[cairo] Thinking towards 1.14

[Chris Wilson]

Having achieved the near-impossible and managed a release out after an
extended delay, I don't want a repeat experience. At this moment in
time, the only development plans I have are towards getting cairo-gl to
an acceptable level and turning it on by default. I certainly have no
wish to rewrite the compositor again, though there is room for
improvement in the handling of curves, hairlines and overhead reduction.
Nothing that can't be done in an incremental fashion. On the X side, we
need to extend the Render protocol before we can realise any significant
improvement.

So, I do not see a need to split development off into a separate branch
and let the stable releases languish. Instead, we should merge features
into the stable branch as they are ready and keep on bugfixing with
regular releases of 1.12, i.e. we need to keep master in a
release-worthy state.

If you have been holding back patches waiting for master to be unfrozen,
consider it so - but lets not break the world.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre