Re: guiledraw (was: problem with compiled module)

[Lars Arvestad <arve@nada.kth.se>]

LM> The thing is giving a lot of other weird errors [like hanging
LM> forever in an SCM_ASSERT], and I'm quite sure it's me doing
LM> something stupid, so I decided to upload the whole stuff to
LM> http://feast.netpedia.net/guiledraw/ for "peer review" :-)

I have also started on libplot bindings, but my work is so incomplete
and buggy that I didn't want to make an announcement just yet. I have
now made my code available as
http://www.nada.kth.se/~arve/drawing.tar.gz for those
interested. Perhaps we can make a joint effort?  Unfortunately, the
main reason I have not completed the coding (I haven't even started
thinking about putting it in a module) is that I have more urgent
things on my mind (such as finishing my degree...) so I will not be
able to code other than on unusual occasions.

I looked at your code, though, and one thing caught my eye: 
    SCM_NEWCELL (plotter);
    SCM_SETCAR (plotter, smob_plotter);
    SCM_SETCAR (plotter, handle);

I believe it is supposed to be SCM_SETCDR in the last line!

Also, is it necessary to put the plotter handle in an smob? The handle 
is only an integer, which can safely be passed to SCM, and since the
actual plotter structures are kept by libplot in a global array, they
should not need further protection, right? I am rusty on the GC issue, 
so please correct me if I am wrong!

	Best regards,
	Lars

-- 
Lars Arvestad               Dept. of Numerical Analysis and Computing Science
                       Royal Institute of Technology (KTH), Stockholm, Sweden