Dag Sverre Seljebotn wrote:
> Any thoughts? (Of course, any implementation burden would fall on me, > the implementor.) (Heh, this didn't make sense. I meant "me, the proposer".) On the issues of the release: SourceDescriptors are a pretty bug-prone change which one could argue for merging right after doing a release rather than right before. Since no actual features depend on these for now, there's no real need for merging them now either. So I'm actually happy for my branch to be merged right after the release rather than right before. Though I suppose that if you consider compiling SAGE testing good enough it shouldn't matter (runtests.py is not sufficient). Anyway, the issues Robert's pointed out should be fixed: - Clearer TreeFragment.py code, and it now copies "pos" (and also copies the substitution arguments, in case they're used in more than one place. This was the simple solution, can optimize/refcount later if needed.) - Fixed the problem pointed out with Sage compilation. I didn't test compiling Sage yet though, I'd like some help with that (see below). - Other SourceDescriptor bugs and fixes. Which brings me to: Could anyone write a sentence or two (not much, just a few pointers) as to where the Cython code in SAGE is located, and the best way to plug in a custom Cython into its build system? I have absolutely no experience with SAGE, and at least the prospect of building all the spkgs seems wrong... -- Dag Sverre _______________________________________________ Cython-dev mailing list [email protected] http://codespeak.net/mailman/listinfo/cython-dev
