Hi,
Since I'm a lisp fan, I thought I'd try my hand at getting a decent free
common lisp environment running on my Linux box running Red Hat 6.0. So
far, I've had no success. From what I've read on c.l.l., it seemed like
clisp and cmucl were the best bets. The clisp home page (clisp.cons.org)
does not appear to be accessible, while the cmucl sources don't seem to
have decent documentation explaining how to build the program from sources,
and I couldn't get the binaries to work either. This is quite frustrating
since I'm used to just having to do "configure --prefix=<whatever" followed
by "make" and "make install". BTW I did get Allegro's free version of ACL
going without problems, but the first program I tried it on (a music
typesetting program) overran the (artificial) heap limitations before it
even finished loading, which was off-putting. I'd rather stick to an
open-source product if possible.
If anyone has any tips that could make installation easier for me (and
others in my situation), I'd appreciate hearing them.
Thanks,
Mike
--------------------------------------------------------------
Mike Vanier ·······@bbb.caltech.edu
Department of Computation and Neural Systems, Caltech 216-76
GNU/Linux: We can't lose; we're on a mission from God.
Michael Vanier <·······@bbb.caltech.edu> writes:
> Hi,
>
> Since I'm a lisp fan, I thought I'd try my hand at getting a decent free
> common lisp environment running on my Linux box running Red Hat 6.0. So
> far, I've had no success. From what I've read on c.l.l., it seemed like
> clisp and cmucl were the best bets. The clisp home page (clisp.cons.org)
> does not appear to be accessible,
I have sporadic problems with the site, but it always comes back.
I was able to install CLISP without too many problems. I'm on 5.1 and
had to get the source rpm because the binaries rpm is linked with a
glibc that 5.1 doesn't have, so you may want to be sure you've got the
correct prerequisites.
CLISP takes a lot of disk space to build, but builds OK.
--
Tom Breton, http://world.std.com/~tob
Not using "gh" since 1997. http://world.std.com/~tob/ugh-free.html
Michael Vanier <·······@bbb.caltech.edu> writes:
> Hi,
>
> Since I'm a lisp fan, I thought I'd try my hand at getting a decent free
> common lisp environment running on my Linux box running Red Hat 6.0. So
> far, I've had no success.
Been there.
> From what I've read on c.l.l., it seemed like clisp and cmucl were
> the best bets.
I found clisp really easy to install. It's just not quite as cool as a
lisp that has a native compiler.
> The clisp home page (clisp.cons.org)
> does not appear to be accessible, while the cmucl sources don't seem to
> have decent documentation explaining how to build the program from sources,
> and I couldn't get the binaries to work either.
I haven't got round to trying to build it from sources. I think you
have to install binaries first anyway, for bootstrapping reasons.
To install binaries, go to
http://ww.telent.net/lisp/howto.html
(when it comes back up...)
and follow the instructions. They work.
You can get ilisp-5.9 at
ftp://ftp.cdrom.com/.1/languages/lisp/ilisp/
in the (seemingly likely) event of ilisp.cons.org being inaccessible.
HTH,
Michael
Michael Hudson <·····@cam.ac.uk> writes:
> Michael Vanier <·······@bbb.caltech.edu> writes:
>
> > The clisp home page (clisp.cons.org)
> > does not appear to be accessible, while the cmucl sources don't seem to
> > have decent documentation explaining how to build the program from sources,
> > and I couldn't get the binaries to work either.
>
> I haven't got round to trying to build it from sources. I think you
> have to install binaries first anyway, for bootstrapping reasons.
You can definitely install CLISP from just the sources. I've done it.
Takes some disk space and time. I don't remember exactly what I had
to do, but I just followed the instructions in the readme.
--
Tom Breton, http://world.std.com/~tob
Not using "gh" since 1997. http://world.std.com/~tob/ugh-free.html
On 29 Oct 1999 15:57:32 -0700, Michael Vanier <·······@bbb.caltech.edu> wrote:
>Since I'm a lisp fan, I thought I'd try my hand at getting a decent free
>common lisp environment running on my Linux box running Red Hat 6.0. So
>far, I've had no success. From what I've read on c.l.l., it seemed like
>clisp and cmucl were the best bets. The clisp home page (clisp.cons.org)
>does not appear to be accessible,
I think a disk crashed, but till the site is up again I took the
liberty of putting a fairly recent version for you (and anyone else) at
http://www.ii.uib.no/~stig/cl/clisp-1999-09-13.tar.gz
CLISP should be easy to install on a Linux-system and is a decent
CL implementation which I use on Linux and Solaris.
> while the cmucl sources don't seem to
>have decent documentation explaining how to build the program from sources,
>and I couldn't get the binaries to work either. This is quite frustrating
>since I'm used to just having to do "configure --prefix=<whatever" followed
>by "make" and "make install".
This is the way one makes and installs CLISP.
CMUCL is also available as RPMs for Redhat and is even easier to install
if you use Debian .deb packages.
> BTW I did get Allegro's free version of ACL
>going without problems, but the first program I tried it on (a music
>typesetting program) overran the (artificial) heap limitations before it
>even finished loading, which was off-putting.
I am not sure whether ACL5.0 is still downloadable, but it has a larger
heap-size limit than ACL5.01 (twice the size).
> I'd rather stick to an open-source product if possible.
Naturally, but ACL is a great system with a very good profiler :)
>--------------------------------------------------------------
>Mike Vanier ·······@bbb.caltech.edu
>Department of Computation and Neural Systems, Caltech 216-76
>GNU/Linux: We can't lose; we're on a mission from God.
I didn't know that Knuth sends people on GNU/Linux missions.. :)
--
------------------------------------------------------------------
Stig Erik Sandoe Institute of Informatics, University of Bergen
····@ii.uib.no http://www.ii.uib.no/~stig/
From: Christopher R. Barry
Subject: Re: getting a lisp running on Red Hat Linux
Date:
Message-ID: <87vh7ptgm6.fsf@2xtreme.net>
····@ii.uib.no (Stig E. Sand�) writes:
> I am not sure whether ACL5.0 is still downloadable, but it has a larger
> heap-size limit than ACL5.01 (twice the size).
There's no heap limit in 5.0 except on Windows. Did you ever actually
hit a heap limit with 5.0? Why do you believe there is one? What's
really annoying is that you usually hit the 5.0.1 heap limit when
doing a GC to try and free space, due to object tenuring:
USER(17): (gc)
Error: An explicit gc call caused tenuring and a need for 9437184 more bytes
of heap. This request cannot be satisfied because you have hit the
ALlegro CL Trial heap limit. If you wish to purchase Allegro CL please
contact Franz Inc. at ·····@franz.com or by calling (510) 548-3600.
[condition type: STORAGE-CONDITION]
Hitting the heap limit doesn't automatically kill the process though.
You can exit from the debugger and then SETF a few large objects to
NIL and retry the GC.
Christopher
On Sat, 30 Oct 1999 08:00:08 GMT, Christopher R. Barry <······@2xtreme.net> wrote:
>····@ii.uib.no (Stig E. Sand�) writes:
>
>> I am not sure whether ACL5.0 is still downloadable, but it has a larger
>> heap-size limit than ACL5.01 (twice the size).
>
>There's no heap limit in 5.0 except on Windows. Did you ever actually
>hit a heap limit with 5.0? Why do you believe there is one?
I have not had problems with it or actually hit it [1], but:
USER(1): (room)
area address(bytes) cons symbols other bytes
8 bytes each 24 bytes each
(free:used) (free:used) (free:used)
Top #x205c8000
New #x204b2000(1138688) 571:5543 188:66 922608:106960
New #x2039c000(1138688) ----- ----- -----
Old #x20000d48(3781304) 512:54514 159:14079 2099600:889976
Root pages: 37
Lisp heap limit: 67108864
And the last line suggests to me that there is a heap limit :)
[1] The 5.0.1 limit however isn't awfully hard to hit.
--
------------------------------------------------------------------
Stig Erik Sandoe Institute of Informatics, University of Bergen
····@ii.uib.no http://www.ii.uib.no/~stig/
From: Christopher R. Barry
Subject: Re: getting a lisp running on Red Hat Linux
Date:
Message-ID: <87n1t0u38p.fsf@2xtreme.net>
····@ii.uib.no (Stig E. Sand�) writes:
> On Sat, 30 Oct 1999 08:00:08 GMT, Christopher R. Barry <······@2xtreme.net> wrote:
> >····@ii.uib.no (Stig E. Sand�) writes:
> >
> >> I am not sure whether ACL5.0 is still downloadable, but it has a larger
> >> heap-size limit than ACL5.01 (twice the size).
> >
> >There's no heap limit in 5.0 except on Windows. Did you ever actually
> >hit a heap limit with 5.0? Why do you believe there is one?
>
> I have not had problems with it or actually hit it [1], but:
>
> USER(1): (room)
> area address(bytes) cons symbols other bytes
> 8 bytes each 24 bytes each
> (free:used) (free:used) (free:used)
> Top #x205c8000
> New #x204b2000(1138688) 571:5543 188:66 922608:106960
> New #x2039c000(1138688) ----- ----- -----
> Old #x20000d48(3781304) 512:54514 159:14079 2099600:889976
> Root pages: 37
> Lisp heap limit: 67108864
>
> And the last line suggests to me that there is a heap limit :)
And it will grow, Stig.
USER(5): (room)
area address(bytes) cons symbols other bytes
8 bytes each 24 bytes each
(free:used) (free:used) (free:used)
Top #x28976000
New #x26920000(33906688) 155:535839 233:21 17343272:12178456
New #x248ca000(33906688) ----- ----- -----
Gap #x248c8000(8192) ----- ----- -----
Old #x216dc000(52346880) 983:90727 0:0 50552512:1031360
Old #x2135c000(3670016) 0:0 0:0 16:3668040
Old #x2131c000(262144) 0:0 0:0 8:261840
...
OTot(Old Areas) 983:1321679 108:14130 50553352:14734288
Root pages: 63
Lisp heap limit: 536346624
^^^^^^^^^
When ACL 5.0.1 starts it uses (AFAICT) 32768k of address space. This
never grows. ACL 5.0 starts up with 47612k and will grow to hundreds
of megabytes if you let it.
Christopher
····@ii.uib.no (Stig E. Sandx) writes:
> On 29 Oct 1999 15:57:32 -0700, Michael Vanier <·······@bbb.caltech.edu> wrote:
>
> >Since I'm a lisp fan, I thought I'd try my hand at getting a decent free
> >common lisp environment running on my Linux box running Red Hat 6.0. So
> >far, I've had no success. From what I've read on c.l.l., it seemed like
> >clisp and cmucl were the best bets. The clisp home page (clisp.cons.org)
> >does not appear to be accessible,
>
> I think a disk crashed, but till the site is up again I took the
> liberty of putting a fairly recent version for you (and anyone else) at
> http://www.ii.uib.no/~stig/cl/clisp-1999-09-13.tar.gz
>
> CLISP should be easy to install on a Linux-system and is a decent
> CL implementation which I use on Linux and Solaris.
>
> > while the cmucl sources don't seem to
> >have decent documentation explaining how to build the program from sources,
> >and I couldn't get the binaries to work either. This is quite frustrating
> >since I'm used to just having to do "configure --prefix=<whatever" followed
> >by "make" and "make install".
>
> This is the way one makes and installs CLISP.
>
Thanks very much for making the clisp sources available. I downloaded them
and installed clisp with no problem at all. The only glitch was that their
configure script seemed to ignore my --prefix argument, but it was trivial
to edit the Makefile to get around that. The installation passed all tests
without difficulty. You can even use clisp as a scripting language! Kudos
to the clisp team -- well done!
>
> > BTW I did get Allegro's free version of ACL
> >going without problems, but the first program I tried it on (a music
> >typesetting program) overran the (artificial) heap limitations before it
> >even finished loading, which was off-putting.
>
> I am not sure whether ACL5.0 is still downloadable, but it has a larger
> heap-size limit than ACL5.01 (twice the size).
>
> > I'd rather stick to an open-source product if possible.
>
> Naturally, but ACL is a great system with a very good profiler :)
>
I'm sure the ACL is great, and frankly, if it weren't for the heap
limit I'd use it. I have a philosophical objection to crippleware.
> >--------------------------------------------------------------
> >Mike Vanier ·······@bbb.caltech.edu
> >Department of Computation and Neural Systems, Caltech 216-76
> >GNU/Linux: We can't lose; we're on a mission from God.
>
> I didn't know that Knuth sends people on GNU/Linux missions.. :)
>
Hmm, never thought of it that way :-) I guess Knuth works in mysterious
ways...
Cheers,
Mike
--------------------------------------------------------------
Mike Vanier ·······@bbb.caltech.edu
Department of Computation and Neural Systems, Caltech 216-76
GNU/Linux: We can't lose; we're on a mission from God.
From: Christopher R. Barry
Subject: Re: getting a lisp running on Red Hat Linux
Date:
Message-ID: <871zaasmy1.fsf@2xtreme.net>
Michael Vanier <·······@sith.bbb.caltech.edu> writes:
> I'm sure the ACL is great, and frankly, if it weren't for the heap
> limit I'd use it. I have a philosophical objection to crippleware.
Well, some of the missing features of the free Lisps are crippling....
You might find that the strengths and weaknesses of crippled
commercial Lisps complement the strengths and weaknesses of the free
ones rather nicely.
Christopher
From: Sam Steingold
Subject: Re: getting a lisp running on Red Hat Linux
Date:
Message-ID: <upuxufbxt.fsf@ksp.com>
>>>> In message <··············@sith.bbb.caltech.edu>
>>>> On the subject of "getting a lisp running on Red Hat Linux"
>>>> Sent on 29 Oct 1999 15:57:32 -0700
>>>> Honorable Michael Vanier <·······@bbb.caltech.edu> writes:
>>
>> The clisp home page (clisp.cons.org)
>> does not appear to be accessible,
the official mirror (with nightly development snapshots as well as
diffs, some binaries and stable versions) is
ftp://cellar.goems.com/pub/clisp/
--
Sam Steingold (http://www.podval.org/~sds/)
Micros**t is not the answer. Micros**t is a question, and the answer is Linux,
(http://www.linux.org) the choice of the GNU (http://www.gnu.org) generation.
The only time you have too much fuel is when you're on fire.
Michael Vanier <·······@bbb.caltech.edu> writes:
>Since I'm a lisp fan, I thought I'd try my hand at getting a decent free
>common lisp environment running on my Linux box running Red Hat 6.0. So
>far, I've had no success. From what I've read on c.l.l., it seemed like
>clisp and cmucl were the best bets. The clisp home page (clisp.cons.org)
>does not appear to be accessible,
The www/cvs machine is down with a broken disk.
You can download both clisp and cmucl stuff from the ftp mirror, which
is on another machine and still intact.
ftp2.cons.org:/pub/languages/lisp
>while the cmucl sources don't seem to
>have decent documentation explaining how to build the program from
>sources,
It's not a matter of documentation, but of complexity. There is not
one way, for each state of the sources manual steps must be inserted
and those steps require a fundamental understanding of the runtime.
And you need a working CMUCL to build one anyway.
>and I couldn't get the binaries to work either.
Please post the error messages.
Usual Linux problems (insert my usual Linux-bashing here) are FPU
control word problems on glibc-2.1 and/or the fact that there is by
default no overcommitment in the VM system anymore.
For the former, you need newer *.deb-files and install them on redhat
(the rpms are outdated), for the latter you need more swap space (min.
1-1.5 G I guess) or switch your kernel to do overcommit.
>If anyone has any tips that could make installation easier for me (and
>others in my situation), I'd appreciate hearing them.
Make sure you post your error messages next time.
Martin
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Martin Cracauer <········@bik-gmbh.de> http://www.bik-gmbh.de/~cracauer/
"Where do you want to do today?" Hard to tell running your calendar
program on a junk operating system, eh?