[cairo] Trapezoid bounding box computations

[Keith Packard]

When filling polygons, we have to compute the bounding box of a set of 
trapezoids.  Currently, that's done after the trapezoids are generated 
through a relatively expensive mechanism.

I've got code sitting here which does this incrementally while the 
trapezoids are being generated.  For the current trapezoid model, it also 
computes a 'lazy' bounding box, using the coordinates of each segment 
endpoint instead of computing approximate trapezoid vertices.  This is 
"correct" for Cairo which uses only polygon vertices for the trapezoid 
segment endpoints, and hence for which the trapezoid segment endpoints are 
sufficient globally, if not locally for each trapezoid.

Should I just throw this away in anticipation of the new tesselation code? 
Or would the incremental bounding box computations make sense in any case?

-keith


-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 228 bytes
Desc: not available
Url : http://lists.freedesktop.org/archives/cairo/attachments/20040907/59f04524/attachment.pgp