[cl-opengl-devel] Project status and request for work
Hello, my name is (as you probably can see!) Odd Rune, and this is my hello-mail to the list. I am a quite fresh Common-Lisp coder, having just finished Seibels "Practical Common Lisp" and also finished working on a few side-projects on my own (ray-tracers and various attempts at getting fast fourier to run, uh, fast). I am, however quite proficient with OpenGL using C and I have great interest in it as a standard. As with all open source projects, there is sort of a wall to climb before one can actually contribute with anything useful -- I would therefore be very grateful if anyone could point me in the direction of "useful stuff for a project-newbie". I find in particular bug-hunting to be an excellent way to get to know the code-base, but anything will do. Another thing I was wondering about was the current status of the project, is there any major revisions coming up? oR
On Sat, May 24, 2008 at 10:42 PM, Odd Rune Strømmen Lykkebø <oddrunesl@gmail.com> wrote:
my name is (as you probably can see!) Odd Rune, and this is my hello-mail to the list.
Welcome!
As with all open source projects, there is sort of a wall to climb before one can actually contribute with anything useful -- I would therefore be very grateful if anyone could point me in the direction of "useful stuff for a project-newbie".
I'm not sure how newbie-friendly these tasks are, but here's what springs to my mind: * A parenscript-like language for GLSL. See <http://advogato.org/person/ingvar/diary/166.html> for an initial attempt at doing that. * Adapt the OpenGL documentation for cl-opengl. (Something like what the Python bindings have, I suppose.) * Translate the rest of the redbook examples to cl-opengl/cl-glut. (Not exactly a big priority, but it'd be nice anyhow.) * Finish the bindings. While everything is available from the %gl package (since that is autogenerated from the spec) not everything has been lispified yet. * Write an automated test suite. * Not exactly specific to cl-opengl, but something better (in particular, more Lisp-friendly) than GLUT would be useful. I learned about LGLFW recently (<http://bollhavet.ath.cx:1080/~mikezor/lglfw/>) but I have not tried it. Porting cl-glut's CLOS-based API to that might be interesting. Hope that helps. Have fun! -- Luís Oliveira http://student.dei.uc.pt/~lmoliv/
On 5/24/08, Luís Oliveira <luismbo@gmail.com> wrote:
* A parenscript-like language for GLSL. See <http://advogato.org/person/ingvar/diary/166.html> for an initial attempt at doing that.
Might be nicer to have something smarter than parenscript, so it could do type inference, and maybe some simple optimizations.
* Adapt the OpenGL documentation for cl-opengl. (Something like what the Python bindings have, I suppose.)
Something that could hook into slimes documentation hotkeys (C-c C-d h, etc.) would be particularly nice :)
* Translate the rest of the redbook examples to cl-opengl/cl-glut. (Not exactly a big priority, but it'd be nice anyhow.)
* Finish the bindings. While everything is available from the %gl package (since that is autogenerated from the spec) not everything has been lispified yet.
I'd probably lean towards writing some examples of more modern OpenGL usage, particularly for someone who uses GL a lot already. Would also be a good way to prioritize which of the missing bindings would be useful to wrap further.
* Not exactly specific to cl-opengl, but something better (in particular, more Lisp-friendly) than GLUT would be useful. I learned about LGLFW recently (<http://bollhavet.ath.cx:1080/~mikezor/lglfw/>) but I have not tried it. Porting cl-glut's CLOS-based API to that might be interesting. see also http://www.cliki.net/cl-glfw , which claims to work with cl-opengl.
Examples of using cl-opengl with other alternatives to glut would also be good, lispbuilder-sdl, CLX, direct glx/wgl/agl/etc... Performance testing/tuning would also be useful, see how close to C we can get, and figure out what sorts of things need work (like texture uploading doesn't look like it will be good until CFFI gets an API for sending lisp arrays to FFI without a copy). -b-
On Sat, 24 May 2008 23:40:28 -0500 "Bart Botta" <00003b@gmail.com> wrote:
Examples of using cl-opengl with other alternatives to glut would also be good, lispbuilder-sdl, CLX, direct glx/wgl/agl/etc...
There's an old post here <http://article.gmane.org/gmane.lisp.cl-lispbuilder.general/648> showing how to use cl-opengl with lispbuilder-sdl; my own tests have been successful as well. As far as CLX is concerned, cl-opengl doesn't work with that, does it? I seem to recall some discussion about this (on IRC maybe) but I don't really remember so I could be wrong. Any further information on CLX + cl-opengl would be appreciated by me as well. Best regards, Mikael Lax
On Fri, May 30, 2008 at 5:55 PM, Mikael Lax <mikael.lax@bredband.net> wrote:
As far as CLX is concerned, cl-opengl doesn't work with that, does it? I seem to recall some discussion about this (on IRC maybe) but I don't really remember so I could be wrong. Any further information on CLX + cl-opengl would be appreciated by me as well.
I think Tim Moore figured out the details of getting CLX and OpenGL (via FFI) working together. See <http://common-lisp.net/project/glouton/>. That reminds me; adding an FFI-less CLX backend to cl-opengl would be an interesting project as well, perhaps. -- Luís Oliveira http://student.dei.uc.pt/~lmoliv/
On 5/30/08, Mikael Lax <mikael.lax@bredband.net> wrote:
On Sat, 24 May 2008 23:40:28 -0500 "Bart Botta" <00003b@gmail.com> wrote:
Examples of using cl-opengl with other alternatives to glut would also be good, lispbuilder-sdl, CLX, direct glx/wgl/agl/etc... There's an old post here <http://article.gmane.org/gmane.lisp.cl-lispbuilder.general/648> showing how to use cl-opengl with lispbuilder-sdl; my own tests have been successful as well.
Yeah, I use it also, but it would be good to have an example with the official code. In particular the bit about making extensions work is probably not obvious (and missing from that post I think. For the record, the magic bit (as of the last time I tried it) is to add (setf %gl:*gl-get-proc-address* 'sdl-cffi::sdl-gl-get-proc-address) after the GL context is valid, and before you call any extensions).
As far as CLX is concerned, cl-opengl doesn't work with that, does it? I seem to recall some discussion about this (on IRC maybe) but I don't really remember so I could be wrong. Any further information on CLX + cl-opengl would be appreciated by me as well.
Not sure about that, I've heard that it should work, but I haven't done any X coding in long enough that I don't know whether I would expect it to work or not... Either way, it would be nice to determine if it works or not, and provide examples if it does, or document it if not -b-
participants (4)
-
Bart Botta -
Luís Oliveira -
Mikael Lax -
Odd Rune Strømmen Lykkebø