[cairo] After 1.3.2, what's going to make cairo faster next?

[Carl Worth]

On Thu, 16 Nov 2006 14:31:43 -0800, "Daniel Amelang" wrote:
> I plan on going back and working on this next week sometime. The
> performance improvements I'm currently working on are more important,
> I think.

Excellent plan on both points.

> I back :) I've been working on Aivars patch all afternoon, and I
> expect to submit my reworking of it sometime tomorrow. This one is a
> pretty big win, BTW. I'll address the text-rendering floating point
> optimizations in a separate email, as it's complicated.

Welcome back! I'll continue to look forward to lots of good stuff from
you.

> While we're talking about performance, what are your thoughts on the
> stencil buffer approach that zrusin is so enamored of? It appears to
> only work with OpenGL, but providing a path that (completely?) avoids
> the tessellator when possible might be interesting.

I really can't form an opinion on it. I saw a blog post which had only
about as much detail as you give above "OpenGL + stencil buffer" with
claims of linear-time performance. I didn't see any pointers to an
explanation of how the approach actually works, what tradeoffs it
might have, any implementation, nor even any pictures of results from
which the quality could be judged.

If the phrase "stencil buffer" makes it obvious to someone how to
write a fast, high-quality replacement for cairo's tessellator, then
please knock yourself out.

In the meantime, I think there will always be reasons that we will
want the all-software paths in cairo to be quite fast. So independent
of any approach requiring OpenGL, I think it's important to work to
optimize the software-rendering paths, so that's something I will
continue to focus on personally.

-Carl
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://lists.freedesktop.org/archives/cairo/attachments/20061117/3c0251c4/attachment.pgp