One of the things that I find most attractive about Lisp is an attitude,
perhaps best exemplified by Steele and Sussman, of "If you don't love
the Lisp you're with, write the Lisp you love."
Few language communities have the self-confidence (or, admittedly, a
language with the expressive power) to say "Hey, don't like our syntax?
Change the lexer to something you're more comfortable with! Don't like
the semantics? Use our language as a base for your own! (Act now and the
compiler is yours to keep.)"
Granted, sexpr addicts do gripe that alternate syntaxes have been tried
and found wanting, but there are precedents such as infix syntax
packages, and the language clearly provides for such play. Ask the
greybeards if a local maximum isn't just an upside-down way of looking
at the rut they're stuck in.
Now, here's what's bothering me. Looking down the language power curve,
there's all kinds of experimentation going on. Much of it, to the
enlightened denizens of this newsgroup, looks like a complete waste of
time. Nonetheless, many people, whether motivated by vanity, unique
needs or the simple love of novelty, take matters into their own hands
and play at Language Architect. But here, everybody plays at "Waiting
for Arc."
Let me be perfectly clear: Were Mr. Graham to contribute nothing further
to Lisp, his place in the Pantheon would be assured. Further, it
certainly cannot be said that Mr. Graham has been anything but
forthright in his description of the schedule for Arc.
The older generation has turned to threats of violence, bribery, funding
by subscription, and ceaseless imploring. The new converts to Lisp pick
up the tune, though with more deference -- their elders await Arc with
eager anticipation, so they do likewise.
Meanwhile, Lisp family language experimentation seems to have declined.
The local maximum maxim could explain that, but for the twitter and buzz
about Arc. Let's face it: If everyone were perfectly satisfied with CL,
Arc wouldn't have the hypnotic effect that it does. There is even talk
of Arc becoming the vehicle to bring a new multitude of users into the fold.
One could say that nothing is stopping anyone from experimenting with
new Lisps. But I don't think that's quite true. There is a subtle but
unmistakable air of waiting for the Messiah about the community, and I
feel that other activity has been curtailed in the meantime.
I would like to emphasize, as strongly as possible, that I believe that
Mr. Graham has nothing but goodwill for the community. I have been of
two minds, while writing this, as to whether Mr. Graham somehow owes the
community more frequent updates; whether the expectations which he has
built, considering his standing in the community, have created a duty of
care. I conclude that he owes us nothing, though others may disagree.
<dose of much needed levity>
While reading c.l.l postings about Arc as background for this, I came
across this URL: http://www.iol.ie/~pcassidy/ARC/guru.html
Entitled "The Guru Papers - a review", this caught my eye: "Part One
examines the most extreme example of one person giving power to another,
the guru/disciple relationship, because this reveals less obvious
occurrences of control." Spoof? Scholarship? Twaddle? Hitting the button
marked "Back to ARC Page" took me to...
The Official alt.recovery.catholicism Webpages
</levity>
Writing and rewriting this has been difficult. I fear that some people
may see it as an attack on an institution. Perhaps Mr. Graham himself
will think that I've gored his ox, though I would like to assure
everyone that this has not been my intent.
I've seen a certain lack of language experiment, combined with a
wistfulness for experimental new Lisps, Arc in particular. Given the
power of the tools we have available to us, and the excellent
scholarship and sense of adventure that earlier Lisp experimenters have
bequeathed us, I think that's a pity.
--
Cameron MacKinnon
Toronto, Canada
As someone interested in what state Arc is in these days, It is pretty
frustrating (in an amused kind of way) at just how little info about
arc is really known. But is arc does not seem to get much attention on
CLL, from what I've seen...
Searching the newsgroups I personally could find not a single thread
(besides this one) on the subject matter of Arc in '04. Searching for
"arc" "paul" "graham" "lisp" yields only 3 or 4 articles that refer to
arc to make some kind of indirect point.
Are you saying Arc his holding up the development of "Common Lisp 2"?
I don't see much of that, really...
Conrad Barski wrote:
> Searching the newsgroups I personally could find not a single thread
> (besides this one) on the subject matter of Arc in '04. Searching for
> "arc" "paul" "graham" "lisp" yields only 3 or 4 articles that refer to
> arc to make some kind of indirect point.
Arc is mentioned once or more in an average of four threads per month.
It doesn't often get its own "Whither Arc?" thread.
> Are you saying Arc his holding up the development of "Common Lisp 2"?
> I don't see much of that, really...
No, I have no power to see what Arc is stifling the development of, if
anything. But I do see people asking after Arc, which to me is an
indicator that they're interested in new Lispy languages. As playing
with Lispy (or even not very Lispy) dialects is as easy as falling off a
log, and the process is well documented to boot, I've made an
observation as to why, perhaps, more people are asking than playing.
--
Cameron MacKinnon
Toronto, Canada
Cameron MacKinnon <··········@clearspot.net> writes:
> As playing with Lispy (or even not very Lispy) dialects is as easy
> as falling off a log, and the process is well documented to boot,
> I've made an observation as to why, perhaps, more people are asking
> than playing.
Asking is easier than playing.
Learning enough Common Lisp to do a serious hobby project takes some
effort. There are packages to download and install, development
environments to configure, books to read, source files to browse for
instruction and inspiration. It's not immediately apparent how
experience in other systems can apply to accelerating the learning of
Common Lisp. This may lead some to think "Hmm, Arc's supposed to be
like a cleaner [whatever that means] Lisp for {web programming,
lightweight scripting, etc.}. I already know {web programming,
lightweight scripting, etc.}, so maybe it'll be easier to learn Lisp,
which everyone says is cool, when Arc comes out. Whee, I don't have to
learn Lisp until then!"
However, once you're over the initial hump, once you have confidence
in your CL environment and abilities, once you feel that you can
complete any project you like given sufficient effort, I don't think
you'd be inclined to wait around, doing nothing, for Arc.
Zach
Conrad Barski wrote:
> Are you saying Arc his holding up the development of "Common Lisp 2"?
> I don't see much of that, really...
The idea of a "Common Lisp 2" doesn't make a lot of sense. It's hard to
find things in CL that can be substantially improved and need
fundamental changes to the language at the same time. Most things can be
achieved by simple add-on libraries.
I have recently discoverd the following entry at Martin Fowler's
"Bliki": http://www.martinfowler.com/bliki/SoftwareDevelopmentAttitude.html
It seems to me that Common Lisp was designed with a strong "enabling
attitude". From such a perspective, it doesn't make a lot of sense to
change the language standard without achieving an essential step forward
in expressivity.
Pascal
--
1st European Lisp and Scheme Workshop
June 13 - Oslo, Norway - co-located with ECOOP 2004
http://www.cs.uni-bonn.de/~costanza/lisp-ecoop/
> The idea of a "Common Lisp 2" doesn't make a lot of sense. It's hard to
> find things in CL that can be substantially improved and need
> fundamental changes to the language at the same time. Most things can be
> achieved by simple add-on libraries.
I put "Common Lisp 2" in quotes because I agree with your sentiment-
It doesn't make much sense. I thought at first that Cameron was
arguing that Arc was having a negative impact on the lisp community,
which I couldn't follow. But I think he was implying instead that
people's attitude to Arc reveals something about the personality of
lispers.
> I have recently discoverd the following entry at Martin Fowler's
> "Bliki": http://www.martinfowler.com/bliki/SoftwareDevelopmentAttitude.html
That's a pretty insightful link- Although I can't understand what the
appeal of the "directing" attitude is, there is no doubt that it is a
very successful one, given the design of the current mainstream
languages.
·····················@yahoo.com (Conrad Barski) writes:
> That's a pretty insightful link- Although I can't understand what the
> appeal of the "directing" attitude is, there is no doubt that it is a
> very successful one, given the design of the current mainstream
> languages.
The appeal is that managers want the software development process to
be reproducible, with costs thereof accurately estimable, so they can
do their business planning.
This, in itself, is not a bad thing. However, they want this
SO BADLY that they've deluded themselves into thinking that
it's not only possible, but the "industry standards best practices".
(i.e. the UML crowd).
Now, the problem as I see it, is that only the very simple,
well understood "me too" software projects are in fact amenable
to this sort of planning, and can be brought to successful conclusion
by "plug compatible, interchangeable engineers" (i.e. the Java crowd).
Also, in any largish organization, the manager will have strong
CYA incentive, and thus will do anything to avoid risk if failure
can be pinned on him. (If the project fails, but he did everything
"the standard way", he's normally safe).
Put all these factors together, and you understand the state of
the industry as it is today. It also explains why it's so darn hard
to get any interesting new software out.
Alain Picard <············@memetrics.com> writes:
> ·····················@yahoo.com (Conrad Barski) writes:
>
> > That's a pretty insightful link- Although I can't understand what the
> > appeal of the "directing" attitude is, there is no doubt that it is a
> > very successful one, given the design of the current mainstream
> > languages.
>
> The appeal is that managers want the software development process to
> be reproducible, with costs thereof accurately estimable, so they can
> do their business planning.
Oh but they do get it! The software development process is highly
reproductible, and its cost is accurately estimable.
--> failure (in 70% of cases) ; thrice the initialy estimated cost.
--
__Pascal_Bourguignon__ http://www.informatimago.com/
There is no worse tyranny than to force a man to pay for what he doesn't
want merely because you think it would be good for him.--Robert Heinlein
http://www.theadvocates.org/
On 28 Mar 2004 09:05:41 +0200, Pascal Bourguignon
<····@thalassa.informatimago.com> wrote:
> Alain Picard <············@memetrics.com> writes:
>> ·····················@yahoo.com (Conrad Barski) writes:
>>
>> > That's a pretty insightful link- Although I can't understand what the
>> > appeal of the "directing" attitude is, there is no doubt that it is a
>> > very successful one, given the design of the current mainstream
>> > languages.
>>
>> The appeal is that managers want the software development process to
>> be reproducible, with costs thereof accurately estimable, so they can
>> do their business planning.
>
> Oh but they do get it! The software development process is highly
> reproductible, and its cost is accurately estimable.
> --> failure (in 70% of cases) ; thrice the initialy estimated cost.
>
It's not restricted to software either, it's what happens
anytime you put bean counters in charge of something
essentially artistic in nature, usually poorly thought out,
and wanted ASAP.
--
dg
Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/
Don Groves wrote:
> On 28 Mar 2004 09:05:41 +0200, Pascal Bourguignon
> <····@thalassa.informatimago.com> wrote:
>
>> Alain Picard <············@memetrics.com> writes:
>>
>>> ·····················@yahoo.com (Conrad Barski) writes:
>>>
>>> > That's a pretty insightful link- Although I can't understand what the
>>> > appeal of the "directing" attitude is, there is no doubt that it is a
>>> > very successful one, given the design of the current mainstream
>>> > languages.
When you program, you're directing the computer (less so in Prolog,
perhaps). Now look at the manual for the most popular lowest level
language, Intel's programming manual for the x86. It's full of
directions and constructs, very few of which seem to have been built for
the convenience of, or to nurture the creativity of the programmer.
So to expect a language designer, a programmer who is accustomed to
having his direction followed and who is writing a translator to a
tyrannical architecture, to create a liberating environment is to expect
the unlikely. That it ever happens is something to give thanks for.
>>> The appeal is that managers want the software development process to
>>> be reproducible, with costs thereof accurately estimable, so they can
>>> do their business planning.
>>
>>
>> Oh but they do get it! The software development process is highly
>> reproductible, and its cost is accurately estimable.
>> --> failure (in 70% of cases) ; thrice the initialy estimated cost.
>>
>
> It's not restricted to software either, it's what happens
> anytime you put bean counters in charge of something
> essentially artistic in nature, usually poorly thought out,
> and wanted ASAP.
Some programming is artistic. But look at the project that kicked off
the extreme programming fad (and which eventually failed). It was a
payroll system! Is it not reasonable to suggest that writing a payroll
system doesn't take a lot of creativity, and that one should be able to
plan and budget such a project?
--
Cameron MacKinnon
Toronto, Canada
On Sun, 28 Mar 2004 16:00:26 -0500, Cameron MacKinnon
<··········@clearspot.net> wrote:
> Don Groves wrote:
>> On 28 Mar 2004 09:05:41 +0200, Pascal Bourguignon
>> <····@thalassa.informatimago.com> wrote:
>>>> The appeal is that managers want the software development process to
>>>> be reproducible, with costs thereof accurately estimable, so they can
>>>> do their business planning.
>>>
>>>
>>> Oh but they do get it! The software development process is highly
>>> reproductible, and its cost is accurately estimable.
>>> --> failure (in 70% of cases) ; thrice the initialy estimated cost.
>>>
>>
>> It's not restricted to software either, it's what happens
>> anytime you put bean counters in charge of something
>> essentially artistic in nature, usually poorly thought out,
>> and wanted ASAP.
>
> Some programming is artistic. But look at the project that kicked off
> the extreme programming fad (and which eventually failed). It was a
> payroll system! Is it not reasonable to suggest that writing a payroll
> system doesn't take a lot of creativity, and that one should be able to
> plan and budget such a project?
>
I should have written "creative process" instead of "essentially
artistic" but yes, I agree that a payroll program, of all things,
should be able to be properly planned and budgeted! Yet, projects
as well defined as payroll are messed up all the time. I still
give the bean counters most of the blame - programmers will never
be as predictable as, say, a machinist cranking out 100s of the
same part hour after hour, yet the bean counters' spreadsheets
treat them as if they were.
--
dg
Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/
Cameron MacKinnon <··········@clearspot.net> writes:
> Some programming is artistic. But look at the project that kicked off
> the extreme programming fad (and which eventually failed). It was a
> payroll system! Is it not reasonable to suggest that writing a payroll
> system doesn't take a lot of creativity, and that one should be able
> to plan and budget such a project?
I think it might take more creativity than you'd think.
There must be some reason that existing payroll products
don't satisfy the needs. If they did, you wouldn't need to design a
new payroll program. And although the abstraction presented to the
end user is rather pedestrian, the implementation could be quite
exotic.
When I started working at True Software on their change management
engine (think source code control on steroids), I figured that it
would be incredibly dull. The model presented to the user was in fact
quite dull. Underneath, however, was some amazingly intricate and
interesting code that involved quite a bit of creativity.
"Joe Marshall" <···@ccs.neu.edu> wrote in message
·················@ccs.neu.edu...
> I think it might take more creativity than you'd think.
>
> There must be some reason that existing payroll products
> don't satisfy the needs. If they did, you wouldn't need to design a
> new payroll program. And although the abstraction presented to the
> end user is rather pedestrian, the implementation could be quite
> exotic.
The reason payroll systems need to be changed is because the layer that most
payroll systems sit on is in constant flux, notably because it's dictated
mostly by several disparate governing bodies. Specifically State government,
Federal Government, Unions, etc. Throw in health care plans, retirement and
pensions, and other benefits, then sprinkle on contractors, out of state
employees (Hot Damn! A new jurisdiction to support!) and you can see how
seemingly simple payroll (Bob worked 8 hours, and earned $20/hr) get very
complicated very quickly.
Add to all of that complexity the fact that folks actually like getting
their paychecks on time, AND they like them to be accurate, and you can see
how these systems need to be rock solid on a bed of continuing change.
Finally, since many of the regulations are simply legislative edicts,
they're not necessarily "logical" or easily abstracted into a generic model,
so invasive code logic changes are required. Kind of like the rules for
Fizbin.
Places like ADP exist for a reason.
Regards,
Will Hartung
(·····@msoft.com)
Pascal Bourguignon <····@thalassa.informatimago.com> writes:
> Alain Picard <············@memetrics.com> writes:
>>
>> The appeal is that managers want the software development process to
>> be reproducible, with costs thereof accurately estimable, so they can
>> do their business planning.
>
> Oh but they do get it! The software development process is highly
> reproductible, and its cost is accurately estimable.
> --> failure (in 70% of cases) ; thrice the initialy estimated cost.
Yes. Strange, isn't it? They say that one definition of insanity
is doing the same thing over and over again, and keep expecting
different results...
Perhaps the result is that you're a manager of a team of tens running a
project with a large budget rather than a manager of a team of two
running a project with a small budget.
Pete
Pascal Bourguignon <····@thalassa.informatimago.com> writes:
> Oh but they do get it! The software development process is highly
> reproductible, and its cost is accurately estimable.
> --> failure (in 70% of cases) ; thrice the initialy estimated cost.
Of course, taking this into account will not increase the accuracy of
the estimate.
Joe Marshall wrote:
> Pascal Bourguignon <····@thalassa.informatimago.com> writes:
>
>
>>Oh but they do get it! The software development process is highly
>>reproductible, and its cost is accurately estimable.
>>--> failure (in 70% of cases) ; thrice the initialy estimated cost.
>
>
> Of course, taking this into account will not increase the accuracy of
> the estimate.
Ya can't fight Hofstatter's (sp?) Law.
kt
--
Home? http://tilton-technology.com
Cells? http://www.common-lisp.net/project/cells/
Cello? http://www.common-lisp.net/project/cello/
Why Lisp? http://alu.cliki.net/RtL%20Highlight%20Film
Your Project Here! http://alu.cliki.net/Industry%20Application
Cameron MacKinnon <··········@clearspot.net> wrote:
> and play at Language Architect. But here, everybody plays at
> "Waiting for Arc."
I wonder if there's a hidden evil twin to cll where this happens.
Or maybe I've just killfiled all those people and threads...
> their elders await Arc with eager anticipation, so they do
> likewise.
Even more so. I can't remember any one person whom I'd qualify
having a long-standing position in the "community" saying anything
to that effect. Please correct me if I've just been sleeping.
It seems that those who express a positive attitude to Arc
do so by effectively saying "It's good that new dialects
are being worked on, but I'm not holding my breath."
Also, on a tangent, it seems to me that cll is in a way gradually
losing it's position as a community centre: there is activity and
discussions on various mailing lists and on IRC that never reach
this newsgroup. Whether this is good or bad -- or even permanent
-- remains to be seen.
Cheers,
-- Nikodemus
In article <··············@midnight.cs.hut.fi>,
·········@random-state.net wrote:
> Cameron MacKinnon <··········@clearspot.net> wrote:
>
> > and play at Language Architect. But here, everybody plays at
> > "Waiting for Arc."
>
> I wonder if there's a hidden evil twin to cll where this happens.
> Or maybe I've just killfiled all those people and threads...
>
> > their elders await Arc with eager anticipation, so they do
> > likewise.
>
> Even more so. I can't remember any one person whom I'd qualify
> having a long-standing position in the "community" saying anything
> to that effect. Please correct me if I've just been sleeping.
>
> It seems that those who express a positive attitude to Arc
> do so by effectively saying "It's good that new dialects
> are being worked on, but I'm not holding my breath."
Same here.
> Also, on a tangent, it seems to me that cll is in a way gradually
> losing it's position as a community centre: there is activity and
> discussions on various mailing lists and on IRC that never reach
> this newsgroup. Whether this is good or bad -- or even permanent
> -- remains to be seen.
This has always been the case. Many lisp people never post to
cll (some of them are reading it though). cll never was THE center of the
Lisp universe. Technical discussions are very often done
via mailing lists. Several mailing list have played an important role
for discussions, work, community building, etc. I'm especially
grateful for all the discussions that have been going on on the
slug mailing list and info-mcl (8000+ messages in my archive since 1993)
- to name a few. Recently IRC has found quite some fans. #lisp at
Freenode has 112 people (or bots) logged on just now.
Rainer Joswig wrote:
> - to name a few. Recently IRC has found quite some fans. #lisp at
> Freenode has 112 people (or bots) logged on just now.
In the ircNet we are only around 5 people, but I started as one :)
Cameron MacKinnon wrote:
> Now, here's what's bothering me. Looking down the language power curve,
> there's all kinds of experimentation going on. Much of it, to the
> enlightened denizens of this newsgroup, looks like a complete waste of
> time. Nonetheless, many people, whether motivated by vanity, unique
> needs or the simple love of novelty, take matters into their own hands
> and play at Language Architect. But here, everybody plays at "Waiting
> for Arc."
<snip>
> One could say that nothing is stopping anyone from experimenting with
> new Lisps. But I don't think that's quite true. There is a subtle but
> unmistakable air of waiting for the Messiah about the community, and I
> feel that other activity has been curtailed in the meantime.
You are just projecting your internal view on what you are reading
here. Paul Grahams dissatisfaction with Common Lisp is his own
own little demon. Others in this group dissatisfied with CL
also have their aesthetic-demon. Do you feel like experimenting
with a new Lisp? Why? If you do, go ahead, feel free. What
is stopping people is reality, its the same for Paul Graham, its hard.
Arc is probably turning out to be like Common Lisp.
Wade
Wade Humeniuk wrote:
> Cameron MacKinnon wrote:
>> One could say that nothing is stopping anyone from experimenting with
>> new Lisps. But I don't think that's quite true. There is a subtle but
>> unmistakable air of waiting for the Messiah about the community, and I
>> feel that other activity has been curtailed in the meantime.
>
> You are just projecting your internal view on what you are reading
> here. Paul Grahams dissatisfaction with Common Lisp is his own
> own little demon. Others in this group dissatisfied with CL
> also have their aesthetic-demon. Do you feel like experimenting
> with a new Lisp? Why? If you do, go ahead, feel free. What
> is stopping people is reality, its the same for Paul Graham, its hard.
> Arc is probably turning out to be like Common Lisp.
I do not yet feel that it is time for me to start experimenting with
Lisp alternatives. Right now, I think I have a year or two, at least, of
using Lisp and Scheme before I would consider personal experimentation
with new dialects for anything but domain specific problems.
But, as I think I did a reasonable job explaining, there are lots of
people here who are waiting expectantly for Arc. If these people are
interested in a new take on Lisp, what's stopping them from
experimenting? Certainly the available tools and the previous research
should be encouraging them.
--
Cameron MacKinnon
Toronto, Canada
Cameron MacKinnon <··········@clearspot.net> wrote:
> But, as I think I did a reasonable job explaining, there are
> lots of people here who are waiting expectantly for Arc. If
Well, you asserted that, but assertion does not make it so.
My _perception_ is that most of those people who are waiting
for Arc are more-or-less using "waiting for Arc" as an excuse
to not use lisp today. Maybe I'm wrong.
Let's behave like children and have show of hands. ;)
Who here is "expactantly waiting for Arc"? And among the teeming
hordes who say "aye", who among you has actively used CL for ...
say more than two years?
I'm surprised if there are more than a few yesses to the first,
and I'd be willing to bet a pint of fine ale that there ain't more
than two who answer both affirmatively.
Arc may turn out great or rubbish, but in the meanwhile most
people have better things to do -- like using CL -- than
"expactantly wait" for something that may or may not happen.
Cheers,
-- Nikodemus
<·········@random-state.net> wrote
> Cameron MacKinnon <··········@clearspot.net> wrote:
>
> > But, as I think I did a reasonable job explaining, there are
> > lots of people here who are waiting expectantly for Arc. If
>
> Well, you asserted that, but assertion does not make it so.
> My _perception_ is that most of those people who are waiting
> for Arc are more-or-less using "waiting for Arc" as an excuse
> to not use lisp today. Maybe I'm wrong.
>
> Let's behave like children and have show of hands. ;)
>
> Who here is "expactantly waiting for Arc"? And among the teeming
> hordes who say "aye", who among you has actively used CL for ...
> say more than two years?
>
> I'm surprised if there are more than a few yesses to the first,
> and I'd be willing to bet a pint of fine ale that there ain't more
> than two who answer both affirmatively.
>
> Arc may turn out great or rubbish, but in the meanwhile most
> people have better things to do -- like using CL -- than
> "expactantly wait" for something that may or may not happen.
Well said. I'm really not waiting for Arc. Anybody can write a new dialect of
Lisp. It's so easy that it's worthless. It's about as dumb as being proud of
being able to hack and deface CLiki. Writing useful code is much harder as
you have to understand at least some problem domain for this.
One of the great idea of Common Lisp is the "Common" word. People not
understanding that will just add a new variant to the Greenspun's tenth law.
And about the popularity of Lisp, it has reached a point where it is low
enough so that I've never seen a dumb Lispnik though high enough to attract
new clever people. So let's leave it that way. ;-)
Marc
Cameron MacKinnon <··········@clearspot.net> writes:
> But, as I think I did a reasonable job explaining, there are lots of
> people here who are waiting expectantly for Arc. If these people are
> interested in a new take on Lisp, what's stopping them from
> experimenting? Certainly the available tools and the previous research
> should be encouraging them.
Writing lisps is fun, but one of the biggest problem with the uncommon
lisps (scheme, arc, goo, maybe dylan counts) is lack of existing
libraries and code. I reason as follows: if arc turns out to be to my
taste, I can help improving it or write/port libraries for it rather
than do my own lisp.
I'm curious to see how arc will be implemented and how it will look,
and which choices will be made.
And I just because I'm eager to see arc as soon as possible doesn't
mean that I'm rolling my thumbs doing nothing.
On Sat, 03 Apr 2004 00:16:55 +0000, Sunnan wrote:
> Cameron MacKinnon <··········@clearspot.net> writes:
>> But, as I think I did a reasonable job explaining, there are lots of
>> people here who are waiting expectantly for Arc. If these people are
>> interested in a new take on Lisp, what's stopping them from
>> experimenting? Certainly the available tools and the previous research
>> should be encouraging them.
>
> Writing lisps is fun, but one of the biggest problem with the uncommon
> lisps (scheme, arc, goo, maybe dylan counts) is lack of existing
> libraries and code. I reason as follows: if arc turns out to be to my
> taste, I can help improving it or write/port libraries for it rather
> than do my own lisp.
Linj might be of interest to you. It looks like Common Lisp but it
gives you access to (probably) the biggest libraries and code repositories
in the entire world: Java.
Antonio Menezes Leitao wrote:
> Linj might be of interest to you. It looks like Common Lisp but it
> gives you access to (probably) the biggest libraries and code repositories
> in the entire world: Java.
As I am doing a lot Java work I am quite interested in Linj. But I have
failed so far to obtain it - the "download" page only contains a (very nice)
tutorial.
Peter
On Sat, 03 Apr 2004 09:08:20 +0100, Antonio Menezes Leitao
<··············@evaluator.pt> wrote:
>Linj might be of interest to you. It looks like Common Lisp but it
>gives you access to (probably) the biggest libraries and code repositories
>in the entire world: Java.
So when will Linj finally be available for download? Its download page
is saying this for months now:
Linj download page
Linj is free for non-comercial use but we are still thinking about the
best license for Linj distribution.
We will provide more information in the next weeks. For the moment, we
only provide a short tutorial on Linj.
Linj Tutorial (in HTML)
Linj Manual (in PDF) (includes the tutorial but is not finished yet)
--Denis.
On Sat, 03 Apr 2004 13:13:37 +0200, Denis Mashkevich wrote:
> On Sat, 03 Apr 2004 09:08:20 +0100, Antonio Menezes Leitao
> <··············@evaluator.pt> wrote:
>
>> [quoted text muted]
>
> So when will Linj finally be available for download? Its download page
> is saying this for months now:
> ...
Yes. I know and it's my fault. I've been too busy doing other things
that I didn't find time to finish the packaging of the compiler. I'll
have some time at the end of the week and 'something' will be
available at the download page.
On Sat, 03 Apr 2004 13:13:37 +0200, Denis Mashkevich wrote:
> On Sat, 03 Apr 2004 09:08:20 +0100, Antonio Menezes Leitao
> <··············@evaluator.pt> wrote:
>
>>Linj might be of interest to you. It looks like Common Lisp but it
>>gives you access to (probably) the biggest libraries and code repositories
>>in the entire world: Java.
>
> So when will Linj finally be available for download?
It is available now! Sorry for the delay.
Antonio Leitao
On 2004-04-11 09:53:10, Antonio Menezes Leitao wrote:
> On Sat, 03 Apr 2004 13:13:37 +0200, Denis Mashkevich wrote:
>> On Sat, 03 Apr 2004 09:08:20 +0100, Antonio Menezes Leitao
>> <··············@evaluator.pt> wrote:
>>>Linj might be of interest to you. It looks like Common Lisp but it
>>>gives you access to (probably) the biggest libraries and code repositories
>>>in the entire world: Java.
>>
>> So when will Linj finally be available for download?
>
> It is available now! Sorry for the delay.
Don't be sorry. The examples were fun to read.
I have to be sorry for not having enough time to give some feedback
atm. But I'll play with it any time soon.
Sunnan wrote:
> Writing lisps is fun, but one of the biggest problem with the uncommon
> lisps (scheme, arc, goo, maybe dylan counts) is lack of existing
> libraries and code.
Not quite. Implementations like DrScheme have rather large libraries,
and together with Swindle you have pretty much the same expressivity
feature-wise as the average Common Lisp implementation. So stating that
the lack of libraries is the main reason that holds you back from using
Scheme is painting a somewhat wrong picture.
Of course, there are other reasons that might hold you back (lack of
portability, different fundamental design choices, and so on).
Pascal
--
1st European Lisp and Scheme Workshop
June 13 - Oslo, Norway - co-located with ECOOP 2004
http://www.cs.uni-bonn.de/~costanza/lisp-ecoop/
Pascal Costanza <········@web.de> writes:
> Not quite. Implementations like DrScheme have rather large libraries,
> and together with Swindle you have pretty much the same expressivity
> feature-wise as the average Common Lisp implementation. So stating
> that the lack of libraries is the main reason that holds you back from
> using Scheme is painting a somewhat wrong picture.
Oh, I'm using scheme right now. (Though CL does have even more
libraries. For example, I've written my own chicken wrappers for a
subset of the Simple DirectMedia Layer, while CL already have, as far
as I'm told, good support for SDL.)
The scheme situation is rapidly improving, compared to only a year
ago. (The same holds true for CL.)
So lack of libraries aren't holding me back from using any lisp. What
it is holding me back from, however, is _writing my own_ lisp. What
would be the point to compete with scheme, arc, goo et cetera and
possibly cannibalize on their "library-writing users"?
So I want to see if arc is good or not. There are several question
marks that I want to have straightened out (the macro system, the
object system, the syntax, the compilation process, performance, the
profiler, and more).
(For the curious, if I were to make my own lisp (not bloody likely) it
would probably be as big of an intersection between scheme and CL I
could muster, so code could at least theoretically be ported from
CL. It would diverge in being single-namespace and having
continuations and all the other loverly scheme-stuff. It would
probably a superset of R5RS. I would probably base it on an existing
scheme, or just help improving an existing scheme.)
(aside no. 2: I'm happy that you're having a lisp workshop in Norway,
I wanted to go but it might be tough. Should I email you about queries
or are you too busy?)
--
One love,
Sunnan
Cameron MacKinnon wrote:
> Meanwhile, Lisp family language experimentation seems to have declined.
Everytime you're writing a macro, you're basically changing the
language. So I wouldn't agree that language experimentation has stopped
happening. It's just that new Lisp "dialects" don't necessarily need a
fanfare and a new catchy name just to be considered new dialects. Most
Lisp dialects are "silent". You can rightfully say that many basic
paradigms have been explored in the 70's and 80's and that this is not
happening anymore. But apart from that, I think Lisp is still used
basically in the same way as it was used back then. The only thing that
has changed is that we have two language standards around which we can
revolve, and this has stabilized some basic concepts. But that's all IMHO.
Pascal
--
1st European Lisp and Scheme Workshop
June 13 - Oslo, Norway - co-located with ECOOP 2004
http://www.cs.uni-bonn.de/~costanza/lisp-ecoop/
Cameron MacKinnon wrote:
.....
> Now, here's what's bothering me. Looking down the language power curve,
> there's all kinds of experimentation going on. Much of it, to the
> enlightened denizens of this newsgroup, looks like a complete waste of
> time. Nonetheless, many people, whether motivated by vanity, unique
> needs or the simple love of novelty, take matters into their own hands
> and play at Language Architect. But here, everybody plays at "Waiting
> for Arc."
.....
> One could say that nothing is stopping anyone from experimenting with
> new Lisps. But I don't think that's quite true. There is a subtle but
> unmistakable air of waiting for the Messiah about the community, and I
> feel that other activity has been curtailed in the meantime.
.....
> I've seen a certain lack of language experiment, combined with a
> wistfulness for experimental new Lisps, Arc in particular. Given the
> power of the tools we have available to us, and the excellent
> scholarship and sense of adventure that earlier Lisp experimenters have
> bequeathed us, I think that's a pity.
This question ties directly into CL history. Common Lisp arose when a
bunch of different lisps were reaching for a common ground, a superset
that included most of the idioms and design points that allowed code
written for any of them to be easily ported to CL.
Things had gotten to the point where the very plasticity of the language
worked against it; people in different shops were developing lisp into very
different languages, with different infrastructure and requirements, and
as a result code written at separate lisp shops was very hard to put
together into any kind of integrated application, because every time you
lifted even a single library from some lisp shop anywhere, you wound up
having to import that shop's entire codebase of infrastructure. Then,
with five or six infrastructures stuck together and (sorta) jerking along
spastically and working, you'd build something out of those libraries - but
then if anybody else wanted to use what you built, they'd have to import
*all* of those five or six redundant infrastructures, in addition to the
ones that were already imported in their project. And that's only if
things were arranged so that you started from the same core implementation;
if one shop was using MacLisp and one shop was using Symbolics Lisp then
there was even more trouble, and sometimes you had interpreters for one
lisp dialect written in another, which would become part of basic
infrastructure as the code moved...
And eventually, some major clients (The DoD in particular, I think) said
"hey, this is getting stupid. Could you guys please all agree on what
Lisp is, so we can write projects that integrate with each other and get
some reusable infrastructure out of them?"
And after a few years of hammering and tail-twisting and gnashing of
teeth, the community did exactly that. Common Lisp is the result. The
community has been burned by language divergence in the past, and these
days there is real reluctance to make code depend on something that changes
the language in any fundamental ways that make it incompatible with the
"basic design" of common lisp. This inhibits experimentation, but
generally is regarded as good because it promotes interoperation and
portable libraries.
The experimenters have been driven away from the CL camp, mainly into
Scheme. And the Scheme camp shows the effects; much that is innovative
gets done there, many new designs and ideas get tried out, and very little
Scheme code is portable between implementations. And so you have the
situation where there is flexibility but no unified direction in the
scheme camp, and a lack of flexibility in the CL camp, and new ideas
and designs are circulating and simmering ... enough new ideas to build
into a powerful new lisp dialect.
People keep waiting, I think, for some luminary, someone who is well-
known in the community, someone whose work will attract a critical mass
of attention, to be the one to do it. It will simply take that big a
force to develop a new direction in Lisp code. The list of names is
short: Steele, Sussman, McCarthy, Moon, Graham, and Knuth are six
acknowledged living Wizards; if any of them developed a new lisp dialect
then Lispers would stand up and take notice and evaluate it seriously as
an alternative to whatever they're doing now. They aren't the only
living wizards, and Knuth isn't even particularly a Lisp wizard - but
they are certainly among those with the greatest name recognition and
idea impact.
There are other luminaries who could do it; If Larry Wall, another living
Wizard, announced tomorrow that Perl 7 would have a fully-parenthesized
prefix notation and data intepretable as code in order to facilitate
lisp-like macrology, it would attract the same kind of attention. If
Kenny Tilton announced that his next project after Cello would be a lisp
system that fixed some of the things he found annoying about CL, it would
at least get people to notice. But neither of those seem likely to
happen; Wall is unlikely to take the final step of making Perl into a
Lisp, and Tilton seems reasonably happy with CL and too preoccupied with
getting-stuff-done to take up the gauntlet in a language-design crusade.
So, since one of the acknowledged wizards appears to be working on a new
Lisp, people are buzzing and muttering and wondering what it's going to
be like.
Bear
Ray Dillinger wrote:
> This question ties directly into CL history.
...
Thanks, Ray. I'd read a lot of that before, but you did an amazing job
of tying it together and putting it in context.
> The experimenters have been driven away from the CL camp, mainly into
> Scheme.
That's very true. Maybe a conservative or reactionary CL camp and a wild
and crazy Scheme hothouse, where every third person seems to be an
implementer, is a symbiotic relationship? The question might then be
"When a Schemer comes up with the next new thing, will the CL community
be too proud to adopt it?"
> People keep waiting, I think, for some luminary, someone who is
> well-known in the community, someone whose work will attract a
> critical mass of attention, to be the one to [build a new dialect].
This got me thinking. Who has built multiple languages which gained
traction? I can only think of Brian Kernighan and Niklaus Wirth, though
there *must* be more. Could language design really be a young man's
game? My first answer would be no, that experience helps save designers
from bad choices. Maybe experience really leads to "analysis paralysis"
though, and hubris wins the day.
--
Cameron MacKinnon
Toronto, Canada
Cameron MacKinnon wrote:
>> The experimenters have been driven away from the CL camp, mainly into
>> Scheme.
>
>
>
> That's very true. Maybe a conservative or reactionary CL camp and a wild
> and crazy Scheme hothouse, where every third person seems to be an
> implementer, is a symbiotic relationship? The question might then be
> "When a Schemer comes up with the next new thing, will the CL community
> be too proud to adopt it?"
>
Just in all things need is what will drive language features. In the case
of programming its the needs of applications that drive what a program
has to do. Ask yourself, what application today cannot be done with the
language features available? (This does not include applications that
have not been currently solved like various AI problems). An example is
OO, it was probably driven by the increasing size of applications. Bottom
up inovation probably will not be used if applications have no pressing
need for the functionality.
And, no, the CL community will not be too proud to adopt it, proof of that is
that it has survived almost 50 years by meeting the needs of applications.
>
> This got me thinking. Who has built multiple languages which gained
> traction? I can only think of Brian Kernighan and Niklaus Wirth, though
> there *must* be more. Could language design really be a young man's
> game? My first answer would be no, that experience helps save designers
> from bad choices. Maybe experience really leads to "analysis paralysis"
> though, and hubris wins the day.
>
Great intellectual systems that survive the small lifespans of human beings
are based on experience. Most things you know today are passed down from
other's experiences and thinking. Its those things that stand the test of time
(like the scientific method or logic) that are important. We shall all have to wait
(and will probably all be dead) before we answer these questions.
Wade
Cameron MacKinnon <··········@clearspot.net> wrote:
+---------------
| This got me thinking. Who has built multiple languages which gained
| traction? I can only think of Brian Kernighan and Niklaus Wirth, though
| there *must* be more. ...
+---------------
Well, Guy Steele co-invented Scheme and was a major cotributor to
Common Lisp (and was peripherally related to Java, too, wasn't he?).
And those are all still around...
-Rob
-----
Rob Warnock <····@rpw3.org>
627 26th Avenue <URL:http://rpw3.org/>
San Mateo, CA 94403 (650)572-2607
Rob Warnock wrote:
> Cameron MacKinnon <··········@clearspot.net> wrote:
> +---------------
> | This got me thinking. Who has built multiple languages which gained
> | traction? I can only think of Brian Kernighan and Niklaus Wirth, though
> | there *must* be more. ...
> +---------------
>
> Well, Guy Steele co-invented Scheme and was a major cotributor to
> Common Lisp (and was peripherally related to Java, too, wasn't he?).
And his PhD "thesis" was an implementation of a computer language based
wholly on constraints.
> And those are all still around...
Welllll...if the dinosaurs are birds, constraint programming languages
are now Cells. Something like that.
kt
Kenny Tilton wrote:
>
>
> Rob Warnock wrote:
>
>> Cameron MacKinnon <··········@clearspot.net> wrote:
>> +---------------
>> | This got me thinking. Who has built multiple languages which gained
>> | traction? I can only think of Brian Kernighan and Niklaus Wirth, though
>> | there *must* be more. ...
>> +---------------
>>
>> Well, Guy Steele co-invented Scheme and was a major cotributor to
>> Common Lisp (and was peripherally related to Java, too, wasn't he?).
>
>
> And his PhD "thesis" was an implementation of a computer language based
> wholly on constraints.
Hmmm, from constraints to lisp to scheme to java... I don't like that
trend line. Hopefully the income trend is the inverse.
kt
Kenny Tilton wrote:
>
>
> Kenny Tilton wrote:
>
>>
>>
>> Rob Warnock wrote:
>>
>>> Cameron MacKinnon <··········@clearspot.net> wrote:
>>> +---------------
>>> | This got me thinking. Who has built multiple languages which gained
>>> | traction? I can only think of Brian Kernighan and Niklaus Wirth,
>>> though
>>> | there *must* be more. ...
>>> +---------------
>>>
>>> Well, Guy Steele co-invented Scheme and was a major cotributor to
>>> Common Lisp (and was peripherally related to Java, too, wasn't he?).
>>
>>
>>
>> And his PhD "thesis" was an implementation of a computer language
>> based wholly on constraints.
>
>
>
> Hmmm, from constraints to lisp to scheme to java... I don't like that
> trend line. Hopefully the income trend is the inverse.
I think the sequence was more like scheme to lisp to java with
constraints somewhere along the way. Steele was just a wee cute tyke
when he designed the one-chip scheme microprocessor (fsvo all of those
words). Does the progression make any more sense?
paul
Paul Wallich wrote:
> Kenny Tilton wrote:
>
>>
>>
>> Kenny Tilton wrote:
>>
>>>
>>>
>>> Rob Warnock wrote:
>>>
>>>> Cameron MacKinnon <··········@clearspot.net> wrote:
>>>> +---------------
>>>> | This got me thinking. Who has built multiple languages which gained
>>>> | traction? I can only think of Brian Kernighan and Niklaus Wirth,
>>>> though
>>>> | there *must* be more. ...
>>>> +---------------
>>>>
>>>> Well, Guy Steele co-invented Scheme and was a major cotributor to
>>>> Common Lisp (and was peripherally related to Java, too, wasn't he?).
>>>
>>>
>>>
>>>
>>> And his PhD "thesis" was an implementation of a computer language
>>> based wholly on constraints.
>>
>>
>>
>>
>> Hmmm, from constraints to lisp to scheme to java... I don't like that
>> trend line. Hopefully the income trend is the inverse.
>
>
> I think the sequence was more like scheme to lisp to java with
> constraints somewhere along the way. Steele was just a wee cute tyke
> when he designed the one-chip scheme microprocessor (fsvo all of those
> words).
But constraints was the PhD thesis. But it sounds as if you are sure
scheme comes first, so then the sequence is:
constraints
/ \
/ \lisp
scheme |
|
|
java
Does the progression make any more sense?
From ivory tower to getting-one's-hands-dirty? I'm down with that.
kt
--
Home? http://tilton-technology.com
Cells? http://www.common-lisp.net/project/cells/
Cello? http://www.common-lisp.net/project/cello/
Why Lisp? http://alu.cliki.net/RtL%20Highlight%20Film
Your Project Here! http://alu.cliki.net/Industry%20Application
> >> Hmmm, from constraints to lisp to scheme to java... I don't like
> >> that trend line. Hopefully the income trend is the inverse.
> > I think the sequence was more like scheme to lisp to java with
> > constraints somewhere along the way. Steele was just a wee cute tyke
> > when he designed the one-chip scheme microprocessor (fsvo all of
> > those words).
>
> But constraints was the PhD thesis. But it sounds as if you are sure
> scheme comes first, ...
Scheme was before constraints
Constraints was his PhD thesis; the Rabbit compiler for Scheme
was a master's thesis; the initial development of Scheme was
even earlier. Common Lisp came after both.
-- jd
Kenny Tilton wrote:
> Welllll...if the dinosaurs are birds, constraint programming languages
> are now Cells. Something like that.
Does Cells handle constraints like SICP 3.3.5 and (to a more
sophisticated extend) languages like Modelica? I got the impression that
it was a framework that automated forward propagation of values only,
since the docs keep saying it's a like spread sheet for objects.
Pete
Pete Kirkham wrote:
> Kenny Tilton wrote:
>
>> Welllll...if the dinosaurs are birds, constraint programming languages
>> are now Cells. Something like that.
>
>
> Does Cells handle constraints like SICP 3.3.5 and (to a more
> sophisticated extend) languages like Modelica? I got the impression that
> it was a framework that automated forward propagation of values only,
> since the docs keep saying it's a like spread sheet for objects.
Right. The dinosaurs lost a little weight, such as partial constraints,
multi-way constraints, and the idea (in some implementations) that
constraints should do everything down to variable assignment. Cells have
the best qualities of constraints (declarative, automatic, and they keep
complex state self-consistent) but are lighter, faster, and do not want
to rule the world.
kt
--
Home? http://tilton-technology.com
Cells? http://www.common-lisp.net/project/cells/
Cello? http://www.common-lisp.net/project/cello/
Why Lisp? http://alu.cliki.net/RtL%20Highlight%20Film
Your Project Here! http://alu.cliki.net/Industry%20Application
Kenny Tilton wrote:
> Right. The dinosaurs lost a little weight, such as partial constraints,
> multi-way constraints, and the idea (in some implementations) that
> constraints should do everything down to variable assignment.
Pity- I've four legacy Fortran application suites to find a cost
effective and maintainable alternative to, and for each of M equations
of N variables has N functions to solve each equation given (N-1)
constant values where N ~= 6; even quite a simple system of constraints
could reduce the complexity from M*N to M, with one entry point per
equation (though that point probably needs a compiler behind it for when
it gets hit too often with a given set of constants).
Pete
Pete Kirkham wrote:
> Kenny Tilton wrote:
>
>> Right. The dinosaurs lost a little weight, such as partial
>> constraints, multi-way constraints, and the idea (in some
>> implementations) that constraints should do everything down to
>> variable assignment.
>
>
> Pity- I've four legacy Fortran application suites to find a cost
> effective and maintainable alternative to, and for each of M equations
> of N variables has N functions to solve each equation given (N-1)
> constant values where N ~= 6; even quite a simple system of constraints
> could reduce the complexity from M*N to M, with one entry point per
> equation (though that point probably needs a compiler behind it for when
> it gets hit too often with a given set of constants).
Hey, I'm just an applications guy, that's way over my BA head. Did you
try Screamer/Screamer+?
Constraint systems are OK where they have a contribution only
constraints systems can make, but they are just too damn hard to figure
out to be used for everyday programming.
The funny thing is that they set out to make programming easier. The sad
thing is that they did not turn back when they saw how hard partial and
multi-way constraints made everything. But if as you seem to think they
ended up able to solve some small class of very hard problem, good for them.
kt
--
Home? http://tilton-technology.com
Cells? http://www.common-lisp.net/project/cells/
Cello? http://www.common-lisp.net/project/cello/
Why Lisp? http://alu.cliki.net/RtL%20Highlight%20Film
Your Project Here! http://alu.cliki.net/Industry%20Application
On Mon, 29 Mar 2004 20:48:50 GMT, Kenny Tilton <·······@nyc.rr.com> wrote:
> Cells [...] do not want to rule the world.
Ooops! When did you change your mind?
Edi Weitz wrote:
> On Mon, 29 Mar 2004 20:48:50 GMT, Kenny Tilton <·······@nyc.rr.com> wrote:
>
>
>>Cells [...] do not want to rule the world.
>
>
> Ooops! When did you change your mind?
I didn't. That analogy started breaking up over Hackensack, I should
have hit the silk instead of trying to make it into Idlewild.
This is one mistake the dinosaurs did not make:
"Hey, this is neat! We should use it for /everything/!"
Next stop, performance and productivity Hell. I guess if that mistake
was good enough for Prolog and Smalltalk it was good enough for constraints.
Peter's invocation of Screamer+ to do his CLP will be quite happy living
in a Cell rule.
kt
--
Home? http://tilton-technology.com
Cells? http://www.common-lisp.net/project/cells/
Cello? http://www.common-lisp.net/project/cello/
Why Lisp? http://alu.cliki.net/RtL%20Highlight%20Film
Your Project Here! http://alu.cliki.net/Industry%20Application
Ray Dillinger <····@sonic.net> writes:
> This question ties directly into CL history. Common Lisp arose when a
> bunch of different lisps were reaching for a common ground, a superset
> that included most of the idioms and design points that allowed code
> written for any of them to be easily ported to CL.
I wouldn't say Common Lisp arose as a superset; it was more a common
subset, intended to unify the MacLisp "branch" of Lisp.
CL ended up unifying more than that for a variety of reasons,
some political (in a broad sense), some technical.
The main reasons there hasn't been a major new Lisp are
that there aren't enough people with the time and funding
to do it, that Common Lisp and Scheme are good enough that
there isn't much pressure to develop a new language, and
that when people do develop new languages not too far
from Lisp (Slate is a recent example), they prefer to
mix in features from other languages.
I don't think "waiting for Arc" - which is what's claimed
in this thread - is a major factor. I don't think there
are people who would be devising a major new Lisp if only
they weren't "waiting for Arc".
> People keep waiting, I think, for some luminary, someone who is well-
> known in the community, someone whose work will attract a critical mass
> of attention, to be the one to do it. It will simply take that big a
> force to develop a new direction in Lisp code. The list of names is
> short: Steele, Sussman, McCarthy, Moon, Graham, and Knuth are six
> acknowledged living Wizards; ...
Something like that happened a while back. There was a new language
that built on the accomplishments of Common Lisp and Scheme, as well
being influenced by other languages. A number of well-known Lisp
people were involved, and there was a lot of excitement and interest.
That language was Dylan.
Dylan still exists, but it failed to become the breakthrough
language that it had once seemed it might be. It also moved
further from the Lisp center, especially by dropping the
parenthesized syntax, in an effort to widen its appeal.
This shows the difficulty. It is not a small effort to develop
a good new language, provide high quality implementations, and
attract a large enough user community; and the involvement of
luminaries is not a big enough factor to make the difference
bwtween success and failure.
While people were working on Dylan, and inferior language,
with fewer luminaries, came along and won big for other reasons.
(I mean Java.) That's how it goes.
> So, since one of the acknowledged wizards appears to be working on a new
> Lisp, people are buzzing and muttering and wondering what it's going to
> be like.
Are they? And who are the people who are waiting for Arc, buzzing,
muttering? This whole thread seems quite strange to me.
-- jd
Jeff Dalton wrote:
> I don't think "waiting for Arc" - which is what's claimed
> in this thread - is a major factor. I don't think there
> are people who would be devising a major new Lisp if only
> they weren't "waiting for Arc".
No, I don't think so either. We're "waiting for arc" in the same
way that a kid might wait to see what new toys come out next year;
there's anticipation, but it's not inhibiting anything. I'm full-
tilt ahead on several projects. I might use a different Lisp if
it was here, but it's not so I'm not.
> > People keep waiting, I think, for some luminary, someone who is well-
> > known in the community, someone whose work will attract a critical mass
> > of attention, to be the one to do it. It will simply take that big a
> > force to develop a new direction in Lisp code. The list of names is
> > short: Steele, Sussman, McCarthy, Moon, Graham, and Knuth are six
> > acknowledged living Wizards; ...
>
> Something like that happened a while back. There was a new language
> that built on the accomplishments of Common Lisp and Scheme, as well
> being influenced by other languages. A number of well-known Lisp
> people were involved, and there was a lot of excitement and interest.
>
> That language was Dylan.
>
> Dylan still exists, but it failed to become the breakthrough
> language that it had once seemed it might be. It also moved
> further from the Lisp center, especially by dropping the
> parenthesized syntax, in an effort to widen its appeal.
As far as I'm concerned Dylan dropped the ball. When you give up
code representable as data, or data that can be interpreted as
code, or at the very least code and data with a common printed
representation, it isn't a lisp anymore. Dylan was a cool thing
for getting some bits of lisp power out to other languages, but
it was a step backward from lisp. I read the manual when it came
out, but decided it wasn't worth my time to learn. It wasn't
low-level enough to replace C as my bit-twiddling and binary-API-
using language of choice and it didn't have the macrology to
compete with Lisp (scheme at the time) as a language-of-choice
for any other jobs I was interested in doing. Finally, it didn't
offer any fundamentally new ideas (like FORTH, APL, or Prolog for
example) that would make it worth learning to sharpen my head;
everything it had was already familiar from Lisp.
> > So, since one of the acknowledged wizards appears to be working on a new
> > Lisp, people are buzzing and muttering and wondering what it's going to
> > be like.
>
> Are they? And who are the people who are waiting for Arc, buzzing,
> muttering? This whole thread seems quite strange to me.
I for one am anticipating Arc. I'm not stopping work; I'm not
refraining from anything or doing anything different because of it;
but I'm curious about it, and if/when it appears, I will check it out.
There are still a couple of places where Lisp hasn't gone (or at least
where no lisp I've used yet has gone other than a prototype); first-
class macros that can be defined and computed at runtime for example.
It's insane, in terms of performance; but no more insane than prolog,
and potentially just as enlightening about new ways to look at the
world. Will Arc have 'em? I don't know. Are they a good idea? I
don't know that too. But I'm curious, and aware that there are ways
a language can be even more powerful than existing lisps.
Bear
Jeff Dalton wrote:
> The main reasons there hasn't been a major new Lisp are that there
> aren't enough people with the time and funding to do it, that Common
> Lisp and Scheme are good enough that there isn't much pressure to
> develop a new language, and that when people do develop new languages
> not too far from Lisp (Slate is a recent example), they prefer to
> mix in features from other languages.
No language community [he baldly asserted] is composed of a majority who
think their tools aren't good enough. When PL/1 and C were developed,
there was not a chorus of Fortran programmers everywhere saying
C FINALLY SOME BETTER TOOLS FOR STRUCTURED PROGRAMMING
When learning a language, people turn to sources which tend to be
relentlessly upbeat on that language's features, are often working with
other users of that language (who have a vested interested in not saying
"my tools is broke"), etcetera. Once they've learned it, admitting it's
crap is an admission that they've chosen badly and wasted their time,
and tends also to have the effect of insulting their associates.
So almost by definition, new language developers tend to be iconoclasts,
except in the rare case where they're being directed to develop a new
language, maybe by IBM, Sun or the US military. I get the impression
that most language architects aren't paid to do it. Were Larry Wall and
K&R paid to develop Perl and C? It seems that they were being paid at
the time, but not specifically as language developers.
> I don't think "waiting for Arc" - which is what's claimed in this
> thread - is a major factor. I don't think there are people who would
> be devising a major new Lisp if only they weren't "waiting for Arc".
That's twice that you've used the phrase "major new Lisp." Writing my
original post, I wasn't really thinking of the case of some lone gun
going public with a radical new and complete product of two years' toil,
more of independent, casual experimenters playing with syntax and
semantics around the margins, and one of them coming up with
something really good.
Was Scheme considered a major new lisp when it was first presented? Some
wags might call it a minor lisp even now, and the original interpreter
has been described as a toy.
--
Cameron MacKinnon
Toronto, Canada
Cameron MacKinnon <··········@clearspot.net> writes:
> Jeff Dalton wrote:
> > I don't think "waiting for Arc" - which is what's claimed in this
> > thread - is a major factor. I don't think there are people who would
> > be devising a major new Lisp if only they weren't "waiting for Arc".
>
> That's twice that you've used the phrase "major new Lisp." Writing my
> original post, I wasn't really thinking of the case of some lone gun
> going public with a radical new and complete product of two years' toil,
> more of independent, casual experimenters playing with syntax and
> semantics around the margins, and one of them coming up with
> something really good.
But people are experimenting with syntax and semantics around
the margins. I assumed you must be thinking of something larger;
you seemed to be talking about something at the level of a new
language or dialect, and it sounded like minor variations on existing
languages or dialects weren't relevant.
Of course, people are developing new languages, some quite close
to Lisp; but I didn't want to get into a debate about whether
they counted as "Lisp" or not.
I didn't say anything about a lone gun developing a radical new and
complete product.
There's a lot of provocative rhetoric in this thread; for instance:
The older generation has turned to threats of violence, bribery,
funding by subscription, and ceaseless imploring. The new converts
to Lisp pick up the tune, though with more deference -- their elders
await Arc with eager anticipation, so they do likewise.
But let's see if it's justified. Who are these people who play at
"Waiting for Arc"?
Name names. You said "here, everyone", so it should be easy.
> Was Scheme considered a major new lisp when it was first presented?
That doesn't matter. Or do you want to switch to saying that people
are playing Language Architect, not Waiting for Arc, but their work
hasn't yet seemed "major"?
-- jd
? Scheme had a large effect on the subsequent development
of Lisp and other languages.
Jeff Dalton wrote:
> But people are experimenting with syntax and semantics around
> the margins. I assumed you must be thinking of something larger;
> you seemed to be talking about something at the level of a new
> language or dialect, and it sounded like minor variations on existing
> languages or dialects weren't relevant.
>
> Of course, people are developing new languages, some quite close
> to Lisp; but I didn't want to get into a debate about whether
> they counted as "Lisp" or not.
People who post to c.l.l with questions about experiments (most about
syntax and few about semantics) are discouraged with the party line
about the local maximum, and how alternates have been tried and found
wanting. This, in combination with the buzz about Arc, gives the
impression that you can experiment with new lisps, but only if you're a
wizard. And *that* goes against the attitude that used to reign in the
lisp world. [Note that Ray Dillinger, elsewhere in this thread, gave an
excellent exposition of how too much diversity in the past has led to
the current attitude.]
Maybe, rather than discouraging these experimenters, we should say
"We're Strict Constructionists here, but experimentation is freely
encouraged over in comp.lang.scheme."
> I didn't say anything about a lone gun developing a radical new and
> complete product.
Well, you said "major new Lisp", which to me precludes mere tinkering.
And I haven't seen a lot of "I'm working on something big and should
have a release in two months" type posts. So if someone IS working on
something major, we have to assume it's a secret right now.
But I take issue with the term. Scheme was major because of its effect.
The original papers introduced it as a series of simple experiments
(though making it look easy is a hallmark of genius), not as anything major.
> There's a lot of provocative rhetoric in this thread; for instance:
>
> The older generation has turned to threats of violence, bribery,
> funding by subscription, and ceaseless imploring. The new converts
> to Lisp pick up the tune, though with more deference -- their elders
> await Arc with eager anticipation, so they do likewise.
>
> But let's see if it's justified. Who are these people who play at
> "Waiting for Arc"?
>
> Name names. You said "here, everyone", so it should be easy.
It would be a long list, and I would doubtless slight someone through
omission. I spent time in the archives to confirm my impressions before
my OP. Lest this now devolve into a bunfight over the term "waiting",
I'll assert that if a reasonable person could infer from a poster's
remarks that the poster is likely to download Arc come the Day, the
poster is waiting.
http://groups.google.com/groups?as_q=arc&as_ugroup=comp.lang.lisp&as_scoring=d&num=100
One could even study the subset of posts which contain an aside to Mr.
Graham about Arc. The device is approaching cliche status.
From the tone of your post, I can tell that I've raised your hackles.
Let me say that I did research to confirm my impressions before posting,
and that I knew I was likely to offend some sensibilities, an effect
that I tried to address in my OP.
I hope that discussion about the issue will lead to somewhat less
growling in future at posters who express an urge to experiment with
language. Experimentation is what got us here, and it would be a shame
if the "local maximum" crowd managed to scare off the next generation of
tinkerers.
--
Cameron MacKinnon
Toronto, Canada
Cameron MacKinnon <··········@clearspot.net> wrote:
> It would be a long list, and I would doubtless slight someone through
> omission. I spent time in the archives to confirm my impressions before
> my OP. Lest this now devolve into a bunfight over the term "waiting",
> I'll assert that if a reasonable person could infer from a poster's
> remarks that the poster is likely to download Arc come the Day, the
> poster is waiting.
Well, your definition of 'waiting' is pretty important to the
discussion. If it means "interested enough in what a luminary like Paul
Graham is doing to read what he says about it and try it out soon after
he pronounces it ready for public consumption," then you'll include a
large percentage of people in c.l.l., perhaps even a majority.
But that's probably also been true of Cells, CL-typesetting, CL-PDF,
CL-PPCRE or any number of other potentially broadly useful libraries or
paradigm experiments, even when their authors haven't been as well known
as PG.
I'd argue that this level of 'waiting' only implies interest, and not
any stoppage of interesting work they might be doing on their own.
Your initial post wondered if people were failing to do interesting work
*because* the community was "waiting for Arc". The weaker you define
'waiting' the easier it is to produce a large chunk of lispers who are
"waiting for Arc", but the harder it is to believe that this 'waiting'
has any affect on their code or idea output.
I'm probably going to download Arc and try it when it comes out. But
the lack of Arc is not stopping me from writing lisp, and it isn't
stopping me from doing language experimentation in lisp. I'm not doing
any language experimentation in lisp because I'm not good enough at it
yet to believe that I could improve on the existing language and even
when I get there, I'm not sure it's worth the large effort involved to
create something fundamental enough to be a dialect change (as opposed
to domain-specific sub-languages and libraries), given that CL is pretty
darned good right now.
That said, if someone else whom I respect has gone to the trouble of
trying, it probably deserves a look. And if when I see it, I actually
like it as well or better than CL, I might use it, or I might have
suggestions about how to improve it.
Since this will probably always be true, I think you could even
characterize me as "waiting", in your sense, for a new dialect even if
Arc were not a proposal and nothing else was in its place.
The question is whether this distinguishes my current actions in any way
from someone who is not so "waiting" -- who has no interest in Arc for
whatever reason.
I don't think it does.
> From the tone of your post, I can tell that I've raised your hackles.
> Let me say that I did research to confirm my impressions before posting,
> and that I knew I was likely to offend some sensibilities, an effect
> that I tried to address in my OP.
> I hope that discussion about the issue will lead to somewhat less
> growling in future at posters who express an urge to experiment with
> language. Experimentation is what got us here, and it would be a shame
> if the "local maximum" crowd managed to scare off the next generation of
> tinkerers.
It seems to me that your "local maximum" crowd are those least likely to
give a hooey about Arc. I've read a lot of people here who seem to
think PG is tilting at windmills with Arc, that he's merely one of many
who've gone down the road of trying to build a better lisp, the results
always falling somewhere on the spectrum between complete failure and
limited success (a dialect which fails but gets a good idea or two
adopted elsewhere, or sticks but is much less popular than CL or
Scheme).
Those people are the very ones who aren't "waiting" for Arc at all,
except perhaps waiting for it to finally come out and flop so that
newbies can tout some other piece of vaporware as the answer to "what's
wrong with lisp?"
Michael
Cameron MacKinnon wrote:
> People who post to c.l.l with questions about experiments (most about
> syntax and few about semantics) are discouraged with the party line
> about the local maximum, and how alternates have been tried and found
> wanting.
You call thirty plus years (leading up to the standardization on CL) of
trying alternatives "local"? And they were exploring with the most
exploration-amenable of languages. Local, schmocal.
This, in combination with the buzz about Arc, gives the
> impression that you can experiment with new lisps, but only if you're a
> wizard.
c.l.l. is stopping you? The only way c.l.l. stops me from doing anything
is by sucking up time I could be spending on Cello or Symbolic Composer.
You are blaming c.l.l. for your not having any ideas to put into code,
just as you blame the absence of Arc. Do some bindings to OpenAL, will
you? I got sidetracked with a neat little Symbolic Composer project
supporting a composer in the music sig of Lisp-NYC. (Macros to the rescue!)
And *that* goes against the attitude that used to reign in the
> lisp world. [Note that Ray Dillinger, elsewhere in this thread, gave an
> excellent exposition of how too much diversity in the past has led to
> the current attitude.]
No, it led to a painful process in which everyone agreed to sit down and
settle on which Lisp ideas to standardize, to eliminate fragmentation.
Ideas good, fragmentation baaaaad! So what are we missing? What have
Java and Python and even Dylan come up with that we'd love to have.
Sounds like Graham is just throwing things out. Scheme tried that and
failed by being a recipe for fragmentation (because they threw out
things people wanted and so had to re-invent each their own way.)
>
> Maybe, rather than discouraging these experimenters, we should say
> "We're Strict Constructionists here, but experimentation is freely
> encouraged over in comp.lang.scheme."
Sure, or Python. Or just spam c.l.l. for a couple of years (Cells!) and
eventually people will go to your web site to see exactly how crazy you
are. But it takes a couple of years because the group just wants to see
if you are serious.
kt
--
Home? http://tilton-technology.com
Cells? http://www.common-lisp.net/project/cells/
Cello? http://www.common-lisp.net/project/cello/
Why Lisp? http://alu.cliki.net/RtL%20Highlight%20Film
Your Project Here! http://alu.cliki.net/Industry%20Application
Kenny Tilton wrote:
> c.l.l. is stopping you? The only way c.l.l. stops me from doing anything
> is by sucking up time I could be spending on Cello or Symbolic Composer.
> You are blaming c.l.l. for your not having any ideas to put into code,
> just as you blame the absence of Arc.
Nowhere did I claim that c.l.l is stopping me from doing anything,
though your "sucking up time" comment hits painfully close to home. If I
wanted to talk about me I'd use the personal pronoun, as the royal we
seems to rub people the wrong way. When I talk about "posters" I'm not
talking about the voice in my head that fills this space Tuesdays,
Thursdays and Saturdays :-)
> Ideas good, fragmentation baaaaad! So what are we missing? What have
> Java and Python and even Dylan come up with that we'd love to have.
> Sounds like Graham is just throwing things out. Scheme tried that and
> failed by being a recipe for fragmentation (because they threw out
> things people wanted and so had to re-invent each their own way.)
Pointing to other languages which haven't come up with anything you want
to borrow is not the same as proving that no future language will
contribute anything novel.
For a Lisper to crank out screen after screen of code which depends on
lexical scope, stopping occasionally to post "Scheme's a failure"
comments to c.l.l seems the height of ingratitude.
I'd suggest, only half jokingly, that Java and Python have come up with
something that we'd love to have: the adulation of the masses. Some
elitist grouch will no doubt rejoin that he doesn't want Lisp to be
popular, but sentiment here says otherwise. Anyway, I don't know how to
bottle the certain je ne sais quoi that Java and Python have, or had.
All of the above to be taken in the same spirit as Kenny's post!
--
Cameron MacKinnon
Toronto, Canada
Kenny Tilton wrote:
>
>
> Cameron MacKinnon wrote:
>
> > People who post to c.l.l with questions about experiments (most about
> > syntax and few about semantics) are discouraged with the party line
> > about the local maximum, and how alternates have been tried and found
> > wanting.
...
...
> c.l.l. is stopping you? The only way c.l.l. stops me from doing anything
> is by sucking up time I could be spending on Cello or Symbolic Composer.
> You are blaming c.l.l. for your not having any ideas to put into code,
> just as you blame the absence of Arc.
I don't think it's as simple as that. Community feedback, interaction, and
reinforcement can help make things happen in a big way. CL conservatism
could very well be inhibiting those kinds of things.
> Sure, or Python. Or just spam c.l.l. for a couple of years (Cells!) and
> eventually people will go to your web site to see exactly how crazy you
> are.
Seems to be working, judging by what Ray Dillinger wrote elsewhere in this
thread:
> If Kenny Tilton announced that his next project after Cello would be
> a lisp system that fixed some of the things he found annoying about
> CL, it would at least get people to notice.
Now we know the real reason no-one's churning out new Lisps - they're all
waiting for KennyLisp (KL)!
Anton
Anton van Straaten wrote:
> Seems to be working, judging by what Ray Dillinger wrote elsewhere in this
> thread:
>
>
>>If Kenny Tilton announced that his next project after Cello would be
>>a lisp system that fixed some of the things he found annoying about
>>CL, it would at least get people to notice.
I saw that. I definitely gotta ease up on the self-promotion.
> Now we know the real reason no-one's churning out new Lisps - they're all
> waiting for KennyLisp (KL)!
Coming soon, as in five years. It will be the same as Common Lisp, except:
(- 42)
=> 42
And you'll be able to use that for everything.
kt
--
Home? http://tilton-technology.com
Cells? http://www.common-lisp.net/project/cells/
Cello? http://www.common-lisp.net/project/cello/
Why Lisp? http://alu.cliki.net/RtL%20Highlight%20Film
Your Project Here! http://alu.cliki.net/Industry%20Application
Cameron MacKinnon wrote:
>
>
> People who post to c.l.l with questions about experiments (most about
> syntax and few about semantics) are discouraged with the party line
> about the local maximum, and how alternates have been tried and found
> wanting. This, in combination with the buzz about Arc, gives the
> impression that you can experiment with new lisps, but only if you're a
> wizard. And *that* goes against the attitude that used to reign in the
> lisp world. [Note that Ray Dillinger, elsewhere in this thread, gave an
> excellent exposition of how too much diversity in the past has led to
> the current attitude.]
>
Many experimental suggestions and observations in c.l.l are plainly naive
and misguided. A lot of "suggestion" come from inexperienced people
with as little as a few weeks experience with ANY Lisp. Questions and
suggestions with some depth and insight are treated with the utmost interest
and examination. That many people here pay ATTENTION and give grounded and
fair comment is a testament to the mindset of those attracted to Lisp. If
your idea has merit it will be gleefully adopted. But when someone says,
"that's been tried before, its a dead end, its not worth it", try to learn
from what they are saying. They give more than enough clues as to what is
wrong. That you are discouraged about the ease which some questions are
shot down, that is human nature.
The attitude that used to reign in the Lisp world was caused by the interaction
of PEERS. People on a equal experiential and contextual footing. This
is no longer the case. You now have people who have decades of difference
in experience and knowledge. People may feel more welcome in the Python
world because it is newer, people are more on a equal footing (mostly
inexperienced). Maybe they feel better because they feel smarter. Maybe
they get the stuffing kicked out of them in the Lisp world because of the
sudden realization that they really do not know much. Its up to
you which kind of world you want to be in and which psychological defense
mechanism you want to put up.
> Maybe, rather than discouraging these experimenters, we should say
> "We're Strict Constructionists here, but experimentation is freely
> encouraged over in comp.lang.scheme."
>
Its easy to label people as "Strict Constructionists" as a way
to control people. Get a grip.
Wade
Wade Humeniuk wrote:
> Many experimental suggestions and observations in c.l.l are plainly naive
> and misguided. A lot of "suggestion" come from inexperienced people
> with as little as a few weeks experience with ANY Lisp. Questions and
> suggestions with some depth and insight are treated with the utmost
> interest
> and examination. That many people here pay ATTENTION and give grounded and
> fair comment is a testament to the mindset of those attracted to Lisp. If
> your idea has merit it will be gleefully adopted.
If these naive and misguided souls were told "that's an interesting
idea, perhaps you could try implementing it in a manner similar to SICP
or The Art of the Interpreter, and report back to us," it might foster
an attitude of exploration and experimentation, and who knows what new
discoveries that would lead to.
Also, the truly clueless might be so intimidated by the suggested
materials, and embarrassed by the admission that they don't understand
them, that they wouldn't bother the newsgroup again. But I wouldn't get
my hopes up. :-)
As to that bit about genuinely insightful questions being treated with
respect, I've seen more than one quick-draw response, eventually
followed by "Oh, sorry. Thought you were a troll."
> But when someone says,
> "that's been tried before, its a dead end, its not worth it", try to learn
> from what they are saying. They give more than enough clues as to what is
> wrong. That you are discouraged about the ease which some questions are
> shot down, that is human nature.
Suddenly you're admonishing *me* to learn and suggesting that *I'm*
discouraged. I would ask that you reread my posts, as you seem to have
misread them.
> The attitude that used to reign in the Lisp world was caused by the
> interaction
> of PEERS. People on a equal experiential and contextual footing. This
> is no longer the case. You now have people who have decades of difference
> in experience and knowledge.
Taken with the rest of this post, you seem to be suggesting that back in
the halcyon days, nobody ever came up with a stupid question. I wasn't
there, but I know enough about human nature to know that:
1. The group doesn't have to be very big, nor the timespan very long,
before a dumb question pops out. Limiting the group to Ivy League grad
students doesn't have as much of an effect as you might think.
2. If people aren't coming up with a certain percentage of stupid
questions, it's because they're not asking enough questions.
Further, the reason we know that some things are misguided is that they
were tried and found wanting. The original posers had to try them.
>> Maybe, rather than discouraging these experimenters, we should say
>> "We're Strict Constructionists here, but experimentation is freely
>> encouraged over in comp.lang.scheme."
>>
>
> Its easy to label people as "Strict Constructionists" as a way
> to control people. Get a grip.
If you have more information on how I can control complete strangers in
faraway places through carefully crafted Usenet posts, please send
details! Oops, gotta go... I've got a thread on line two where the
poster points out that some people don't consider Scheme a Lisp.
--
Cameron MacKinnon
Toronto, Canada
Cameron MacKinnon <··········@clearspot.net> writes:
> If these naive and misguided souls were told "that's an interesting
> idea, perhaps you could try implementing it in a manner similar to SICP
> or The Art of the Interpreter, and report back to us," it might foster
> an attitude of exploration and experimentation, and who knows what new
> discoveries that would lead to.
I would respectfully suggest those souls to try implementing the
outstanding items listed in the FIXME/TODO comments of projects such
as SBCL or McCLIM. I don't know about new discoveries, but the new
features or improved functionality would add up, with benefits for
both the projects and those souls.
Paolo
--
Why Lisp? http://alu.cliki.net/RtL%20Highlight%20Film
Cameron MacKinnon <··········@clearspot.net> writes:
> [Note that Ray Dillinger, elsewhere in this thread,
> gave an excellent exposition of how too much diversity in the past has
> led to the current attitude.]
I don't agree with that part of his account.
Or with your characterization of the current attitude.
> Well, you said "major new Lisp", which to me precludes mere
> tinkering.
Because *you* appeared to be talking about more than "mere tinkering";
I assumed you knew many experiments around the margins were going on
but didn't think it counted, especially since your contrast with
"waiting for Arc" was "language architect", not "casual tinkerer".
> But I take issue with the term. Scheme was major because of its
> effect. ...
It has its effect for reasons; and it was a major (that word again)
departure from what was then mainstream Lisp. Some people -- people
well-informed about the languages and issues -- even claim that Scheme
isn't a Lisp.
> > There's a lot of provocative rhetoric in this thread ...
> > But let's see if it's justified. Who are these people who play at
> > "Waiting for Arc"?
> > Name names. You said "here, everyone", so it should be easy.
> It would be a long list,
Start with a few of the worst offenders, or the ones with
thicker skins, if you're worried that naming might offend.
> and I would doubtless slight someone through
> omission.
They they can speak up and say "My name is ___ and I am
waiting for Arc".
> I spent time in the archives to confirm my impressions
> before my OP. Lest this now devolve into a bunfight over the term
> "waiting", I'll assert that if a reasonable person could infer from a
> poster's remarks that the poster is likely to download Arc come the
> Day, the poster is waiting.
Your claim seemed to be that they were playing "Waiting for Arc"
rather than Language Architect -- that they were waiting *rather
than* doing language design -- not that they were merely interested
in Arc or were likely to download it.
> From the tone of your post, I can tell that I've raised your
> hackles. Let me say that I did research to confirm my impressions
> before posting, and that I knew I was likely to offend some
> sensibilities, an effect that I tried to address in my OP.
I just don't think it's true. What's the evidence, apart from
your impression of the archives? Who are the "waiters for Arc",
so we can check whether your interpretation of their behaviour is
correct?
> I hope that discussion about the issue will lead to somewhat less
> growling in future at posters who express an urge to experiment with
> language. Experimentation is what got us here, and it would be a shame
> if the "local maximum" crowd managed to scare off the next generation
> of tinkerers.
Fair enough. If the issue is discouraging experimentation in general,
then I agree that it should not be discouraged; but I was addressing
specifically the claims about "waiting for Arc".
-- jd
Jeff Dalton wrote:
> Cameron MacKinnon <··········@clearspot.net> writes:
>>But I take issue with the term. Scheme was major because of its
>>effect. ...
>
> It has its effect for reasons; and it was a major (that word again)
> departure from what was then mainstream Lisp. Some people -- people
> well-informed about the languages and issues -- even claim that Scheme
> isn't a Lisp.
That's a whole other thread, and an unproductive one at that. There was
a brief period (perhaps only hours) when I thought that Lisp-1, Lisp 1.5
and Lisp-2 were three early Lisp systems. :-)
Suffice it to say that, while I can concede some of the points in the
Scheme is Not a Lisp (SNL) doctrine, I consider the schism
counterproductive given the already low population of users in the
lisp-like-languages archipelago.
In the context of the current topic, I suspect that SNL people would
generally be hostile to experimentation, on the theory that if SNL, then
one can't experiment too much without creating another heresy. But
that's just a guess -- maybe SNL people don't mind what people build, as
long as it isn't claimed to be a Lisp.
>>I spent time in the archives to confirm my impressions
>>before my OP. Lest this now devolve into a bunfight over the term
>>"waiting", I'll assert that if a reasonable person could infer from a
>>poster's remarks that the poster is likely to download Arc come the
>>Day, the poster is waiting.
>
>
> Your claim seemed to be that they were playing "Waiting for Arc"
> rather than Language Architect -- that they were waiting *rather
> than* doing language design -- not that they were merely interested
> in Arc or were likely to download it.
Well, there's multiple groups.
1. There's the "local maximum" crowd. These people are happily coding in
Lisp, and don't care much about Arc, which they see as "vapourware" or
"tilting at windmills". They are usually too polite, deferential or
uninterested to sling mud at Arc, though it occasionally happens. They
WILL, however, likely be the first to respond when someone new shows an
interest in alternate syntax, with "Was tried. Sucked. Don't bother."
2. There's the pragmatic but happy Lispers. Coding in Lisp, but will try
Arc if it appears. They exchange all the latest Arc gossip and news with
each other, and will cheerfully answer strangers' Arc questions. These
are the ones who are always making asides to Mr. Graham. I don't think
these people would be playing language god themselves if it weren't for
Arc, but maybe some of them would.
3. It is the interaction, no, parallel action of the above two groups
which creates the group I describe. People relatively new to Lisp who
are simultaneously discouraged from their own language experiments and
told that Arc is just around the corner. Given that there's respectable
members of the community in the adherents of both of the above camps,
the effect on the new reader to c.l.l is powerful.
To round out the guidebook, but otherwise apropos of nothing, I'll add a
species others have sighted, and whose existence I can now confirm, but
which hasn't figured in my population dynamics studies to date:
4. The Gadfly Malcontent: Evaluating existing languages for his Next Big
Project, he shows up in the newsgroup with a confident swagger and a
moderate list of indispensable features. As the suggestions roll in, he
starts adding to the requirements or raising previously unvoiced
objections, leading to the suspicion that he's just looking for excuses
to delay the start of coding. Confirmation comes when he's observed in
the exact same behaviour in other newsgroups.
> I just don't think it's true. What's the evidence, apart from
> your impression of the archives? Who are the "waiters for Arc",
> so we can check whether your interpretation of their behaviour is
> correct?
Well, you've found me out. I don't have an existence proof, someone who
has been discouraged from conducting his own language experiments, but
is also waiting for Arc. I just suspect they exist, based on list
dynamics I've observed. Anyone here still reading who'd like to stand up
and be counted???
The Google link I included in the last post had lots of evidence of
group 2, and of strangers who inquire about Arc. It also had posts from
group 1, that's where my quotes are from.
The interaction between group 1 and people keen on playing with language
wasn't documented in the Google search, but I suspect you've seen that.
If it appears that my position has evolved somewhat since my OP, it
probably has, but only a little. Some of the responses got me thinking
and researching a bit more, but I stand by my original post.
Personally, I'd hope that every language experimenter, no matter how
naive, was encouraged to experiment in Lisp. From the point of view of
spreading Lisp to third parties and demonstrating the benefits of
writing languages in Lisp, I think poor language X, done in Lisp is
better than if done in C, or not at all. And one of those experimenters
might hit on something good.
--
Cameron MacKinnon
Toronto, Canada
"Cameron MacKinnon" <··········@clearspot.net> wrote in message
···························@golden.net...
> 1. There's the "local maximum" crowd. These people are happily coding in
> Lisp, and don't care much about Arc, which they see as "vapourware" or
> "tilting at windmills". They are usually too polite, deferential or
> uninterested to sling mud at Arc, though it occasionally happens. They
> WILL, however, likely be the first to respond when someone new shows an
> interest in alternate syntax, with "Was tried. Sucked. Don't bother."
>
> 2. There's the pragmatic but happy Lispers. Coding in Lisp, but will try
> Arc if it appears. They exchange all the latest Arc gossip and news with
> each other, and will cheerfully answer strangers' Arc questions. These
> are the ones who are always making asides to Mr. Graham. I don't think
> these people would be playing language god themselves if it weren't for
> Arc, but maybe some of them would.
>
> 3. It is the interaction, no, parallel action of the above two groups
> which creates the group I describe. People relatively new to Lisp who
> are simultaneously discouraged from their own language experiments and
> told that Arc is just around the corner. Given that there's respectable
> members of the community in the adherents of both of the above camps,
> the effect on the new reader to c.l.l is powerful.
Here's what I see from all of this.
You have two camps, those in the Common Lisp camp, and those outside of it.
You can be inspired by Common Lisp, based on Common Lisp, taking ideas from
Common Lisp, etc, but you won't be Common Lisp when you're done.
The reason is that if your final result used Common Lisp as a core, then
you're still Common Lisp.
Kenny's Cells is Common Lisp. Uncommon SQL with its embedded SQL reader
macros is Common Lisp. The LOOP Macro is Common Lisp. Uncommon Web, with its
limitations and limited CPS macros, is Common Lisp. CLOS is Common Lisp.
Garnets KR is Common Lisp.
But, for example, AspectJ, is not Java. Java like, Java inpired, sorta looks
like Java, Works with Java, but not Java. They could have done Aspect C++,
or AspectPython, or AspectIntercal.
Why the distinction? They had to go outside the domain to implement it. If I
throw an AspectJ program at a system that "knows Java", it will go "WTF?
THIS isn't Java."
You can not extend the Java language.
The built in Common Lisp compiler compiles all of the previous contstructs.
AspectJ needed its own compiler.
It is said that the lathe is the only machine that can actually reproduce
itself. That's how Common Lisp is. It can reproduce itself, expand itself,
grow itself.
Once you've defined a new Common Lisp construct, the borg that is CL absorbs
and consumes it, and makes it one with itself.
You'd like to think that the Scheme's of the world are similar, but, going
on the most base RxRS versions, the language itself isn't rich enough to
extend itself. A simple example is things like structures/records. Basically
in Scheme, they're implemented as either lists or arrays, with a layer of
macros on top of it. But if you ask the type of the object, you won't get
structure/record, you inevitably get list or array. It's difficult to make
the new type opaque to the system.
So, while many packages exist to add structures to Scheme, most
implementations pick a representation and build it into the system, into the
compiler, which is typically NOT written in Scheme. The point here is while
Scheme is a dynamic and extensible language, I don't think that Standard
gives enough of a foundation and most implementors must move beyond it.
Something like Arc COULD be written on top of CL. A more succinct system,
shorter names, even its own reader and listener. Then, Arc becomes a domain
specific extension of CL, like, perhaps, the HTML exstensions, but you
always have CL if you ever needed it. Run into an brick wall? You can fall
down into CL and fix it.
But, Arc doesn't want to be a CL macro, package, and library layer. Best to
reinvent the entire wheel.
Lisp has reinvented itself, and been renewed with changes that can only be
done by restarting from scratch, and that's fine, all hands on deck, balls
out, full steam ahead. But every proposal regarding CL I've seen wants to
toss out the baby, bathwater, basin, sponge and soap to get their new idea.
They want a sorta kinda maybe thing like a little almost Common Lisp.
"I love Common Lisp, that's why I want to toss half of it out and restart
the other half from scratch to I can slowly rebuild it to 75% of what Common
Lisp was before I started!"
I'd be more interested in some hacker that wants to implement something like
Dylans sealing into a CLOS-esque system by hacking CLOS code and the CMUCL
compiler. Something like this needs compiler support to gain the actual
benefits, but that doesn't necessarily mean that everything else has to go
with it at the same time. Not that this is a particular pet peeve, but this
is one of the things that Dylan was trying to accomplish and felt they
needed to redo the entire thing (perhaps they did, I don't really know).
Most of the other CL complaints have to do with the implementations and
portable functionalities (sockets, threads) (which single implementation
languages don't have), and holes in implementations (like the MOP -- which,
to be fair, like sockets, isn't in the Standard). Those can be, and have
been, fixed without tossing out CL.
Regards,
Will Hartung
(·····@msoft.com)
"Will Hartung" <·····@msoft.com> writes:
> I'd be more interested in some hacker that wants to implement
> something like Dylans sealing into a CLOS-esque system by hacking
> CLOS code and the CMUCL compiler.
You mean Gerd Moellmann? (see CMUCL CVS)
Christophe
--
http://www-jcsu.jesus.cam.ac.uk/~csr21/ +44 1223 510 299/+44 7729 383 757
(set-pprint-dispatch 'number (lambda (s o) (declare (special b)) (format s b)))
(defvar b "~&Just another Lisp hacker~%") (pprint #36rJesusCollegeCambridge)
Cameron MacKinnon <··········@clearspot.net> writes:
> Jeff Dalton wrote:
> > Your claim seemed to be that they were playing "Waiting for Arc"
> > rather than Language Architect -- that they were waiting *rather
> > than* doing language design -- not that they were merely interested
> > in Arc or were likely to download it.
>
> Well, there's multiple groups.
>
> 1. There's the "local maximum" crowd. These people are happily coding
> in Lisp, and don't care much about Arc, which they see as "vapourware"
> or "tilting at windmills". They are usually too polite, deferential or
> uninterested to sling mud at Arc, though it occasionally happens. They
> WILL, however, likely be the first to respond when someone new shows
> an interest in alternate syntax, with "Was tried. Sucked. Don't
> bother."
>
> 2. There's the pragmatic but happy Lispers. Coding in Lisp, but will
> try Arc if it appears. They exchange all the latest Arc gossip and
> news with each other, and will cheerfully answer strangers' Arc
> questions. These are the ones who are always making asides to
> Mr. Graham. I don't think these people would be playing language god
> themselves if it weren't for Arc, but maybe some of them would.
>
> 3. It is the interaction, no, parallel action of the above two groups
> which creates the group I describe. People relatively new to Lisp who
> are simultaneously discouraged from their own language experiments and
> told that Arc is just around the corner. Given that there's
> respectable members of the community in the adherents of both of the
> above camps, the effect on the new reader to c.l.l is powerful.
That's a more plausible analysis, but very different from what you
said initially.
Before, it was ` here, everybody plays at "Waiting for Arc"' and not
only "people relatively new to Lisp": "their elders await Arc with
eager anticipation, so they do likewise." Arc had a "hypnotic
effect" and there was a "subtle but unmistakable air of waiting
for the Messiah".
Anyway, I wonder how much of this is about syntax. That seems to be
your chief complaint about the `"local maximum" crowd':
the first to respond when someone new shows an interest in alternate
syntax, with "Was tried. Sucked. Don't bother."
I don't think there are any interesting discoveries to me made
in syntax.
(That doesn't mean I think people shouldn't experiment with syntax.)
-- jd
Jeff Dalton wrote:
>
> Anyway, I wonder how much of this is about syntax. That seems to be
> your chief complaint about the `"local maximum" crowd':
>
> the first to respond when someone new shows an interest in alternate
> syntax, with "Was tried. Sucked. Don't bother."
>
> I don't think there are any interesting discoveries to me made
> in syntax.
I don't know if it's "syntax" particularly, but I think many kinds of
data should be callable. So, for example, a string is also a function,
and what the function does depends on its arguments. I can easily
think of a dozen argument combinations that have useful and easily
understandable semantics for functional strings.
Likewise, arrays of all kinds would be functions, as would structs
and possibly even characters (Numbers, I doubt: but someone else may
have different ideas).
Graham appears to be introducing a similar idea in Arc, but he's
limiting it to one or two very restricted usages and treating it
strictly as syntax.
Bear
Ray Dillinger <····@sonic.net> writes:
> Likewise, arrays of all kinds would be functions, as would structs
> and possibly even characters (Numbers, I doubt: but someone else may
> have different ideas).
Arrays "were" functions in MacLisp, weren't they? I bet there's
stuff written about why that wasn't carried over to CL, though
it might very well be worth reconsidering.
'mr
--
[Emacs] is written in Lisp, which is the only computer language that is
beautiful. -- Neal Stephenson, _In the Beginning was the Command Line_
Ray Dillinger <····@sonic.net> writes:
> Jeff Dalton wrote:
> > I don't think there are any interesting discoveries to me made
> > in syntax.
>
> I don't know if it's "syntax" particularly, but I think many kinds of
> data should be callable. So, for example, a string is also a function,
> and what the function does depends on its arguments. I can easily
> think of a dozen argument combinations that have useful and easily
> understandable semantics for functional strings.
I think of that as semantics, because there could be pretty much
any syntax for it. Also, in a Lisp, functionp (or whatever it's
called) should return true for those objects.
You could make it just syntax if the objects were not functions
as objects, but you could still put them in "function positions" in
the syntax. You could think of it as being implemented by
rewriting.
> Likewise, arrays of all kinds would be functions, as would structs
> and possibly even characters (Numbers, I doubt: but someone else may
> have different ideas).
One functional use of positive integers that's been suggested is that
they return the nth element of their argument.
It would be interesting to collect cases. I think someone mentioned
arrays in MacLisp. (I don't recall whether that's correct.) But
arrays were in some ways functions in Franz Lisp, which is fairly
close to MacLisp, and it looks like Franz arrays were meant for
compatibility with MacLisp.
The Pop language (Pop-2, Pop-11, etc) treats more things as
functions than Lisps typically do.
(The Pop languages are much like Lisp but with a different syntax
and with a different "take" on many things. Poplog is a system that
provides a number of languages with a common VM: Pop-11, Common Lisp,
Prolog, and I think ML.)
I'm not sure about all of these, but here are some of the things
that might be functions or usable as functions in Pop-11:
* hash tables
* arrays
* field refs /accessors: obj.f is a way to call f on arg obj.
* generators (?)
For access functions, there's a function that returns an
"updater" function. This is similar to setters in T,
and different from Common Lisp where the operation is
macro-like.
-- jd
Jeff Dalton wrote:
> Ray Dillinger <····@sonic.net> writes:
>>I don't know if it's "syntax" particularly, but I think many kinds of
>>data should be callable. So, for example, a string is also a function,
>>and what the function does depends on its arguments. I can easily
>>think of a dozen argument combinations that have useful and easily
>>understandable semantics for functional strings.
>
> I think of that as semantics, because there could be pretty much
> any syntax for it. Also, in a Lisp, functionp (or whatever it's
> called) should return true for those objects.
Aren't funcallable instances supposed to be used for these things? (-> MOP)
Pascal
--
ECOOP 2004 Workshops - Oslo, Norway
*1st European Lisp and Scheme Workshop, June 13*
http://www.cs.uni-bonn.de/~costanza/lisp-ecoop/
*2nd Post-Java Workshop, June 14*
http://prog.vub.ac.be/~wdmeuter/PostJava04/
Ray Dillinger <····@sonic.net> writes:
> Jeff Dalton wrote:
> >
>
> > Anyway, I wonder how much of this is about syntax. That seems to be
> > your chief complaint about the `"local maximum" crowd':
> >
> > the first to respond when someone new shows an interest in alternate
> > syntax, with "Was tried. Sucked. Don't bother."
> >
> > I don't think there are any interesting discoveries to me made
> > in syntax.
>
> I don't know if it's "syntax" particularly, but I think many kinds of
> data should be callable. So, for example, a string is also a function,
> and what the function does depends on its arguments. I can easily
> think of a dozen argument combinations that have useful and easily
> understandable semantics for functional strings.
If you flip this thought on its side, you'll note that arrays, hash
tables, functions, etc, are all kind of similar, in that they're
subscriptable -- or you can dereference them. I use a GF called REF
for this purpose; a couple years ago, the topic came up here, it turns
out there are some CL'ers happily using such things. Erann recently
mentioned it in a post on another topic.
> Likewise, arrays of all kinds would be functions, as would structs
> and possibly even characters (Numbers, I doubt: but someone else may
> have different ideas).
>
> Graham appears to be introducing a similar idea in Arc, but he's
> limiting it to one or two very restricted usages and treating it
> strictly as syntax.
The syntax is the least appealing part of it, for me. I already have
what I need built on top of CL: REF subscripts arrays, works with
alists, funcalls functions, and is extensible for my own types.
WITH-REF-LABELS lets me use a function-calling syntax:
(defun foo (a b c)
(with-ref-labels (a b c)
(list (a 1) (b 2) (c 3))))
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
Ray Dillinger <····@sonic.net> writes:
> I don't know if it's "syntax" particularly, but I think many kinds of
> data should be callable. So, for example, a string is also a function,
> and what the function does depends on its arguments. I can easily
> think of a dozen argument combinations that have useful and easily
> understandable semantics for functional strings.
I don't see the advantage of this.
(let ((mystring "foo"))
(char mystring 1)) => #\o
(let ((mystring "foo"))
(funcall mystring 1)) => #\o
--
~jrm
Joe Marshall <·············@comcast.net> writes:
> Ray Dillinger <····@sonic.net> writes:
>
>> I don't know if it's "syntax" particularly, but I think many kinds of
>> data should be callable. So, for example, a string is also a function,
>> and what the function does depends on its arguments. I can easily
>> think of a dozen argument combinations that have useful and easily
>> understandable semantics for functional strings.
>
> I don't see the advantage of this.
>
> (let ((mystring "foo"))
> (char mystring 1)) => #\o
>
> (let ((mystring "foo"))
> (funcall mystring 1)) => #\o
It would be nice to have:
(let ((mystring "foo"))
(mystring 1)) => #\o
Yes, I know, this does not work in CL, but you could do something like
this in CL:
(with-string ((mystring "foo"))
(mystring 1))
Writing this simple macro is left as an exercise to the reader.
Regards,
--
Julian Stecklina
Signed and encrypted mail welcome.
Key-Server: pgp.mit.edu Key-ID: 0xD65B2AB5
FA38 DCD3 00EC 97B8 6DD8 D7CC 35D8 8D0E D65B 2AB5
Any sufficiently complicated C or Fortran program
contains an ad hoc informally-specified bug-ridden
slow implementation of half of Common Lisp.
- Greenspun's Tenth Rule of Programming
Joe Marshall wrote:
>
> Ray Dillinger <····@sonic.net> writes:
>
> > I don't know if it's "syntax" particularly, but I think many kinds of
> > data should be callable. So, for example, a string is also a function,
> > and what the function does depends on its arguments. I can easily
> > think of a dozen argument combinations that have useful and easily
> > understandable semantics for functional strings.
>
> I don't see the advantage of this.
>
> (let ((mystring "foo"))
> (char mystring 1)) => #\o
>
> (let ((mystring "foo"))
> (funcall mystring 1)) => #\o
>
(let ((mystring "foo"))
(mystring 1)) =>#\o
was more what I had in mind. Clearly this data-value-is-also-a-function-value
thing is not Common Lisp; but I think it's more expressive and succinct.
Bear
"Ray Dillinger" <····@sonic.net> wrote in message
······················@sonic.net...
> Joe Marshall wrote:
> >
> > Ray Dillinger <····@sonic.net> writes:
> >
> > > I don't know if it's "syntax" particularly, but I think many kinds of
> > > data should be callable. So, for example, a string is also a
function,
> > > and what the function does depends on its arguments. I can easily
> > > think of a dozen argument combinations that have useful and easily
> > > understandable semantics for functional strings.
> >
> > I don't see the advantage of this.
> >
> > (let ((mystring "foo"))
> > (char mystring 1)) => #\o
> >
> > (let ((mystring "foo"))
> > (funcall mystring 1)) => #\o
> >
>
> (let ((mystring "foo"))
> (mystring 1)) =>#\o
>
> was more what I had in mind. Clearly this
data-value-is-also-a-function-value
> thing is not Common Lisp; but I think it's more expressive and succinct.
It is also not a Lisp 2, ie it requires a single namespace. In the above
alternate universe you could no longer do:
(let ((list "My List"))
(list 1))
and get => (1)
--
Coby Beck
(remove #\Space "coby 101 @ big pond . com")
Coby Beck wrote:
>
> "Ray Dillinger" <····@sonic.net> wrote in message
> > (let ((mystring "foo"))
> > (mystring 1)) =>#\o
> >
> > was more what I had in mind. Clearly this
> data-value-is-also-a-function-value
> > thing is not Common Lisp; but I think it's more expressive and succinct.
>
> It is also not a Lisp 2, ie it requires a single namespace. In the above
> alternate universe you could no longer do:
>
> (let ((list "My List"))
> (list 1))
>
> and get => (1)
Nope. You'd get => #\y instead. (or "y" -- characters are a special
case of strings, and might as well be written using string syntax).
I think I'm okay with that. I don't think separate function and data
values are terribly desirable.
Bear
BTW, it's interesting that this thread is complaining that people
aren't being language architects, while over in the thread about
Groovy, there are "who needs it?" complaints because it's too
like Jython and JRuby.
-- jd
OK, I admit it - I'm the guy waiting for Arc.
http://groups.google.com/groups?q=group:comp.lang.lisp+author:Ralph+author:Richard+author:Cook&hl=en&lr=&ie=UTF-8&oe=UTF-8&scoring=d&selm=406490f8.4735048%40newsgroups.bellsouth.net&rnum=1
And even I'm still getting on with my life, learning and programming
in Common Lisp.
A couple of things I'm anticipating:
Graham's writing about languages, some being more powerful than
others, and the "Blub" programmers not seeing what their language is
missing, makes me wonder if Arc is going to have something in it that
will put it above Lisp on the "power ladder".
If Arc is a decent Lisp but without the parentheses, will this be the
language that makes Lisp more popular? You can say "let the cowardly
dunderheads tremble before the ()'s", but don't most of us also say
we'd like to use Lisp more, say at work? If Arc is just as powerful as
Lisp but more palatible to the masses that may become more possible.
Ralph Richard Cook wrote:
> If Arc is a decent Lisp but without the parentheses, will this be the
> language that makes Lisp more popular?
It didn't work with Dylan.
Frankly, syntax is overrated as a determiner of programming language
popularity. What's important is what you can do with the language.
For Lisp (or a hypothetical successor) this means better libraries
and development environments.
Paul
I read up some on ARC and it seems the parentesises stay.
LOGO removes that top level parethesis.
This simplifies the code somewhat without removing power.
On Fri, 02 Apr 2004 05:39:44 GMT, Ralph Richard Cook
<······@bellsouth.net> wrote:
> OK, I admit it - I'm the guy waiting for Arc.
> http://groups.google.com/groups?q=group:comp.lang.lisp+author:Ralph+author:Richard+author:Cook&hl=en&lr=&ie=UTF-8&oe=UTF-8&scoring=d&selm=406490f8.4735048%40newsgroups.bellsouth.net&rnum=1
>
> And even I'm still getting on with my life, learning and programming
> in Common Lisp.
>
> A couple of things I'm anticipating:
> Graham's writing about languages, some being more powerful than
> others, and the "Blub" programmers not seeing what their language is
> missing, makes me wonder if Arc is going to have something in it that
> will put it above Lisp on the "power ladder".
>
> If Arc is a decent Lisp but without the parentheses, will this be the
> language that makes Lisp more popular? You can say "let the cowardly
> dunderheads tremble before the ()'s", but don't most of us also say
> we'd like to use Lisp more, say at work? If Arc is just as powerful as
> Lisp but more palatible to the masses that may become more possible.
--
Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/