do UFFI at all.
http://uffi.b9.com/platforms.html SBCL didn't work at all on Windows,
it wasn't even alpha. It was clear that there was nothing "common"
about Common Lisp with respect to C FFIs on Windows. Each
implementation was a law unto itself, no different than the Scheme
universe. So...
- I went to Chicken Scheme and stayed there. It worked on Windows well
enough to be worth improving. I didn't really think it would take me
10 months when I got started, otherwise I might have chosen to live and
die by something else. But what's done is done, and the Windows
support is now first class. In the sense that all basic capabilities
are working. I'm not trying to say it's an alternative to .NET or has
oodles of Windows-centric libraries or anything like that.
All that CL portability stuff will be nice someday when the work has
actually been done on Windows. Let's see SBCL get to beta on Windows.
Then let's see CFFI get to beta on SBCL on Windows. Until then, I
really can't take your portability sales pitch seriously.
I'm also wondering why I'm wasting so many words in a thread that has
gotten so long. It's not like I'm waiting for a build to finish at the
moment. People either "get it" as to what constitutes good Windows
support, or they don't. I think I'm going to spend less time arguing
with people in comp.lang.lisp who aren't serious about Windows
development, and more time identifying and converting people who need
an open source language implementation that is well supported on
Windows.
Cheers,
Brandon Van Every
Pascal Bourguignon <···@informatimago.com> writes:
> Ari Johnson <·········@gmail.com> writes:
>
> > The point is simply this: if your business plan is made or broken by
> > $3,900 of software and you cannot cost-justify the man-hours of
> > implementing what you need using free software, then your problem is
> > not the software prices but instead is your business plan as a whole.
>
> You're missing the bootstrapping process. $3,900 can kill the
> bootstrapping.
If it can, then again, as Ari puts it: "your problem is not the
software prices but instead is your business plan as a whole." This
isn't rocket science.
/Jon
--
'j' - a n t h o n y at romeo/charley/november com
jayessay <······@foo.com> writes:
> Pascal Bourguignon <···@informatimago.com> writes:
>
>> Ari Johnson <·········@gmail.com> writes:
>>
>> > The point is simply this: if your business plan is made or broken by
>> > $3,900 of software and you cannot cost-justify the man-hours of
>> > implementing what you need using free software, then your problem is
>> > not the software prices but instead is your business plan as a whole.
>>
>> You're missing the bootstrapping process. $3,900 can kill the
>> bootstrapping.
>
> If it can, then again, as Ari puts it: "your problem is not the
> software prices but instead is your business plan as a whole." This
> isn't rocket science.
Onasis started with $0.10. He bought an apple. He spent the day
making it shine. Then he sold it for $0.20. The next day, he bought
two apples. A few years later he was milliardaire. Imagine he bought
instead a new clothe for $0.15 to make them shine better or faster
instead of two apples.
--
__Pascal Bourguignon__ http://www.informatimago.com/
"A TRUE Klingon warrior does not comment his code!"
Pascal Bourguignon wrote, On 7/9/06 7:55 AM:
> Onasis started with $0.10. He bought an apple. He spent the day
> making it shine. Then he sold it for $0.20. The next day, he bought
> two apples. A few years later he was milliardaire. Imagine he bought
> instead a new clothe for $0.15 to make them shine better or faster
> instead of two apples.
I don't understand... Are you talking about Aristotle Onassis? Here's an
excerpt I found from a biography, beginning with his leaving Smyrna
after the Turkish invasion:
"After this episode, Ari felt a sense of disillusionment in Europe and
headed for the new world and Argentina. There he started to work as a
telephone switchboard operator. This allowed him to listen in on
conversations and learn the language. He especially liked to listen in
to some calls between Argentina and New York. One of these yielded an
attractive deal where he made some money speculatively.
Ari saw an opportunity to import some of his father's tobacco products
and promptly organized it into a profitable business. When Greece would
later change their export regulations, Ari would become a consul for
Greece in Argentina (aka. a spy) so that he could continue the business
without the additional taxes and duties.
It was during his time in Argentina that Ari started to build his fleet.
His first purchase was six ships from a Canadian shipping company during
the Great Depression at scrap metal prices."
No mention there of a business plan that depends on exponential growth.
No mention of it taking an entire day to make an apple shine if you
don't have a $0.15 cloth to use as a tool.
sidney <······@sidney.com> writes:
> Pascal Bourguignon wrote, On 7/9/06 7:55 AM:
>> Onasis started with $0.10. He bought an apple. He spent the day
>> making it shine. Then he sold it for $0.20. The next day, he bought
>> two apples. A few years later he was milliardaire. Imagine he bought
>> instead a new clothe for $0.15 to make them shine better or faster
>> instead of two apples.
>
> I don't understand... Are you talking about Aristotle Onassis? Here's an
> excerpt I found from a biography, beginning with his leaving Smyrna
> after the Turkish invasion:
Perhaps it was Rockfeller. It doesn't matter, it's just the
archetipal milliardaire who made his fortune starting with 10 cents.
--
__Pascal Bourguignon__ http://www.informatimago.com/
"What is this talk of "release"? Klingons do not make software
"releases". Our software "escapes" leaving a bloody trail of
designers and quality assurance people in its wake."
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2fyf4pr8t.fsf@hermes.theari.com>
Pascal Bourguignon <···@informatimago.com> writes:
> sidney <······@sidney.com> writes:
>
>> Pascal Bourguignon wrote, On 7/9/06 7:55 AM:
>>> Onasis started with $0.10. He bought an apple. He spent the day
>>> making it shine. Then he sold it for $0.20. The next day, he bought
>>> two apples. A few years later he was milliardaire. Imagine he bought
>>> instead a new clothe for $0.15 to make them shine better or faster
>>> instead of two apples.
>>
>> I don't understand... Are you talking about Aristotle Onassis? Here's an
>> excerpt I found from a biography, beginning with his leaving Smyrna
>> after the Turkish invasion:
>
> Perhaps it was Rockfeller. It doesn't matter, it's just the
> archetipal milliardaire who made his fortune starting with 10 cents.
$3,900 is not 10 cents, but it's also not $10 million. Furthermore,
if he could have borrowed the 15 cents to get a better polishing cloth
and increased his productivity to the extent necessary to pay back the
15 cents plus interest, he should have bought the cloth. If you need
a $3,900 tool to be productive but cannot pay for it plus interest or
create your own tool to replace it in a cost-effective way, then you
need to consider alternatives to your current business plan that
incurs irrational (as in not financially justifiable) "needs."
By the way, if you can't convince anyone at all to lend you $3,900
after seeing your business plan, then the business-world consensus is
that your business plan sucks. If you disagree with the business
world, you'll have to make do without the $3,900 tool until you can
afford it. But seriously - anyone can borrow $3,900 if he is willing
to pay the interest rate necessary to offset the speculation-related
risk inherent in his plan to pay it back.
From: Thomas Lindgren
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <87k643y3tq.fsf@dev.null>
Ari Johnson <·········@gmail.com> writes:
> By the way, if you can't convince anyone at all to lend you $3,900
> after seeing your business plan, then the business-world consensus is
> that your business plan sucks. If you disagree with the business
> world, you'll have to make do without the $3,900 tool until you can
> afford it. But seriously - anyone can borrow $3,900 if he is willing
> to pay the interest rate necessary to offset the speculation-related
> risk inherent in his plan to pay it back.
The gentleman's solution is to max out one's credit card(s). Another
option is to do contracting work for a while first. For example, if
you can charge $100/hour, you can buy your $3900 Lisp (fully
deductible) after about a regular week's work. And while you're saving
up, you can always prototype things in a free Lisp.
Best,
Thomas
--
Thomas Lindgren
Pascal Bourguignon wrote, On 7/9/06 9:28 AM:
> Perhaps it was Rockfeller. It doesn't matter, it's just the
> archetipal milliardaire who made his fortune starting with 10 cents.
I still don't understand why he had to spend the entire day to get one
apple to shine. Clearly it would not have worked to buy a cloth for
$0.15 when he had turned that first $0.10 for an apple into $0.20. But
why didn't he pick up an open source cloth from the freeware bin, put in
the effort to clean it up and get the bugs out, and then use that to
polish apples better and quicker?
If one wasn't available, then he would have had to put the work into
doing without the tools until he sold two more apples, then spent the
$0.15 to let him proceed with his exponential growth. He could not have
sustained the growth in his business polishing just one apple per day.
Without the details, I can't tell if your parable is for or against
using free Lisp tools, or perhaps is just neutral to the topic.
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m27j0gpqfi.fsf@hermes.theari.com>
sidney <······@sidney.com> writes:
> Pascal Bourguignon wrote, On 7/9/06 9:28 AM:
>> Perhaps it was Rockfeller. It doesn't matter, it's just the
>> archetipal milliardaire who made his fortune starting with 10 cents.
>
> I still don't understand why he had to spend the entire day to get one
> apple to shine. Clearly it would not have worked to buy a cloth for
> $0.15 when he had turned that first $0.10 for an apple into $0.20. But
> why didn't he pick up an open source cloth from the freeware bin, put in
> the effort to clean it up and get the bugs out, and then use that to
> polish apples better and quicker?
He takes an entire day to shine one apple because he has no capital to
buy more apples to work on the first day. Otherwise it would only be
linear growth later on, when he can only polish one apple a day.
He wouldn't get a free cloth, though, because of the opportunity
costs. If it took him a week to turn a free cloth into the equivalent
of a 15-cent cloth, and he can polish apples as fast as he can afford
to buy them them with a cloth but can only polish at most one apple an
hour (or, working long hours, 10 a day) without the cloth:
(Note that he starts with 10 cents in his pocket, and for the free
cloth spends the first week working on the cloth instead of buying
apples; for the 15-cent cloth he waits until he can afford both the
cloth and one apple to get started, which is at the end of Day 2 when
he has sold two apples for 40 cents total and buys the cloth for 15
cents, leaving 25 cents to buy apples on Day 3. He takes a 5-cent
dividend for personal use on that day, as well.)
All figures are in apples, and it is assumed that he buys the lesser
of the number of apples he can afford and the number he can polish.
He buys apples at the end of the day to polish them on the following
day.
Day No cloth Free cloth 15-cent cloth
0 1 1 1
1 2 1 2
2 4 1 2
3 8 1 4
4 10 1 8
5 10 1 16
6 10 1 32
7 10 2 64
8 10 4 128
9 10 8 256
10 10 16 512
And that's why you buy the cloth. You will always be exponentially
ahead of the guy without one and a factor of 32 ahead of the guy who
took the cheap way out.
From: Ken Tilton
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <Y7JLg.78$8t.4@newsfe12.lga>
Ari Johnson wrote:
> sidney <······@sidney.com> writes:
>
>
>>Pascal Bourguignon wrote, On 7/9/06 9:28 AM:
>>
>>>Perhaps it was Rockfeller. It doesn't matter, it's just the
>>>archetipal milliardaire who made his fortune starting with 10 cents.
>>
>>I still don't understand why he had to spend the entire day to get one
>>apple to shine.
I have high hopes for any thread started by Ron, but with a nice juicy
argument from analogy I think the sky is really the limit on how
effective it can be at keeping you dorks from getting any work done and
competing with my Lisp start-up. Allow me to do my small part to fan the
flames...
So PB's brilliant idea is that I should have taken the code I had
written at the end of the first day and offered it for sale?
(That crumbling sound you hear is the analogy.*)
hth, kenny
* Windows 3.1 is acceptable as a counter-argument.
--
Cells: http://common-lisp.net/project/cells/
"I'll say I'm losing my grip, and it feels terrific."
-- Smiling husband to scowling wife, New Yorker cartoon
Ari Johnson wrote, On 7/9/06 10:42 AM:
> He wouldn't get a free cloth, though, because of the opportunity
> costs. If it took him a week to turn a free cloth into the equivalent
> of a 15-cent cloth
Unless he got it from the Free Softwear Foundation, who volunteer to
wash and donate clothing for poor people. The terms of the Garment
Public License would not have allowed him to, for example, appliqu� a
fancy design and sell it as designer clothing (you need the BVD license
for that, which is a lot briefer and more open), but did not prevent him
from cutting them up and using them as polishing cloths for his
proprietary Apple business. Which would, in fact, have allowed him to
hire apple polishers a little earlier in the process than if he would
first have had to earn the money to buy multiple $0.15 polishing cloths.
On the other hand, according to the chart you posted that would only
have made the difference of one day in his progress, not a make or break
thing either way. The critical point is how ready to use is the free
product, vs how much time would have to be spent removing the bugs, all
that expressed as opportunity costs, as you point out.
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2y7swo91q.fsf@hermes.theari.com>
sidney <······@sidney.com> writes:
> On the other hand, according to the chart you posted that would only
> have made the difference of one day in his progress, not a make or break
> thing either way. The critical point is how ready to use is the free
> product, vs how much time would have to be spent removing the bugs, all
> that expressed as opportunity costs, as you point out.
Precisely. You have to weigh all the alternatives available to you
and decide which one will maximize your profit (or minimize your risk
while still turning a profit, if you are risk averse). If you have a
free cloth available to you that does not require any work to start
using, then you will obviously choose that path. However, I think
that the original genesis of this "but Lispworks costs too much!"
thread was someone saying that none of the free alternatives have
feature X and someone telling him to buy Lispworks if he needs feature
X. If getting feature X to work on a free Lisp involves less total
cost (financial, opportunity, and risk) than buying Lispworks, then
that is the way to do it.
If I had the time and UI expertise (neither of which I do have), I
would buy a commercial Lisp and develop at least one of the
applications I have in mind which has a definite market. If I had the
capital, I would hire a few people to write it for me.
Actually, heck ...
Who here wants to become a financial success using Lisp? I want to
develop a cross-platform (at least Windows and Mac) client-server
(preferably both ends can run on each platform and are very easy to
install and maintain; a web client would also be nice although a
"real" client is mandatory) semi-connected (so that you can take the
client side on the road and synchronize over the Internet when you get
the chance, but work in real time when connected to the server
directly) application the details of which I will not share on Usenet.
I have the idea and the contacts to market the finished product, and I
know that the market is there. E-mail me your resume and choice of
development tools (so that I can factor that expense into any business
plan that develops). English proficiency, GUI experience, and
willingness to trade time for equity in the resulting company are
required.
sidney <······@sidney.com> writes:
> Pascal Bourguignon wrote, On 7/9/06 9:28 AM:
>> Perhaps it was Rockfeller. It doesn't matter, it's just the
>> archetipal milliardaire who made his fortune starting with 10 cents.
>
> I still don't understand why he had to spend the entire day to get one
> apple to shine. Clearly it would not have worked to buy a cloth for
> $0.15 when he had turned that first $0.10 for an apple into $0.20. But
> why didn't he pick up an open source cloth from the freeware bin, put in
> the effort to clean it up and get the bugs out, and then use that to
> polish apples better and quicker?
>
> If one wasn't available, then he would have had to put the work into
> doing without the tools until he sold two more apples, then spent the
> $0.15 to let him proceed with his exponential growth. He could not have
> sustained the growth in his business polishing just one apple per day.
>
> Without the details, I can't tell if your parable is for or against
> using free Lisp tools, or perhaps is just neutral to the topic.
Obviously, he used some available free cloth.
If he bought one professionnal apple-shining-cloth, he wouldn't be a legend now.
--
__Pascal Bourguignon__ http://www.informatimago.com/
"Specifications are for the weak and timid!"
Ari Johnson wrote:
<snipped>
> The point is simply this: if your business plan is made or broken by
> $3,900 of software and you cannot cost-justify the man-hours of
> implementing what you need using free software, then your problem is
> not the software prices but instead is your business plan as a whole.
The thing is, where I am it would be cheaper to throw
more programmers at the problem than purchase the
cheapest commercial cross-platform lisp.
You have not got the same perspective as we (3rd-world)
do when looking at $3900.
Let me put it to you this way:
What did you pay for your house (the selling price
you bought at?)?
Now divide that figure by 12. Would you consider
*that* figure as excessive? Because that is what
it effectively costs *us* in the third world.
The biggest threat to the commercial lisp vendors
is not the free/opensource lisps, it's the leagues
of cheap developers "overseas" who might, out of
necessity, one day actually develop a lisp that
sells for a tenth of the price because their
costs are a tenth of the current vendors.
(Hey, that's actually quite a good idea, now that
I think about it - I know of no other language
where people *want* a cheap commercial implementation
rather than a free one).
--
goose
Have I offended you? Send flames to ····@localhost
real email: lelanthran at gmail dot com
website : www.lelanthran.com
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2fyf7899j.fsf@hermes.theari.com>
goose <··············@webmail.co.za> writes:
> Ari Johnson wrote:
>
> <snipped>
>
>> The point is simply this: if your business plan is made or broken by
>> $3,900 of software and you cannot cost-justify the man-hours of
>> implementing what you need using free software, then your problem is
>> not the software prices but instead is your business plan as a whole.
>
> The thing is, where I am it would be cheaper to throw
> more programmers at the problem than purchase the
> cheapest commercial cross-platform lisp.
In that case, look through my other posts and you will see that I have
made it explicit that, in the event it is more cost-justifiable to do
the missing bits with a free Lisp than to purchase a commercial Lisp,
your decision is clear.
I personally have never paid for a Lisp. I cannot rationally justify
the purchase for my own purposes. Each person should look at the
costs and benefits of each of his alternatives and choose the best one
for him.
goose wrote:
>
> The biggest threat to the commercial lisp vendors
> is not the free/opensource lisps, it's the leagues
> of cheap developers "overseas" who might, out of
> necessity, one day actually develop a lisp that
> sells for a tenth of the price because their
> costs are a tenth of the current vendors.
>
> (Hey, that's actually quite a good idea, now that
> I think about it - I know of no other language
> where people *want* a cheap commercial implementation
> rather than a free one).
I think that would be a great thing to happen! I don't think a Lisp
system would be a very good target as a business proposition though.
Develop some kind of enterprise system first, where there's a
significant market with lots of money to spend. But by all means, if
you could develop a system with a comparable feature set & quality of
LW or Allegro which sells for a tenth the price, I'd buy licenses for
several platforms!
--tim
Tim Bradshaw wrote:
> goose wrote:
<snipped>
>>
>>(Hey, that's actually quite a good idea, now that
>>I think about it - I know of no other language
>>where people *want* a cheap commercial implementation
>>rather than a free one).
>
>
> I think that would be a great thing to happen! I don't think a Lisp
> system would be a very good target as a business proposition though.
> Develop some kind of enterprise system first, where there's a
> significant market with lots of money to spend. But by all means, if
> you could develop a system with a comparable feature set & quality of
> LW or Allegro which sells for a tenth the price, I'd buy licenses for
> several platforms!
Well, the tedious work in developing an ANSI Common Lisp
implementation is in the elbow grease required to write
all the libraries (CLOS, loop, etc).
I suspect that once one has a decent lisp
runtime minus library functions[1], then one has
completed about 5% of an ANSI Common Lisp.
I think that if I can complete that step, I'll be
able to do the rest *mostly* in parallel (hire
a ton of people just to write the ACL functions).
[1] By this I mean the ability to read, parse into
a tree (and expand macros while doing so) and traverse
a lisp program *and* do function lookup, defun, etc.
The functions themselves will take a considerable amount
of time but that doesn't stop a runtime from having
a uniform ability to lookup functions, symbols, etc.
--
goose
Have I offended you? Send flames to ····@localhost
real email: lelanthran at gmail dot com
website : www.lelanthran.com
Ari Johnson <·········@gmail.com> writes:
> "Javier" <·······@gmail.com> writes:
>
>> Ari Johnson ha escrito:
>>
>>> "Javier" <·······@gmail.com> writes:
>>> > I've been not in software business, but in other areas and not of a
>>> > such technical level.
>>> > Elsewhere, I still think that if you are a freelance worker, or even
>>> > associating to another person, $7800 (for 2 seats) is big amount of
>>> > money, taking in consideration that there are free tools and that it is
>>> > not the only thing you have to buy. If you are starting with little
>>> > money (and most of starting business does), you are going to take care
>>> > of all the money you are going to spend.
>>>
>>> What other area of business have you been in that the tools of the
>>> trade were free?
>>
>> I owned a cybercafe and the operating system was Linux. It had 16 seats
>> and saved me a good amount of money in licensing, and a lot of
>> headaches with virus, reconfigurations and more. I give up the
>> business because of some problematic clients and to have a better job,
>> but not because it was not success economically speaking. I know it is
>> not at the technical level of a software company, but it allowed me to
>> realised what people want, what the costs are, and how to maximize it
>> while not being tied to a software company.
>
> The operating system was free and did everything that any other
> operating system would have in that application with not a lot of
> extra work on your part. How about all the other equipment? Were the
> chairs, tables, coffee equipment, building, and staff all free, as
> well?
>
> The point is simply this: if your business plan is made or broken by
> $3,900 of software and you cannot cost-justify the man-hours of
> implementing what you need using free software, then your problem is
> not the software prices but instead is your business plan as a whole.
++ !
From: Andreas
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m36uoF43h7uU2@uni-berlin.de>
Javier wrote:
> Ari Johnson ha escrito:
>
>
>>That's not what he meant. He meant that the cost of a developer is
>>that amount, not the salary. There are three general categories of
>>expenses in having an employee (salary/wages, benefits such as group
>>health insurance, and tools/workspace), and salary is generally the
>>smallest factor. For the 15,000 euros you are paid in a year, how
>>much more do you think it costs your employer to provide you with a
>>desk and either an office or cubicle, plus a computer and electricity
>>to run it, plus lavatory facilities, plus a parking space, plus
>>lighting, plus air conditioning, etc.? It is almost certainly more
>>than 15,000 euros.
>
>
> Sure, but I think that �100K is too far from �15k (almost 7x more),
> and don't think that having to pay everything you say will sum so much!
> At the most, perhaps �30k for me. Here we pay 200� for insurance,
> and 10% for wages every month. The rest are colateral costs, which are
> shared from all other employers in the same office, so this might be
> not so much in comparation.
According to these numbers, costs ARE around 5-7 times higher here.
But these numbers are hard to compare and moreover it is a well-known
secret that production in germany is not cheap.
Regards
AHz
Ari Johnson <·········@gmail.com> writes:
> "Javier" <·······@gmail.com> writes:
>
>> Andreas ha escrito:
>>> A software developer here (germany) is calculated with around
>>> 100k/year. This includes salary, employer's contribution and rents
>>> but no license costs.
>>
>> Oh my god!!! Please, look for a job in Germany for me, here in Spain we
>> earn 15.000/year!! (And only after working for 800/month for
>> several years until you've got enought experience!!!)
>> So you're right. In Germany it is possible to buy it. In Spain, it's
>> almost impossible if you are a normal guy (and I'm pretty sure that
>> this is the same in almost any country).
>
> That's not what he meant. He meant that the cost of a developer is
> that amount, not the salary. There are three general categories of
> expenses in having an employee (salary/wages, benefits such as group
> health insurance, and tools/workspace), and salary is generally the
> smallest factor. For the 15,000 euros you are paid in a year, how
> much more do you think it costs your employer to provide you with a
> desk and either an office or cubicle, plus a computer and electricity
> to run it, plus lavatory facilities, plus a parking space, plus
> lighting, plus air conditioning, etc.? It is almost certainly more
> than 15,000 euros.
As this is run off based to some extent on what I wrote, I'd like to
clarify what I meant - well at least try to.
The cost of something like LW is, in itself, not prohibitive. As
pointed out by others, there are many other costs involved, including
other staff, marketing and packaging, rent, insurance, etc etc which
are far more expensive and a bigger concern. We can even ignore other
posts on this group that have indicated PG didn't use a commercial
lisp because of licensing and cost issues (I make no claim as to the
accuracy of that - just seen it reported in this group).
The point I was trying to make is that if you wanted to get venture
capital to cover your startup costs until you started shipping the
product, using lisp was a problem because many venture capitalists
won't touch you (essentially because of FUD regarding lisp and
partially because their is a tendency to go with trends by venture
capitalists - they are actually fairly conservative in their thinking
and it takes a good sales person to convince them otherwise (I've not
met many good hackers who are also good sales people).
The fundamental requirement still remains having a good and usually
original idea. It also helps if the idea is one of those which others
can appreciate easily and quickly.
Finally, I still think there is an issue regarding free versus
commercial lisps. Kenny and others will say that unless your using the
commercial flavors, you really are just in the hobby category. He
probably has a valid point. However, you do have to ask the question,
"Why is this the case with CL and no other languages?". Given that CL
has the reputation of being X times more productive than other
languages, how come we don't have a well accepted set of standard
libraries for things that are frequently used/required in modern
development languages (ie my XML parsing reference in an earlier
post). I can get a C/C++ environment for free which is as productive
and powerful as the commercial versions, I can download Java, python,
perl and a number of sophisticated IDEs and I get a whole range of
standard, well documented, well supported and used libraries with next
to no financial risk.
I actually think I may have an answer to this question. Could it
possibly be that because CL is so powerful and you are so much more
productive with it, that people just roll their own libraries rather
than use someone elses? The "not built here" syndrome taken to its
extreme. I personally have found myself doing this and find that its
even worse because doing something like writing your own XML parser in
CL is not only faster than most other languages, its far more
rewarding. I actually get lots and lots of "cool moments" doing stuff
like this in CL, where in other languages, such as Java or C++, even
fairly trivial things seem to be a pain and quite tedious.
If this is the case, then we are really our own worst enemy!
Tim
--
tcross (at) rapttech dot com dot au
Tim X ha escrito:
> I actually get lots and lots of "cool moments" doing stuff
> like this in CL, where in other languages, such as Java or C++, even
> fairly trivial things seem to be a pain and quite tedious.
>
> If this is the case, then we are really our own worst enemy!
I understand what you want to say, and agree up to some point (just the
political one disturbs me, but it is not so important, really).
I'm in the point that I have totally resigned to program in anything
that I don't like, even if it does mean that I'm not going to work as a
programmer anymore.
Somewhere the last year, while looking for an open source project for
joining, I discovered Lisp and was impressed at its power.
Just for learning, I'm writing an object system (with no other ambition
as having some fun). One year ago could not even dream that modifying a
language would be so "easy".
I'm studing the posibility to join any project related to Lisp or to
start a new one, altought I don't feel prepared myself to program
serius code.
Yes, we are our own worst enemy, but who worries.
Javier wrote:
> Tim X ha escrito:
>
> > I actually get lots and lots of "cool moments" doing stuff
> > like this in CL, where in other languages, such as Java or C++, even
> > fairly trivial things seem to be a pain and quite tedious.
> >
> > If this is the case, then we are really our own worst enemy!
>
> I understand what you want to say, and agree up to some point (just the
> political one disturbs me, but it is not so important, really).
> I'm in the point that I have totally resigned to program in anything
> that I don't like, even if it does mean that I'm not going to work as a
> programmer anymore.
> Somewhere the last year, while looking for an open source project for
> joining, I discovered Lisp and was impressed at its power.
> Just for learning, I'm writing an object system (with no other ambition
> as having some fun). One year ago could not even dream that modifying a
> language would be so "easy".
> I'm studing the posibility to join any project related to Lisp or to
> start a new one, altought I don't feel prepared myself to program
> serius code.
>
> Yes, we are our own worst enemy, but who worries.
You're worried when your bank account is overdrawn, you're going to
food banks, you're in danger of eviction, and you have a dog to
support. Those realities can trigger some serious rethinking of how
you invest your time. Right now I'd be happy with any programming job
that's vaguely proximate to the skills I've developed over the past 14
years. Like, 3D graphics or build engineering or anything with an
interesting language or something scientifc or any old game or quite a
number of things really. What I'm not willing to do, is start
completely from scratch and dance to some monkey tune called out by
Redmond, doing the complete miserable shitwork of the computer industry
I've been trying to build a better mousetrap for. I'd sooner dig
ditches.
Cheers,
Brandon Van Every
Tim X wrote:
>
> Finally, I still think there is an issue regarding free versus
> commercial lisps. Kenny and others will say that unless your using the
> commercial flavors, you really are just in the hobby category. He
> probably has a valid point. However, you do have to ask the question,
> "Why is this the case with CL and no other languages?". Given that CL
> has the reputation of being X times more productive than other
> languages, how come we don't have a well accepted set of standard
> libraries for things that are frequently used/required in modern
> development languages (ie my XML parsing reference in an earlier
> post). I can get a C/C++ environment for free which is as productive
> and powerful as the commercial versions, I can download Java, python,
> perl and a number of sophisticated IDEs and I get a whole range of
> standard, well documented, well supported and used libraries with next
> to no financial risk.
>
> I actually think I may have an answer to this question. Could it
> possibly be that because CL is so powerful and you are so much more
> productive with it, that people just roll their own libraries rather
> than use someone elses?
Nope. It's because Lisp was used heavily in its heyday, and some
industries still have longstanding interest and experience with doing
things the Lisp way. Supply and demand has settled on fairly
exhorbitant tools to serve the needs of those industries. Those
industries are vertical markets, so off-the-shelf stuff isn't so
important. And when people around here have talked about how to
actually make money on Lisp development, they've mostly talked about
vertical markets / custom software.
Cheers,
Brandon Van Every
From: Andreas
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m36dcF488qfU1@uni-berlin.de>
Javier wrote:
> Andreas ha escrito:
>
>>A software developer here (germany) is calculated with around
>>�100k/year. This includes salary, employer's contribution and rents
>>but no license costs.
>
>
> Oh my god!!! Please, look for a job in Germany for me, here in Spain we
> earn �15.000/year!! (And only after working for �800/month for
> several years until you've got enought experience!!!)
In Germany the employers's contribution is very high (i.e. you have to
pay 50% of the cost for health insurance for your employees) so the
employee, after taxes etc, will see much less. Moreover living is
expensive so i don't think it's a big win for you to change over :-)
But due to EU regulations you are free to work in germany, too. Just
check the web for jobs (i.e. http://www.ingenieurweb.de)
> So you're right. In Germany it is possible to buy it. In Spain, it's
> almost impossible if you are a normal guy (and I'm pretty sure that
> this is the same in almost any country).
Sorry, wrong again. A very misleading idea is that the cost for software
is very high. Look at other bussiness areas. Compare the cost for LWPro
& a PC for developing software with the cost for an oven for a bakery
(u have bakers in spain ?)or a lifting ramp for a car repair shop (you
have car repair shops !). Walk around in your city and look what
equipment the different companies need to have. You will find that
software is not expensive.
Moreover, owners of car repair shops are not going to become rich :-)
Btw: What did you said is the TCO/year of your car ? ^^
> So now I understand why there are so much open source developers in
> Spain, and why you paid us so much to enter in the European Union.
No, thats because there is so nice weather for surfing in Tarifa :-)
> Unfortunaly, LW cost the same for us than for you. That's life.
>
Yes, of course. Because the bottom line for the price of a product
is normally the cost of development and maintainment (and the upper
line is the worth the product will have to you).
Regards
AHz
Andreas ha escrito:
> But due to EU regulations you are free to work in germany, too. Just
> check the web for jobs (i.e. http://www.ingenieurweb.de)
Naa, thanks, I thought it better and I prefer spanish sun and people,
even if a earn 2x less. :)
> Sorry, wrong again. A very misleading idea is that the cost for software
> is very high.
It is a cost, elsewhere, and the goal of any company is to minimize
costs and maximize sales.
> Look at other bussiness areas. Compare the cost for LWPro
> & a PC for developing software with the cost for an oven for a bakery
> (u have bakers in spain ?)
Ok, but I also can compare the cost of SBCL+PC and it is 3900 eur
cheaper than LW+PC.
This doesn't mean that if there is a feature of LW that SBCL doesn't
have, and I need, then I should consider 1) write that feature if it is
reasonable the amount of work required 2) if point 1 fails, buy LW.
But I don't need to go to point 2 if my goal is achieved at point 1.
And there are other goals, like being able to modify the source or
optimize the compiler if you need some more advanced control of your
program, and being able to distribute it open source for others to
modify your program, etc. (Yes, I know that you can distribute your
work even if using a comercial compiler, but people are not going to be
interested in contributing to it if they have to buy a dependant
comercial compiler to do it).
> Moreover, owners of car repair shops are not going to become rich :-)
Some of them are. :)
> Btw: What did you said is the TCO/year of your car ? ^^
Well, ok, but this comparison is not right. You may say that bread is
not expensive because it only cost 3 eur and LW 3900. And you may say
that bread is expensive because a pencil costs only 50 cents. But none
of them are comparable. They are expensive if you can find in other
place a bread which cost 75 cents, a pen which cost 25 cents, and a
Lisp compiler which cost 0 eur.
Particulary, It may even be more expensive if you have found a bug in
the comercial Lisp compiler and the sheller regrects or is inefficent
or takes too long to resolve that bug. With the free one, at least you
know that you can resolve it.
> > So now I understand why there are so much open source developers in
> > Spain, and why you paid us so much to enter in the European Union.
> No, thats because there is so nice weather for surfing in Tarifa :-)
That's nice, sun and waves in exchange for money. We've got plenty of
them. :)
From: Andreas
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m50btF4c0joU2@uni-berlin.de>
Javier wrote:
> Particulary, It may even be more expensive if you have found a bug in
> the comercial Lisp compiler and the sheller regrects or is inefficent
> or takes too long to resolve that bug. With the free one, at least you
> know that you can resolve it.
>
What you imply is that you are able to fix a bug i.e. in CMUCL's code
optimizer? Sorry, i seriously doubt that :-)
Moreover, all i have seen so far LispWorks is very fast in bug fixing
and you are able to buy support calls for whatever fix you want.
(I can't say how good Franz is here but i expect the same good service)
I absolutly agree that this is quite different with MS or Sun.
Regards
AHz
Andreas ha escrito:
> What you imply is that you are able to fix a bug i.e. in CMUCL's code
> optimizer? Sorry, i seriously doubt that :-)
You doubt well, but not everybody is a newbie like me.
> Moreover, all i have seen so far LispWorks is very fast in bug fixing
> and you are able to buy support calls for whatever fix you want.
> (I can't say how good Franz is here but i expect the same good service)
I don't know much about LW and Franz, so you're probably right.
> I absolutly agree that this is quite different with MS or Sun.
At least Sun is turning out into open source now.
From: Andreas
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m5bldF4i7ndU1@uni-berlin.de>
Javier wrote:
> Andreas ha escrito:
>
>
>>What you imply is that you are able to fix a bug i.e. in CMUCL's code
>>optimizer? Sorry, i seriously doubt that :-)
>
>
> You doubt well, but not everybody is a newbie like me.
Seems that i was really unclear here. Please see my answer to
Christophe's mail.
Regards
AHz
From: Christophe Rhodes
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <sqveo24luw.fsf@cam.ac.uk>
Andreas <···@snafu.de> writes:
> Javier wrote:
>> Particulary, It may even be more expensive if you have found a bug in
>> the comercial Lisp compiler and the sheller regrects or is inefficent
>> or takes too long to resolve that bug. With the free one, at least you
>> know that you can resolve it.
>>
> What you imply is that you are able to fix a bug i.e. in CMUCL's code
> optimizer? Sorry, i seriously doubt that :-)
Why? Many people with no previous CMUCL experience have submitted bug
fixes or patches, even to the code optimizer. Are you saying that
Javier is much stupider than all of those people?
Christophe
From: Andreas
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m56qqF4jhnjU1@uni-berlin.de>
Christophe Rhodes wrote:
> Andreas <···@snafu.de> writes:
>
>
>>Javier wrote:
>>
>>>Particulary, It may even be more expensive if you have found a bug in
>>>the comercial Lisp compiler and the sheller regrects or is inefficent
>>>or takes too long to resolve that bug. With the free one, at least you
>>>know that you can resolve it.
>>>
>>
>>What you imply is that you are able to fix a bug i.e. in CMUCL's code
>>optimizer? Sorry, i seriously doubt that :-)
>
>
> Why? Many people with no previous CMUCL experience have submitted bug
> fixes or patches, even to the code optimizer. Are you saying that
> Javier is much stupider than all of those people?
>
> Christophe
Obviously not.
I don't mean that he is unable to fix that bug in a technical sense
but in a comercial one.
He need to understand how the compiler works, then locate
the bug, fix the bug, test the fix, test if he breaks anything else and
finally publish the fix.
Assuming he is not an expert in compiler design with intimate knowledge
of CMUCL's IR, this will take a good amount of time.
And what i doubt that the cost of this time is below the cost for buying
the fix from LW (as chargable support call).
Please keep in mind that the time spended for bug-fixing the compiler is
not available for developing his product while all running costs do
still apply.
Sorry to be unclear here
Regards
AHz
"Javier" <·······@gmail.com> writes:
> Path: uni-berlin.de!fu-berlin.de!postnews.google.com!m73g2000cwd.googlegroups.com!not-for-mail
> From: "Javier" <·······@gmail.com>
> Newsgroups: comp.lang.lisp
> Subject: Re: Any Lisp financial success stories since Naughty Dog?
> Date: 4 Sep 2006 10:01:56 -0700
> Organization: http://groups.google.com
> Lines: 17
> Message-ID: <·······················@m73g2000cwd.googlegroups.com>
> References: <·······························@news.gha.chartermi.net>
> <······················@parsec.no-spoon.de>
> <··············@rowlf.interhack.net>
> <··············@lion.rapttech.com.au>
> <·············@newsfe08.lga>
> <························@b28g2000cwb.googlegroups.com>
> <··············@uni-berlin.de>
> <························@p79g2000cwp.googlegroups.com>
> <··············@uni-berlin.de>
> Mime-Version: 1.0
> Content-Type: text/plain; charset="iso-8859-1"
> Content-Transfer-Encoding: quoted-printable
> X-Trace: posting.google.com 1157389322 14943 127.0.0.1 (4 Sep 2006 17:02:02 GMT)
> X-Complaints-To: ············@google.com
> NNTP-Posting-Date: Mon, 4 Sep 2006 17:02:02 +0000 (UTC)
> In-Reply-To: <··············@uni-berlin.de>
> User-Agent: G2/0.2
> X-HTTP-UserAgent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; es-ES; rv:1.8.0.6) Gecko/20060728 Firefox/1.5.0.6,gzip(gfe),gzip(gfe)
> Complaints-To: ············@google.com
> Injection-Info: m73g2000cwd.googlegroups.com; posting-host=83.32.32.3;
> posting-account=wA24ug0AAACvCKJk2i67f5BnFA4OZ_yV
> Xref: uni-berlin.de comp.lang.lisp:211934
>
> Andreas ha escrito:
> > A software developer here (germany) is calculated with around
> > 100k/year. This includes salary, employer's contribution and rents
> > but no license costs.
Javier, please configure your software to send the right charset!
128 is not a ISO-8859-1 code for any character. It's a control code!
There is no euro character in ISO-8859-1. If you want to cater to
standards, use ISO-8859-15 or UTF-8 to get the euro character.
The code for euro € in ISO-8859-15 is 164, and in unicode it is 8364.
Also, take note that the euro sign must be written AFTER the amount,
not before. There are european norms about this...
--
__Pascal Bourguignon__ http://www.informatimago.com/
IMPORTANT NOTICE TO PURCHASERS: The entire physical universe,
including this product, may one day collapse back into an
infinitesimally small space. Should another universe subsequently
re-emerge, the existence of this product in that universe cannot be
guaranteed.
Pascal Bourguignon ha escrito:
> Javier, please configure your software to send the right charset!
I'm using Google Groups and this is not configurable, sorry.
(The only selection I can do is the language, and I have it set to
english).
"Javier" <·······@gmail.com> writes:
> Pascal Bourguignon ha escrito:
>
>> Javier, please configure your software to send the right charset!
>
> I'm using Google Groups and this is not configurable, sorry.
> (The only selection I can do is the language, and I have it set to
> english).
They write 100k EUR or 100k euro.
--
__Pascal Bourguignon__ http://www.informatimago.com/
In deep sleep hear sound,
Cat vomit hairball somewhere.
Will find in morning.
From: Espen Vestre
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m1zmden8fn.fsf@vestre.net>
"Javier" <·······@gmail.com> writes:
> Oh my god!!! Please, look for a job in Germany for me, here in Spain we
> earn [EUR] 15.000/year!! (And only after working for [EUR] 800/month for
> several years until you've got enought experience!!!)
Hmm. Are you sure about those numbers? I saw an ad for java
programmers in Spain recently, and they were offering at least 20K, I
think, with modest experience requirements.
An average salary of EUR 15000 would make you cheaper than Russia on
this list:
http://www.informationweek.com/blog/main/archives/2006/08/us_ranks_a_midd.html
Even India has skyrocketing salaries these days. Remember, the $9898
on this list is the average - I assume the better ones are far above
that level. In fact, a few years ago when the german government
thought they were going to attract indian programmers with special
work permissions, almost nobody came, the economic incentive simply
wasn't strong enough when they compared costs of living...
--
(espen)
Espen Vestre ha escrito:
> Hmm. Are you sure about those numbers? I saw an ad for java
> programmers in Spain recently, and they were offering at least 20K, I
> think, with modest experience requirements.
In Spain it is used to say salaries including insurance and wages, so
that 20K would be easyly 12-15k.
> An average salary of EUR 15000 would make you cheaper than Russia on
> this list:
> http://www.informationweek.com/blog/main/archives/2006/08/us_ranks_a_midd.html
The normal average is not the same as the real average. If 1 out of 10
programmers are earning 100k, but the rest ones 12K, this makes the
average quite little trustworthy.
>
> Even India has skyrocketing salaries these days. Remember, the $9898
> on this list is the average - I assume the better ones are far above
> that level
And the worst ones, which will with no doubt surprass in numbers, will
be inferior.
> In fact, a few years ago when the german government
> thought they were going to attract indian programmers with special
> work permissions, almost nobody came, the economic incentive simply
> wasn't strong enough when they compared costs of living...
Not really the costs of living... may be better explained as quality of
people (and I'm not saying that german people are bad, just talking
about a general sensation that have some foreign people living in that
country).
From: Rajappa Iyer
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <86mz9e8nj1.fsf@panix.com>
"Javier" <·······@gmail.com> writes:
> Espen Vestre ha escrito:
>> In fact, a few years ago when the german government
>> thought they were going to attract indian programmers with special
>> work permissions, almost nobody came, the economic incentive simply
>> wasn't strong enough when they compared costs of living...
> Not really the costs of living... may be better explained as quality of
> people (and I'm not saying that german people are bad, just talking
> about a general sensation that have some foreign people living in that
> country).
Neither reason, actually. The problem was that Germany offered a
temporary work permit with no possibility of immigration. This is
pretty unattractive proposition for someone especially compared to the
policies of Australia, NZ, Canada, US etc. "Kinder, not Inder" didn't
help either.
rsi
--
<···@panix.com> a.k.a. Rajappa Iyer.
Absinthe makes the tart grow fonder.
Rajappa Iyer <···@panix.com> wrote:
> Neither reason, actually. The problem was that Germany offered a
> temporary work permit with no possibility of immigration. This is
> pretty unattractive proposition for someone especially compared to the
> policies of Australia, NZ, Canada, US etc. "Kinder, not Inder" didn't
> help either.
And that everything was just a big lie to get some cheap
workers.
The first greencard owners became unemployed within a year.
Andreas wrote:
>
> A software developer here (germany) is calculated with around
> 100k/year. This includes salary, employer's contribution and rents
> but no license costs.
> This rough number may change, dependend on the region, by some 10%.
>
> He is used to work on around 200days/year (due to weekends, vacation,
> illness etc).
>
> So we end up with costs of 500/working day (!).
>
> And that's what you need to state in your bussiness plan (excepting
> you create a one-person-working-at-home-as-consulting company. Thats
> a different story. And you will often have no bussiness plan at all).
>
> Finally lets assume 3 developer (300k) and one sales person (100k too -
> because im too lazy to look it up).
> Then we are talking about 400k/year and you still have no secretary
> for the telephone nor support ppl or anything else.
> And you really care about $1100/seat for the software ?
>
> If you can speed up development by only 3 days because they uses LWPro
> then you have earned 400 (3*500-1100) per seat.
>
>
> I think that was it what Kenny wants you to say.
This is all true, but it's also the BIG CAPITAL way of doing a startup.
As opposed to the bootstrapping, self-reliant method of doing a
startup. If you go for BIG CAPITAL, then you're going to put a lot of
hours into chasing other people's money. Say, half your first year.
It took a friend of mine 1/4 of his year to chase after game
publishers, and that's already a pretty structured system of deals, so
I anticipate much worse for VC.
VCs are also going to limit what you can get funding to do. Maybe
you've got a great engineering idea that's definitely VC compatible.
Great, have at it. Those of us going after the more creative pursuits,
and trying to retain creative control of our IP, don't have much pull
with VCs. Especially since guys like John Romero fucked over the
industry once upon a time, in the days when publishers threw silly
money at guys with "star power," and they didn't deliver. So those
days are gone. People are primarily interested in paying the Chinese
to do canned things.
I'm sure indie film directors can relate. "Why don't you have millions
to make your movie? You're not a serious entrepreneur unless you're
already loaded and do everything with money."
Let's put it another way. The commercial practice of computers is
GODAWFUL EXPENSIVE. This is because the commercial practice of
computers is GODAWFUL BAD. And my jury is out on whether Lisp really
has anything to say about this.
Cheers,
Brandon Van Every
Andreas wrote:
> >
> Sorry Javier but Kenny is right.
> If you want to create a bussiness and $1100 for LWPro becomes a problem
> then there is something very wrong with you bussiness plan.
Why don't you multiply that per seat? How many bodies in your
business? How many other "oh it's only $1K" tools per seat?
I have a plan, a crazy plan. I believe that individuals should be able
to author commercial 3D games, without the ridiculous art teams and
production headaches common in industry today. It has turned out to be
more of a research plan than a business plan.
Cheers,
Brandon Van Every
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2hcznxmb5.fsf@hermes.theari.com>
"Mallor" <···········@gmail.com> writes:
> Andreas wrote:
>> >
>> Sorry Javier but Kenny is right.
>> If you want to create a bussiness and $1100 for LWPro becomes a problem
>> then there is something very wrong with you bussiness plan.
>
> Why don't you multiply that per seat? How many bodies in your
> business? How many other "oh it's only $1K" tools per seat?
>
> I have a plan, a crazy plan. I believe that individuals should be able
> to author commercial 3D games, without the ridiculous art teams and
> production headaches common in industry today. It has turned out to be
> more of a research plan than a business plan.
A few points...
1. If you are talking about "individuals," then you shouldn't multiply
things per seat.
2. The art teams are the expensive part, not the development tools.
You need to spend a lot of money on development tools before your
programming budget exceeds your art budget.
3. The vast majority of individuals are not capable of doing the work
of an art team. I am a very talented individual, but I am not capable
of it. Add to that the fact that a solid art budget and
reasonable-sized art team will actually get the content of your game
done in time to get to market before the whole thing is obsolete,
whereas you sitting at home and creating levels and so forth by
yourself would take too long to be useful, and you get to the
inescapable conclusion that, while individuals *can* author 3D games
without any help or money down beyond their Linux box and gcc, to do
so is analogous to single-handedly rowboating across the Atlantic.
From: Andreas Hinze
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m50atF4c0joU1@uni-berlin.de>
Ari Johnson wrote:
> 3. The vast majority of individuals are not capable of doing the work
> of an art team. I am a very talented individual, but I am not capable
> of it. Add to that the fact that a solid art budget and
> reasonable-sized art team will actually get the content of your game
> done in time to get to market before the whole thing is obsolete,
> whereas you sitting at home and creating levels and so forth by
> yourself would take too long to be useful, and you get to the
> inescapable conclusion that, while individuals *can* author 3D games
> without any help or money down beyond their Linux box and gcc, to do
> so is analogous to single-handedly rowboating across the Atlantic.
But, on the other side, there is a big comunity of indy developers that
work with small teams and have develop very nice games. Ok, nothing to
get rich but they earn enough to pay the tools (the cheap ones :-)
You can have have a working low-end toolset for around 1k-2k to make
some cute thinks :-)
Have a look at www.garagegames.com
Regards
AHz
From: Andreas
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m50c6F4c0joU3@uni-berlin.de>
Mallor wrote:
>
> I have a plan, a crazy plan. I believe that individuals should be able
> to author commercial 3D games, without the ridiculous art teams and
> production headaches common in industry today. It has turned out to be
> more of a research plan than a business plan.
>
I never meet ppl that are so good artists to make scenes and chars and
at the same time are so experienced programmer that they are able to add
i.e. a particle effect class to a scene-graph. So, for me, game
developing involves at least two ppl (a artist and a programmer).
And this is only creating the game. No test, no story design, no sales.
Regards
AHz
Andreas wrote:
> Mallor wrote:
> >
> > I have a plan, a crazy plan. I believe that individuals should be able
> > to author commercial 3D games, without the ridiculous art teams and
> > production headaches common in industry today. It has turned out to be
> > more of a research plan than a business plan.
> >
> I never meet ppl that are so good artists to make scenes and chars and
> at the same time are so experienced programmer that they are able to add
> i.e. a particle effect class to a scene-graph.
I'm a perfectly good visual artist in traditional media. In a parallel
universe, Brandon is a starving artist, not a starving programmer.
Perhaps this is evidence of cosmic convergence, that I've found a way
to do the latter. ;-) Programming + visual arts is not unheard of
among people I've met. I do think programming + musical ability is
more common.
I'm no good at things like Maya. Not because I couldn't be, but
because such packages have enormous learning curves that I've been
unwilling to go up. I've got code to write. Similarly, I could have
been a much better painter by now if I had sunk the time into it. I've
got code to write. It is difficult to dance to the tune of 2 masters.
And I'm a skilled martial artist, although in recent years I've not
trained as intensely. Still I keep up with it. My last instructor was
equally good at martial arts and sculpture. Among creatively gifted
people, I think it is exceedingly common for them to be gifted at at
least 2 things.
The unknown experiment is how easy is it for me to write algorithms
that produce tons of visual art that's acceptable to consumers. If I
achieve it I'm sure you'll read about it. There are precursors, both
in algorithmically generated art and in the somewhat better known "Demo
Scene." The problem with the Demo Scene is it lacks a culture of
reproducible, large scale results, i.e. the Open Source mentality.
They're too interested in being 3133t c0d3r5. If you do things with
off-the-shelf stuff, you are treated with derision.
Will Wright of Maxis is going in the demo coder direction, and he's
chasing these content problems, for the same reason of "art is too
godawful expensive to produce." We are all looking forward to his
upcoming "Spore" to see what he's able to do. But he's doing it with
the full resources of EA behind him, so I seriously doubt he's going to
nail it down to the level of 1 solo coder. He will probably point the
way though. If he makes money, he'll create the business environment
that will accept the techniques in the game industry.
Cheers,
Brandon Van Every
Ken Tilton wrote:
>
> Buckets of venture capital to buy LW Pro? That simply makes no sense.
> You are confusing the silly hobbyists who hang around c.l.l with a
> serious entrepreneur, who will take a second mortgage, borrow from
> family, use $$$ from a severance package, sell some stuff on eBay, etc
> etc to get reasonably small amounts of initial capital.
And you are assuming someone has stuck it out in industry long enough
to amass significant wealth doing "the boring stuff." I declared
bankruptcy to the tune of $82K last year. That from $55K of credit
card debt, the rest was interest. House, what house? I was always
undercapitalized, but that's often reality when you strike out on your
own. I had / have my $1400 office chair, I had my $1500 trip to Game
Developer Conference 2002. I even paid for MSDN at $500/pop once upon
a time, and I've never pirated any software I actually use.
I've burned a lot of money and I see little point in this-or-that
pricey technology, in the absent of provables about what it'll do for
me. Most of computerdom is junk, and the question is often whether
you're stupid enough to pay for the junk (especially in Microsoft's
case) or just use free junk. I could see forking over a few $K for
Alias Maya or 3DStudio MAX, if one wanted to commit to the full 3D
animation learning curve, as there is a big market for those skills.
No such market for Common Lisp. You go with CL, then you're going on a
faith that it's gonna do magical things for you alone.
Cheers,
Brandon Van Every
Mallor ha escrito:
> You go with CL, then you're going on a
> faith that it's gonna do magical things for you alone.
They would better have some fun becoming fans of Harry Potter. :)
Mallor wrote:
> Ken Tilton wrote:
> >
> > Buckets of venture capital to buy LW Pro? That simply makes no sense.
> > You are confusing the silly hobbyists who hang around c.l.l with a
> > serious entrepreneur, who will take a second mortgage, borrow from
> > family, use $$$ from a severance package, sell some stuff on eBay, etc
> > etc to get reasonably small amounts of initial capital.
>
> And you are assuming someone has stuck it out in industry long enough
> to amass significant wealth doing "the boring stuff." I declared
> bankruptcy to the tune of $82K last year. That from $55K of credit
> card debt, the rest was interest. House, what house? I was always
> undercapitalized, but that's often reality when you strike out on your
> own. I had / have my $1400 office chair, I had my $1500 trip to Game
> Developer Conference 2002. I even paid for MSDN at $500/pop once upon
> a time, and I've never pirated any software I actually use.
>
> I've burned a lot of money and I see little point in this-or-that
> pricey technology, in the absent of provables about what it'll do for
> me. Most of computerdom is junk, and the question is often whether
> you're stupid enough to pay for the junk (especially in Microsoft's
> case) or just use free junk. I could see forking over a few $K for
> Alias Maya or 3DStudio MAX, if one wanted to commit to the full 3D
> animation learning curve, as there is a big market for those skills.
> No such market for Common Lisp. You go with CL, then you're going on a
> faith that it's gonna do magical things for you alone.
>
>
> Cheers,
> Brandon Van Every
Are you fighting an unwinnable battle? What kind of game are you
trying to make and why do you have to use scheme to do it? Are you
trying to write an engine in Scheme, or do you want to use Scheme as a
scripting language? It seems that you're so bogged down in the minor
detail of your style of programming language, that you're unable to see
the forest through the trees. Shouldn't you concentrate on getting a
game out there to bring in some revenue, rather than fighting with
Chicken Scheme forever in order to get it to work on windows the way
you want it to?
I'm not trying to bash you, I'm just curious about your situation
because I see you posting a lot about your financial woes, wanting to
do game development, and fighting with Chicken Scheme. What do you do
to put food on the table?
From: Ken Tilton
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <V25Lg.51$7U4.24@newsfe12.lga>
> Mallor wrote:
>
>>Ken Tilton wrote:
>>
>>>Buckets of venture capital to buy LW Pro? That simply makes no sense.
>>>You are confusing the silly hobbyists who hang around c.l.l with a
>>>serious entrepreneur, who will take a second mortgage, borrow from
>>>family, use $$$ from a severance package, sell some stuff on eBay, etc
>>>etc to get reasonably small amounts of initial capital.
>>
>>And you are assuming someone has stuck it out in industry long enough
>>to amass significant wealth doing "the boring stuff." I declared
>>bankruptcy to the tune of $82K last year. That from $55K of credit
>>card debt, the rest was interest. House, what house? I was always
>>undercapitalized, but that's often reality when you strike out on your
>>own. I had / have my $1400 office chair,...
:)
kt
--
Cells: http://common-lisp.net/project/cells/
"I'll say I'm losing my grip, and it feels terrific."
-- Smiling husband to scowling wife, New Yorker cartoon
From: Ari Johnson
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <m2lkozxmir.fsf@hermes.theari.com>
"Mallor" <···········@gmail.com> writes:
> I've burned a lot of money and I see little point in this-or-that
> pricey technology, in the absent of provables about what it'll do for
> me. Most of computerdom is junk, and the question is often whether
> you're stupid enough to pay for the junk (especially in Microsoft's
> case) or just use free junk. I could see forking over a few $K for
> Alias Maya or 3DStudio MAX, if one wanted to commit to the full 3D
> animation learning curve, as there is a big market for those skills.
> No such market for Common Lisp. You go with CL, then you're going on a
> faith that it's gonna do magical things for you alone.
If you are striking out on your own, having marketable skills comes in
second to being effective on your own. It's always good to have a
solid Plan B, but not at the expense of a solid Plan A. Paul Graham
is a wealthy man because Common Lisp was the right tool for him to put
his vision into motion with. It is not the right tool for every
problem or for every person, but if you have a good idea it is not
going to hold you back.
That isn't even to mention the fact that you can do nearly everything
with Common Lisp, including determine if it's the right tool for the
job or not, with a free implementation. You can't get a free
implementation of Maya to determine if it's right for you.
Ari Johnson wrote:
>
> That isn't even to mention the fact that you can do nearly everything
> with Common Lisp, including determine if it's the right tool for the
> job or not, with a free implementation. You can't get a free
> implementation of Maya to determine if it's right for you.
You can certainly get a pirated one, it's dead easy. I'm not the
slightest bit squeamish about pirating for educational purposes.
Nobody gets $1K..$4K of my money sight unseen! Alias, having
recognized this fact, actually puts out free Personal Learning
Editions. They do have a problem in that their C++ SDK isn't included,
so again, back to piracy.
The real barrier, however, was the huge learning curve. I wasn't gonna
do it. I'd download these things, try to look at them, and quickly say
"Jesus." These UIs have nothing to do with what I think an artistic
process should be like. In fact the only things I've ever liked at the
SIGGRAPHs and GDCs I've been to, have been the haptic VR displays of
one kind or another. A lot of this stuff, we should be sculpting in
the air, or on a resistant surface. Actually that's what a lot of
artists do in practice, they scan clay models. I've lost track of
whether the price of 3D digitizers has come down. They used to be
pretty pricey.
Even sculpting in clay takes too long though. I need a lot less anal
specification and a lot more visual product. I wonder at what point
will algorithms achieve economies of scale over building things
manually.
Cheers,
Brandon Van Every
From: Pascal Costanza
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <4m25fqF446b7U1@individual.net>
Tim X wrote:
> Matt Curtin <········@interhack.net> writes:
>
>> Stefan Scholl <······@no-spoon.de> writes:
>>
>>> Why are so many Lisp people so obsessed about getting rich?
>> We often say that Lisp is so much more productive than other
>> languages. In a commercial context, where we're getting paid for
>> developing products or enabling services, this should translate into
>> significantly greater returns than on systems developed with more
>> common languages.
>>
>> PG's ViaWeb story is a good example of what should be generally taking
>> place among Lispers if we really are that much better.
>>
>> The fact is that even using computers is about economics, speeding
>> things up, doing things more cheaply than by hand or with some other
>> machines.
>>
>
> Although I don't disagree, I think the often overlooked critical
> ingrediant is actually having the right idea - the implementation
> language can make it easier to get that idea up and profitable, but
> you have to have the idea.
>
> Putting it another way, I suspect PG would have been successful with
> viaweb even if he had used another implementation language. certainly,
> he might have had to do some things a little differently and maybe it
> would have been a bit harder to have the flexibility he feels was a
> big part of the secret to the success, but I believe a talented
> development team with a good idea would probably have pulled it off
> using another language just as successfully.
>
> Another factor which I think is relevant is that lisp, while one of
> the easiest languages I've tried to learn, is also one of the hardest
> languages I've tried to master (and one of the most enjoyable because
> I seem to learn something new every day). Of course, this progress is
> artificially delayed by the fact I can only practice at home in my own
> time - unlike many other languages I've learnt, where I was paid and
> able to spend full-time getting to know the language.
>
> finally, although many get upset at the following claim, I still think
> it holds - most of the currently popular languages have lots of
> libraries available for doing what is common in many applications
> these days, making it fast to develop at least the prototype. CL is
> lacking in some of these - while it is true that for most things you
> can find a CL library if you search enough, you have to do the
> searching and often you have to evaluate different solutions to find
> the one you want. for example, parsing XML, something which is often
> remarked upon as being natural for lisp, but there is no standard
> library for it like there is in perl, java or python. Worse yet, when
> you go searching, you find quite a few and have to spend some time
> working out which one has the right balance of features and complexity
> for what you need. then you try it out and find some minor issues and
> have to go back and reevaluate.
>
> Some will argue the solution is to use one of the commercial
> implementations and maybe they are right. However, compared to perl,
> java and python, this is likely to be quite expensive (and since you
> are unlikely to get buckets of venture capital to use because you are
> implementing using lisp, money is an issue right now).
You have to invest some money sooner or later. Until then, you can get
quite far with the free/trial/personal editions of the commercial
implementations, or negotiate an evaluation period with the vendor. If
you have a reasonable plan, they should probably go with this, because
they are interested in new customers, and not in making everyone
sticking to each and every word of their licenses. (People seem to
forget that we're not dealing with Microsoft, Sun or IBM here. The Lisp
vendors are much more flexible than them.)
Apart from that, I think you're right on the spot: Having a good idea is
much more important than having the right tools.
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/
Ron Garret wrote:
> I concede (happily) that Naughty Dog is a Lisp success story. So now I
> have two in the last decade. Are there any others that I've missed?
>
> By "success story" I mean that 1) someone made a significant amount of
> money (more than just a salary) and 2) the principals attribute their
> success at least in part to the use of Lisp.
You know, if you look at Franz's http://www.franz.com/success/ , you'll
see they bill these successes as "Recent Stories." Well, some are
recent, but the Naughty Dog article is from 2001. "Recent" apparently
means 5 years to Franz and the Lisp world. Whereas by anyone else's
standards, that's the stone ages. I should know, I still develop on
2001 era hardware. Consider that Ruby On Rails didn't even exist.
Cheers,
Brandon Van Every
From: David Steuber
Subject: Re: Any Lisp financial success stories since Naughty Dog?
Date:
Message-ID: <87fyf7pfzw.fsf@david-steuber.com>
Ron Garret <·········@flownet.com> writes:
> Is ITA turning a profit?
Last summer they had a big campaign going to hire more developers. I
imagine they are doing alright.
--
This post uses 100% post consumer electrons and 100% virgin photons.
At 2.6 miles per minute, you don't really have time to get bored.
--- Pete Roehling on rec.motorcycles
I bump into a lot of veteran riders in my travels.
--- David Hough: Proficient Motorcycling
As the only person who actually answered Ron's actual question, and the
only person(as far as I know) reading (or at least discussing) this who
actually made significant money on Lisp [I don't think that Ron's money
came through Lisp, although I could be wrong], let me claim some small
authority and try to bring this around to a higher level. Maybe it's
the wrong question. (Someone else said this, but I'm going to repeat it
in a little more detail.) Maybe the question shouldn't be: How many
people made money on Lisp? But, rather something like: What is the
success ratio of people who seriously try to make money on Lisp and
fail, v. some other language. Of course, this isn't any easier to
quantify, and may indeed be more difficult, but given the evidence so
far, we have 3/1 for lisp: Graham, Afferent, and ND tried and won;
Reddit tried and Lost (actually, as far as I know they haven't made any
money yet, and it's not likely that they will, so the jury is still out
on them, and whether you count them as having actually tried with lisp
is also questionable since they changed horses in mid stream). Now,
given the uncountably large number of businesses that TRY to make money
on something OTHER than lisp and FAIL, I'm just guessing here, but the
ratio is probably much smaller than 3/1. Of course, if we could
actually count, I'm sure that the 3/1 would come down significantly,
ineed, my guess is that, as someone else said here, Lisp isn't the
issue. Yet anyway. Again, I htink that it WILL be in the future, but
that's another thread! :-)
From: Pascal Costanza
Subject: Re: From Counts to Ratios (Was: Any Lisp financial success stories since Naughty Dog?)
Date:
Message-ID: <4m8m7nF505b3U1@individual.net>
········@gmail.com wrote:
> As the only person who actually answered Ron's actual question, and the
> only person(as far as I know) reading (or at least discussing) this who
> actually made significant money on Lisp [I don't think that Ron's money
> came through Lisp, although I could be wrong], let me claim some small
> authority and try to bring this around to a higher level. Maybe it's
> the wrong question. (Someone else said this, but I'm going to repeat it
> in a little more detail.) Maybe the question shouldn't be: How many
> people made money on Lisp? But, rather something like: What is the
> success ratio of people who seriously try to make money on Lisp and
> fail, v. some other language. Of course, this isn't any easier to
> quantify, and may indeed be more difficult, but given the evidence so
> far, we have 3/1 for lisp: Graham, Afferent, and ND tried and won;
> Reddit tried and Lost (actually, as far as I know they haven't made any
> money yet, and it's not likely that they will, so the jury is still out
> on them, and whether you count them as having actually tried with lisp
> is also questionable since they changed horses in mid stream). Now,
> given the uncountably large number of businesses that TRY to make money
> on something OTHER than lisp and FAIL, I'm just guessing here, but the
> ratio is probably much smaller than 3/1. Of course, if we could
> actually count, I'm sure that the 3/1 would come down significantly,
> ineed, my guess is that, as someone else said here, Lisp isn't the
> issue. Yet anyway. Again, I htink that it WILL be in the future, but
> that's another thread! :-)
I doubt that you get very far in this forum no matter how you pose the
question. The reason is that comp.lang.lisp only represents a very small
fraction of the actual Lisp community. [1]
About two years ago, Arthur Lemmens has created a database of Lisp
users, and has collected over 200 entries in a relatively short amount
of time. The database still exists, as far as I know, and it may be
worthwhile to update it. It could then be used to distribute a
questionnaire and get some more reliable data. For example, this could
be something that the ALU could do for the Lisp community, or so. Just
an idea.
Pascal
[1] I am using the term "community" in a very broad sense. Some doubt
that there even is a community.
--
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 wrote:
>
> I doubt that you get very far in this forum no matter how you pose the
> question. The reason is that comp.lang.lisp only represents a very small
> fraction of the actual Lisp community. [1]
>
> About two years ago, Arthur Lemmens has created a database of Lisp
> users, and has collected over 200 entries in a relatively short amount
> of time. The database still exists, as far as I know, and it may be
> worthwhile to update it. It could then be used to distribute a
> questionnaire and get some more reliable data. For example, this could
> be something that the ALU could do for the Lisp community, or so. Just
> an idea.
Are you going to show up at the ALU and volunteer to do it then? Or
say you'll do it if they'll pay you to do it or something like that?
Ideas like this are nice, but they require leadership. The reason that
database isn't useful is because nobody took leadership, they didn't
keep it alive. As it stands, it paints Lisp as bad and useless,
something dead that only 200 souls on the ENTIRE INTERNET saw fit to go
on record as caring about, at one brief moment in history.
I'm saying if you think it's a good idea, put your own back into it.
I put my back into Chicken Scheme.
Cheers,
Brandon Van Every
Mallor wrote:
> Pascal Costanza wrote:
>> I doubt that you get very far in this forum no matter how you pose the
>> question. The reason is that comp.lang.lisp only represents a very small
>> fraction of the actual Lisp community. [1]
>>
>> About two years ago, Arthur Lemmens has created a database of Lisp
>> users, and has collected over 200 entries in a relatively short amount
>> of time. The database still exists, as far as I know, and it may be
>> worthwhile to update it. It could then be used to distribute a
>> questionnaire and get some more reliable data. For example, this could
>> be something that the ALU could do for the Lisp community, or so. Just
>> an idea.
>
> Are you going to show up at the ALU and volunteer to do it then? Or
> say you'll do it if they'll pay you to do it or something like that?
No, I am already doing other stuff.
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 wrote:
> Mallor wrote:
> > Pascal Costanza wrote:
> >> I doubt that you get very far in this forum no matter how you pose the
> >> question. The reason is that comp.lang.lisp only represents a very small
> >> fraction of the actual Lisp community. [1]
> >>
> >> About two years ago, Arthur Lemmens has created a database of Lisp
> >> users, and has collected over 200 entries in a relatively short amount
> >> of time. The database still exists, as far as I know, and it may be
> >> worthwhile to update it. It could then be used to distribute a
> >> questionnaire and get some more reliable data. For example, this could
> >> be something that the ALU could do for the Lisp community, or so. Just
> >> an idea.
> >
> > Are you going to show up at the ALU and volunteer to do it then? Or
> > say you'll do it if they'll pay you to do it or something like that?
>
> No, I am already doing other stuff.
That's pretty much the reason that database projects like that don't go
anywhere. You've got something more important to do, I've got
something more important to do, so does the next guy....
Cheers,
Brandon Van Every
On Thu, 07 Sep 2006 15:55:07 -0700, Mallor wrote:
>
> Pascal Costanza wrote:
>> Mallor wrote:
>> > Pascal Costanza wrote:
>> >> I doubt that you get very far in this forum no matter how you pose
>> >> the question. The reason is that comp.lang.lisp only represents a
>> >> very small fraction of the actual Lisp community. [1]
>> >>
>> >> About two years ago, Arthur Lemmens has created a database of Lisp
>> >> users, and has collected over 200 entries in a relatively short
>> >> amount of time. The database still exists, as far as I know, and it
>> >> may be worthwhile to update it. It could then be used to distribute a
>> >> questionnaire and get some more reliable data. For example, this
>> >> could be something that the ALU could do for the Lisp community, or
>> >> so. Just an idea.
>> >
>> > Are you going to show up at the ALU and volunteer to do it then? Or
>> > say you'll do it if they'll pay you to do it or something like that?
>>
>> No, I am already doing other stuff.
>
> That's pretty much the reason that database projects like that don't go
> anywhere. You've got something more important to do, I've got something
> more important to do, so does the next guy....
>
>
> Cheers,
> Brandon Van Every
What about a wiki-thing; let the people do most of the work themselves:
http://www.cl-user.net/asp/tags/11006
(maybe I should get myself added there .. O_o)
--
Lars Rune Nøstdal
http://lars.nostdal.org/
Lars Rune Nøstdal wrote:
> On Thu, 07 Sep 2006 15:55:07 -0700, Mallor wrote:
> > Pascal Costanza wrote:
> >>
> >> No, I am already doing other stuff.
> >
> > That's pretty much the reason that database projects like that don't go
> > anywhere. You've got something more important to do, I've got something
> > more important to do, so does the next guy....
>
> What about a wiki-thing; let the people do most of the work themselves:
> http://www.cl-user.net/asp/tags/11006
>
> (maybe I should get myself added there .. O_o)
Anything like the "Lisp Users Questionnaire" has a basic problem of
utility. I did some thinking on this problem for the would-be LispSea
chapter.
http://common-lisp.net/pipermail/seattle/2006-June/000034.html
My thoughts on the matter went unanswered. I'll repost them here.
[ahem]
As I've talked, I've realized I have 10..15 different personal
interests of relevance to LispSea. And that's not even counting all the
ones of a purely social value, such as sociocultural anthropology,
world history, martial arts, film, etc. A group such as LispSea is not
just a technical organization, it is also a social one, so I don't see
the listing of such interests as out of bounds. SeaFunc meets over beer
for a reason, after all.
It occurs to me, however, that my list of interests gets complicated.
Let's say all the other active LispSea participants are equally
complicated. How can all these interests be made accessible?
One way is socialization. People blab at each other, people find out
about their common interests. That's largely the format that SeaFunc
has followed. Although I would note, significantly, that I have not
personally engaged in any social activities with anyone from SeaFunc,
outside of SeaFunc. Or MLSIG-Seattle before that. The accident of
personal chemistry? Or geography? Or available time? Anyways, it says
to me that the meme of socialization is not fully utilized. On the
other hand, I have a pretty good mental inventory of most of the
regulars' technical interests. So it's not under-utilized either. I
couldn't tell you much about people who just showed up a few times
though.
So, socialization is not entirely sticky. It requires the ongoing
ritual of showing up to a meeting, and furthermore actively talking to
people. Not difficult at SeaFunc since that's all we do, but it will be
more of a challenge for LispSea, since we're aiming at more lecture
formats.
A bio page is the quick way to get information about 1 person. It's
not, however, the quick way to get information about 300 people. My
main objection to looking at the Lisp Questionnaire is, "You expect me
to look manually through 230 people in hopes I find something?" No way,
man.
[Edit: furthermore, consider the implications for the entire internet,
as opposed to a local interest group.]
How do I search other people's interests, in a painless way? Tooltips
spring to mind. I crank up the website, there's a randomly featured
member in a sidebar. A few of his major interests are given. I suppose
the definition of interests is the whim of the database designer, so
that people can have checkboxes to click on, and they aren't required
to play "guess the term." I suppose we'd need ongoing maintenance by
humans, designing new and useful categories according to user feedback.
But I also wonder if Lisp can be used to construct some categories out
of thin air. That would be damn useful, because the data could
potentially be extracted from any old crapola a Lisper was willing to
write, and database designers might need to invent less and do less
maintenance.
Emphasis on "might." I'd happily start with augmented manual labor
rather than AI or heavy duty text swizzling.
How do you even get a Lisper to write some sample crapola, so that
there's even something to sample? No participation, no data! Eliza
springs to mind. If the extraction of user data is implemented as a
silly game, perhaps users would bite the lure.
So now that I've laid out 1 year of R&D for someone, and the basis for
an internet startup ala Meetup.com... anyone see the quick and dirty
version of any of this?
Cheers,
Brandon Van Every