I took a look at several of the Lisps mentioned in the FAQ and was a
bit disappointed.
I'm looking for the PC equivalent of Macintosh Common Lisp. That is,
a nice development environment, compilation to native code, graphics
functions, long file names, and a low, low, low student price.
My ideal lisp would also be able to run interpreted code quickly.
Any suggestions? I've tried Harlequin's and ... Franz?'s demo
versions. I'm aware of the CMU CL ports to Linux and *BSD, but I need
something that runs under Windows 95 and lets me create simple
windows and graphics.
Please respond to me via email. If there's interest, I will summarize
the responses in this group.
--Pat
In article <··········@new-news.cc.brandeis.edu> ·····@cs.brandeis.edu (Patrick Tufts) writes:
> From: ·····@cs.brandeis.edu (Patrick Tufts)
>
> I took a look at several of the Lisps mentioned in the FAQ and was a
> bit disappointed.
>
> I'm looking for the PC equivalent of Macintosh Common Lisp. That is,
> a nice development environment, compilation to native code, graphics
> functions, long file names, and a low, low, low student price.
Unfortunately, I don't think there is any good PC equivalent of
Macintosh Common Lisp.
Franz's ACL for Windows still doesn't implement all of CLtL2 (at least
not the last time I looked at it about 18 months ago). The other major
problem with it is that it is HIDEOUSLY EXPENSIVE. For the price of a
single copy of ACL for the PC, we were able to get a site license for
MCL. Franz is blowing a great marketing opportunity here, since they
want more for their software than the typical hardware it will run on!
Digitool has a nice, low-priced version that they use for evangelists
to introduce the product into their companies. Lisp often has a hard
enough time making it into companies without pricing itself out of the
market. As it stands now, the only people who will generally purchase
the Franz ACL package are those who absolutely need to have lisp on the
PC platform. It's not priced to help develop new markets, which is a
major strategic blunder on Franz' part.
Now I suppose the standard Franz rebuttal would point to their free Lisp
product for the PC. In some ways this is potentially an even worse
idea. Since the free lisp doesn't have a compiler, it is likely to help
perpetuate the "Lisp is too slow for real work" perception. To the
extent that that happens, it ends up being a lisp aversion therapy (or
at least anti-marketing) tool.
> My ideal lisp would also be able to run interpreted code quickly.
>
> Any suggestions? I've tried Harlequin's and ... Franz?'s demo
> versions. I'm aware of the CMU CL ports to Linux and *BSD, but I need
> something that runs under Windows 95 and lets me create simple
> windows and graphics.
I'm not familiar with Harlequin's product, so I can't comment on it.
--
Thomas A. Russ, USC/Information Sciences Institute ···@isi.edu
From: Jim Veitch
Subject: Re: good cheap Lisp for Win95?
Date:
Message-ID: <324AADC1.208@franz.com>
I would like to point out an inaccuracy: the free Allegro CL for
Windows is a full functional Lisp with a native machine 32 bit
compiler. There is no interpreter; like MCL everything is compiled on
fly. In independent tests posted to the net not that long ago I was
surprised to see us sometimes running faster than equivalent C++
programs.
Jim Veitch
Franz Inc.
Thomas A. Russ wrote:
> Now I suppose the standard Franz rebuttal would point to their free Lisp
> product for the PC. In some ways this is potentially an even worse
> idea. Since the free lisp doesn't have a compiler, it is likely to help
> perpetuate the "Lisp is too slow for real work" perception. To the
> extent that that happens, it ends up being a lisp aversion therapy (or
> at least anti-marketing) tool.
From: Jim Veitch
Subject: Re: good cheap Lisp for Win95?
Date:
Message-ID: <324AD0AB.18BB@franz.com>
We checked and under Windows our pricing is quite similar to Digitool
for the low price entries (all prices list):
MCL trial version: $320
Allegro CL Lite: free
MCL "developer" version: $595 (3.0) or $720(3.9)
Allegro CL for Windows developer version: $595
We are more expensive than MCL for some variations of the high end
"Professional" offering. But not too different from other PC
development tools (like Delphi or PowerBuilder).
We are very committed to making Lisp succeed. This does take money!
But we feel we deliver high value and have set things up so as to be
able to attract new (non-Lisp) programmers into the fold without making
price an objection. I would guess the MCL guys have done a similar
thing for similar reasons.
Jim.
Thomas A. Russ wrote:
> The other major
> problem with it is that it is HIDEOUSLY EXPENSIVE. For the price of a
> single copy of ACL for the PC, we were able to get a site license for
> MCL. Franz is blowing a great marketing opportunity here, since they
> want more for their software than the typical hardware it will run on!
>
> Digitool has a nice, low-priced version that they use for evangelists
> to introduce the product into their companies. Lisp often has a hard
> enough time making it into companies without pricing itself out of the
> market. As it stands now, the only people who will generally purchase
> the Franz ACL package are those who absolutely need to have lisp on the
> PC platform. It's not priced to help develop new markets, which is a
> major strategic blunder on Franz' part.
Hi!
JV>But we feel we deliver high value and have set things up so as to be
JV>able to attract new (non-Lisp) programmers into the fold without making
JV>price an objection.
Yes, indeed. The Web-Release is really nice. But there is some problem with
it - because of it's limitations, you have to put the source into files and
(load) them. And that's a bit unconvenient even for small experiments ;-)
I think putting in the (save-image) function wouldn't give up too much but
would help to be able to experiment a bit more.
I understand that you put in limitations, but it's a bit hard to loose all
your changes after restart of the system. It's not the Lisp-way-of-life ;-)
Have a look at what some Smalltalk-Vendors do, they sell full products for
private use only (for example IBM with their Visual Age family), or give
away full (but older) versions for free (Parcplace-Digitalk did this with
Smalltalk Express these days). I don't think that so much users of free
versions will sell programs they did write with the free version. And if
they do so - then you actually have some programs that are written in
Common Lisp and need to be maintained. So there will be a future version
sold.
bye, Georg
In article <·············@franz.com>, ···@franz.com wrote:
> We checked and under Windows our pricing is quite similar to Digitool
> for the low price entries (all prices list):
Digitool's price list is at http://www.digitool.com/MCL-price-list.html.
> MCL trial version: $320
> Allegro CL Lite: free
>
> MCL "developer" version: $595 (3.0) or $720(3.9)
> Allegro CL for Windows developer version: $595
PPC MCL 3.9 (with 4.0 upgrade) would cost my company $720.00 .
As an "individual" I would have to pay $355.00 or as a
student $160.00 . These are *full* versions of MCL
(the latter versions without printed docs).
> We are more expensive than MCL for some variations of the high end
> "Professional" offering. But not too different from other PC
> development tools (like Delphi or PowerBuilder).
For a company using Lisp as a commercial development tool $595
for ACL Windows seems not overly expensive. For a student it
is to high (unless the university has a site license).
The Unix stuff usually is way more expensive. I think
in not so distant time Windows 9x and Windows NT
will be the preferred development platforms. So
prices there will have to be lower compared to the Unix
level.
Rainer Joswig
>Now I suppose the standard Franz rebuttal would point to their free Lisp
>product for the PC. In some ways this is potentially an even worse
>idea. Since the free lisp doesn't have a compiler, it is likely to help
>perpetuate the "Lisp is too slow for real work" perception. To the
>extent that that happens, it ends up being a lisp aversion therapy (or
>at least anti-marketing) tool.
>
This is, fortunately, untrue. The free version of Allegro has always
included a compiler. I've used it to run benchmarks ever since it first
shipped, and the results have consistently been within a hair of optimized
C++. (1.25 seconds vs. 0.75 seconds to generate all the primes below
30000 using an inefficient algorithm). The limitations on the product
are: inability to save the image, inability to use compile-file, and an
unspecified heap-size limitation that I've never managed to exceed.
······@lavielle.com (Rainer Joswig) writes:
> For a company using Lisp as a commercial development tool $595
> for ACL Windows seems not overly expensive. For a student it
> is to high (unless the university has a site license).
But I would have thought that a student would be satisfied with the
free version.
- Marty
Lisp Resources: <URL:http://www.apl.jhu.edu/~hall/lisp.html>
In article <··············@rsi.jhuapl.edu>, Marty Hall <····@apl.jhu.edu> wrote:
> ······@lavielle.com (Rainer Joswig) writes:
> > For a company using Lisp as a commercial development tool $595
> > for ACL Windows seems not overly expensive. For a student it
> > is to high (unless the university has a site license).
>
> But I would have thought that a student would be satisfied with the
> free version.
Maybe.
Until he wants to do a bit more than his CS1 homework. We need
to encourage these people.
Greetings,
Rainer Joswig
Hi!
MH>But I would have thought that a student would be satisfied with the
MH>free version.
No - it's not that much usable because you can't at least save your image -
you have to reload a program every time you want to use it. That adds much
overhead to the startup time of ACL. I think the limitations of the free
version are a bit to restrictive. I can live without compile-file (since
the system compiles on loading of sources), but to be unable to save-image
is really a problem. I can't say how limiting the limited heap really is,
because I didn't do any bigger programs with it (mainly because my notebook
is a bit limited in memory).
bye, Georg
······@lavielle.com (Rainer Joswig) writes:
> Marty Hall <····@apl.jhu.edu> wrote:
>
> > ······@lavielle.com (Rainer Joswig) writes:
> > > For a company using Lisp as a commercial development tool $595
> > > for ACL Windows seems not overly expensive. For a student it
> > > is to high (unless the university has a site license).
> >
> > But I would have thought that a student would be satisfied with the
> > free version.
>
> Maybe.
>
> Until he wants to do a bit more than his CS1 homework. We need
> to encourage these people.
Well, I'm primarily a UNIX user, not a Windows-95 user, so I can't
speak from direct experience. But I was under the impression that the
free version, since it had a compiler and a quite large stack-size limit
was fine for even reasonably complex programs (just not huge
industry-sized projects). Can anyone who has used it speak to this?
- Marty
Lisp Resources: <URL:http://www.apl.jhu.edu/~hall/lisp.html>
Marty Hall <····@apl.jhu.edu> wrote:
>I was under the impression that the
>free version, since it had a compiler and a quite large stack-size limit
>was fine for even reasonably complex programs (just not huge
>industry-sized projects). Can anyone who has used it speak to this?
>
I used ACL windows free version for an MSc Project having found XLisp too restricted
and byte code compiled C-Lisp too slow.
The program included the calculation of a Markov chain for 150 states, and a path
search of the same number of nodes and even on my ancient 25mhz 386 with 16mb and
windows 3.1 the optimised code ran OK if a little slow.
David Clark