Having scoured the net for a free ANSI Common Lisp implementation, I
just want to check what others know.
Seems like CMU CL is the most up to date free CL, and even it is no
longer under development, and it won't run on my Linux box. <Sigh> I
might have to go in and use the Suns at work :-(
Otherwise, GCL seems to be the runner up, and I'm not entirely sure how
close it is to XJ3.
I need something that will take the place of Lucid, as I want to use a
large NLP package that has been developed under Lucid. GCL-2.2 won't
compile it, although it gets further than CLISP does.
Any messages of hope, or confirmations of my findings greatly appreciated.
Mark
--
-----------------------------------------------------------------------------
Mark Pedersen | email : ·····@cs.uq.oz.au | http://student.uq.edu.au/~s303080
"When freedom destroys order, the yearning for order will destroy freedom."
In article <················@best.best.com>, . <···@best.com> wrote:
>In article <··········@miso.cs.uq.edu.au> ·····@cs.uq.edu.au (Mark Pedersen) writes:
>
> Seems like CMU CL is the most up to date free CL, and even it is no
> longer under development, and it won't run on my Linux box. <Sigh> I
> might have to go in and use the Suns at work :-(
>
>CMU CL now runs under FreeBSD, meaning that it has a [345]86 code
>generator. A Linux port would probably not be all that difficult.
Doesn't it still lack garbage collection?
> Otherwise, GCL seems to be the runner up, and I'm not entirely sure how
> close it is to XJ3.
>
> I need something that will take the place of Lucid, as I want to use a
> large NLP package that has been developed under Lucid. GCL-2.2 won't
> compile it, although it gets further than CLISP does.
>
>In my experience, you would probably have to put in considerable
>effort even moving between two compliant implementations...
I move between Lucid and GCL-1.1 all the time, which is not to say
that Lucid-specific code will be easy to convert. (I try to write
mostly portable code in the first place.)
>Have you considered helping with making GCL more ANSI CL compliant?
>You could add/modify those functions that aren't behaving in the way
>you need them to and share the results with the GCL effort. That way,
>you get your program running and GCL moves closer to ANSI CL.
Neither Lucid nor GCL are fully ANSI. It helps to add the full MIT
loop macro, defpackage, and some other readily available things to GCL.
-- jd
In <················@iiltd.demon.co.uk> ····@interactive.co.uk (Jeff Dalton) writes:
>In article <················@best.best.com>, . <···@best.com> wrote:
>I move between Lucid and GCL-1.1 all the time, which is not to say
>that Lucid-specific code will be easy to convert. (I try to write
>mostly portable code in the first place.)
>Neither Lucid nor GCL are fully ANSI. It helps to add the full MIT
>loop macro, defpackage, and some other readily available things to GCL.
Do you (or anyone else?) know if the MIT loop macro etc. has been
incorporated into GCL-2.2. I'm not near a gcl-2.2. source tree right now
to check myself. Is there common wisdom about what additional packages
are useful to add to (particular) lisps?
The problem I ran into compiling XTAG under GCL-2.2 was that it compiled
(or appeared to compile) a package, and then when it tried to include a
struct from that package in the next package it was compiling, it said
that the struct was undefined.
Any clues about this? Esp from experienced gcl users?
Mark
--
-----------------------------------------------------------------------------
Mark Pedersen | email : ·····@cs.uq.oz.au | http://student.uq.edu.au/~s303080
"When freedom destroys order, the yearning for order will destroy freedom."
Mark Pedersen wrote:
>
> In <················@iiltd.demon.co.uk> ····@interactive.co.uk (Jeff Dalton) writes:
>
> >In article <················@best.best.com>, . <···@best.com> wrote:
>
> >Neither Lucid nor GCL are fully ANSI. It helps to add the full MIT
> >loop macro, defpackage, and some other readily available things to GCL.
>
> Do you (or anyone else?) know if the MIT loop macro etc. has been
> incorporated into GCL-2.2. I'm not near a gcl-2.2. source tree right now
> to check myself. Is there common wisdom about what additional packages
> are useful to add to (particular) lisps?
You can find my "port" of the Symbolics Loop macro and a defpackage package
at http://www.webbuild.com/~rtoy.
---------------------------------------------------------------------------
----> Raymond Toy http://www.webbuild.com/~rtoy