Gary Klimowicz wrote:
I was planning on doing some development on win32, as that's how my system is set up right now.
I can be flexible about which Lisp you want me to port it to. I've been using ACL, but am amenable to working on whatever you feel you are most in need of. I want to help and am not too fussy about how.
Make a wish. :-)
Uh-oh, I sent it last night, 11pm NYC time. I think I'll just put it up on my web site, because most people have had trouble getting it via email. Plz confirm you did not receive and I will do that.
As for making a wish, that is very generous of you. It would be fine if you just played the part of user (vs. porter) and helped me figure out what documentation or examples to write. (I always prefer examples, and those are easier for me to pull together.) If you are really interested in porting, then Corman is the one implementation remaining to cover. CLisp (I think!) does not handle callbacks from C, which we need to use Freeglut.
Hey, who else is working on Corman? So many have asked for distros, I have lost track. :)
kenny
On Sun, 22 Feb 2004 22:10:27 -0500, Kenny Tilton ktilton@nyc.rr.com wrote:
CLisp (I think!) does not handle callbacks from C,
http://clisp.cons.org/impnotes/dffi.html
See DEF-CALL-IN.
Edi.
Edi Weitz wrote:
On Sun, 22 Feb 2004 22:10:27 -0500, Kenny Tilton ktilton@nyc.rr.com wrote:
CLisp (I think!) does not handle callbacks from C,
http://clisp.cons.org/impnotes/dffi.html
See DEF-CALL-IN.
Thx, Edi.
"A foreign function description is written as a Lisp file, and when compiled it produces a #P".c" file which is then compiled by the C http://www.eskimo.com/%7Escs/C-faq/top.html compiler and may be linked together with lisp.a."
Wow, so /any/ FFI stuff requires relinking CLisp before one can get at it? ie, Cello would come with a special CLisp FFI file to be run through a build process culminating in a new Cello-ready CLisp. That is not the end of the world. I do notice that even this much does not work on all platforms. And UFFI does not support much CLisp. That would not be the hardest workaround either.
I think the last time I looked at doing the CLisp port myself I concluded there was too much trouble for too little benefit (I do not use CLisp) and that it would be better to just make Cello good enough that a hardcore CLisper (or its maintainers) would take the trouble. I'd certainly help any way I could.
Right now I want to see if I can extend the widget set: menus, pop-up menus, context menus, scrolling, text edit, listview, treeviw, tab control... a "file open" dialog uses most of that, so maybe that is my next project....hmmm, no, I can first change the object inspector into a REPL. It certainly needs pop-up menus and scrolling already.
kt
On Mon, 23 Feb 2004 08:56:45 -0500, Kenny Tilton ktilton@nyc.rr.com wrote:
I think the last time I looked at doing the CLisp port myself I concluded there was too much trouble for too little benefit (I do not use CLisp)
Same here. I just wanted to note that callbacks in CLISP are possible, at least in theory.
Edi.
Edi Weitz wrote:
On Mon, 23 Feb 2004 08:56:45 -0500, Kenny Tilton ktilton@nyc.rr.com wrote:
I think the last time I looked at doing the CLisp port myself I concluded there was too much trouble for too little benefit (I do not use CLisp)
Same here. I just wanted to note that callbacks in CLISP are possible, at least in theory.
Yes, and thx a ton for that. I was unaware. And if at some point I get all fired up about having Cello ubiquitous (well, as far as the CLisp FFI reaches), it is nice to know I have this option. Oy, I just noticed (I think this is right): one also has to re-link CLisp with a static library of the package itself (such as FTGL) and not just the glue. ie, CLisp does not load dynamic libraries.
Well, the good news would be that this would be a one-time deal: rebuild CLisp with GL, GLU, Freeglut, ImageMagick, and FTGL and yer done.
kt
Edi.