[cairo] Pending Patches

Carl Worth cworth at east.isi.edu
Sat Aug 14 07:20:56 PDT 2004


On Sat, 14 Aug 2004 00:36:07 -0600, "Peter Dennis Bartok" wrote:
> Now that we've (hopefully) left the license discussion behind, how about
> getting some of those pending patches reviewed and committed,

Yes, I would very much like to get back to coding myself.

> and maybe even getting a new snapshot of cairo released?

The current tessellation bugs in CVS leave things too broken even for a
"pre-release snapshot" in my point-of-view. The long-term approach is to
finish the new tessellation routine, (which I have been working on). A
shorter-term way to get a snapshot out would be to revert Keith's exact
line intersection code from 2004-05-28. Oh, that looks like it might be
as easy as backing up a single file since cairo_traps.c hasn't changed
again since then, so I might do that.

> I've submitted a patch in May,
> resubmitted in June, and it never made it into CVS, but it wasn't rejected
> either.

Yes, I'm sorry about that.

With that specific patch, I believed it could fix the symptom, but I
also believed it indicated deeper bugs regarding the memory management
semantics of the libpixman interface. So I had delayed the fix hoping to
look at the deeper problem.

I've now committed the fix and just added a comment indicating the
deeper problem.

> Not sure what else I can do at this point short of forking the last
> pre-LGPL cairo version and just changing what is needed, but it's starting
> to become frustrating.

I am sorry about your frustration. But if you're serious about wanting
to host a bug-fix fork or something that would be a fine idea. I don't
think that cairo is moving fast enough, nor does it have a user base
large enough to really need something like that. But when we get there I
would be happy to see people hosting alternate trees like this.

> Also, it'd be a great thing to get an API that
> provide means to measure glyph sizes. Or expose the existing one...

Is there something insufficient about the functionality of
cairo_text_extents and cairo_glyph_extents?

-Carl




More information about the cairo mailing list