[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[no subject]

Note: it send this mail to the ML and not to Joerg since some ideas
put herein could be criticized by all of the ML members.
Furthermore, I've lost Joerg EMail :(.

Let's go ....

Joerg Hoehle wrote:
-------------------

>I'd recommend using the cmulisp+comint.el package from Olin Shivers
Ok, thanks. Note that I currently use the GNUEMACS 18.xx (not the beta
version on Aminet). I'll see if cmulisp.el is part of the package I have.

>Q1:what do you mean by clisp.el?
I meant some kind of package dedicated to clisp. You answered me: cmulisp.el

>>(I wrote)
>> I know EMACS uses inferior-lisp-program for the path of the program he uses
>>  when one types in M-x run-lisp. Is there some kind of inferior-lisp-arguments to
>>  specify the arguments of the command ?
>(You answered)
>No, but cmulisp.el+comint.el solves this problem, using another variable.
Hope, I've got it yet.
It seem that I'll do some kind of (autoload ...) in my .emacs tonight :)
 
>> When will the next version of CLISP will be released on Amiga ?
>When I'll have a somewhat usable FFI. That's what's missing.
(Silly question) What is a FFI ?

>Well, to be more precise: no public Amiga binary has been _released_
>since then, although I have distributed several newer versions to a few
>people with specific needs. I want to release a new version soon, but
>the current developers know that it's still missing some things that
>I'd just like to "be in there".
I use CLISP for research work, last year I made my project  of Master of
Science with it (a Lambda-Calculus interpreter) and since I was VERY
pleased with it, I made some advertising at the University for it.
I'll use it this year to do research (I am in DEA (between Master of
Science and PhD, I guess)). There is and abstract:
Applying Rewriting to Object Specifications
Use rewriting to give operational semantics to a Formal Object Model
(formal classes) developped at the University.
Language: CLOS.
Team: Equipe de Recherche en Technologie Objet
      Object Technologies Research Team
      IRIN - Universite de Nantes, France
I don't have special need, apart from beeing able to use a Common Lisp
respecting the norms (ANSI X3J13 I think ...) and making my programs
running both at the University (DEC Alpha) and at Home (A12oo).

>Questions:
>1. how many users are out there that would like or need the set of .o
>files to link together with own objects?
I'm interested in everything that makes Amiga-CLISP closer than
-say- the UNIX version.
As far as the .o are concerned, I'm also interested in interoperability
problems (i.e. using object classes between different languages, see
SOM, and the Formal Class Model). Could be useful.

>2. who has been able so far to build an own CLISP?
I guess I could try but
- don't have time right now
- got GCC2.6.2 (+2.6.3 but: to install)

>3. Where to put special archives containing .o files? Remember that
>there are now three different binaries (-low, -high and -wide) and all
>would need own object files.
I've go -high. What is -wide ?
I'd put .o files with the sources archive (.lsp/.fas)
I'd add emacs' stuff and some kind of LispTOLateX driver with it.
I'd put also pointers to the mail-list, news-lists, books...
It could be some kind of package for intensive users. Others could cope
with the current way of doing I think.

>4. would people like myself not distributing an archive with a binary,
>but with the .o files and GNU-ld instead? That would take more place
>and would need one more step to install. (Distributing GNU-ld in an
>archive would also require having GNU-ld source available near it,
>needing yet more ftp resources.)
Ok !
But, as far as GNU-ld is concerned, I think more than 3/4 of the CLISP
users that will use this package should have GCC at home.
Note that GNU-ld comes from GCC-2.6.3.
My solution:
 - put GNU-ld with it for say, one or two versions
 - after that, when everyone should have GCC-2.6.3 (or higher) GNU-ld
   should no longer be needed.

Well, bye.
I'm going back home to see if I have the emacs files you told me about.

Pascal

           |||
          oO oo
     --oOO-\_/-OOo-- A1200/68030MMU-28Mhz

poizat@ensinfo.univ-nantes.fr