[cairo] Updated ROADMAP for cairo 1.2.0 (and beyond)

[Mike Shaver]

On 4/26/06, Carl Worth <cworth at cworth.org> wrote:
> And, the initial section (Target) should really be driven by people
> outside of cairo, so maybe that section really should be moved to a
> wiki page rather than just a file under revision control in the source
> tree.

As far as the Mozilla targets go, we're going to need improved
performance well before December of 2006.  We could probably take
performance fixes up until around then, but we'll need to be relative
confident in our final performance state by the fall.  I'll let Vlad
and others comment more on our detailed needs there, but history
dictates that we depend on performance in the hand, even if there
might be twice the performance coming soon from a bush. :)

We're also hoping to get some tesselator improvements before the end
of Q2, since the tesselator accounts for >50% of some
moderately-complex SVG testcases reduced from the wild.  I'm not sure
how far along the design of an improved tesselator is -- there was a
promising prototype from keithp at one point, I seem to recall, so I
have hope! -- but it would be good to know what the timeline is for
that work.

What's the best way for us to provide more detail around our
performance needs?  If the Target section moves into a Wiki, someone
could update it there, but otherwise should Vlad just check in
elaborations?

Mike