Boris Zbarsky wrote:
Robert O'Callahan wrote:
Sounds like a step in the right direction. However, what about a stack
of nested inlines that contain a block at the bottom? It seems like
your approach can't detect that.
Hmm. That's a good point. That issue doesn't come up for the other
three cases, I guess.
I suppose we could more or less keep the {ib} wrapping separate as it is
for now...
Another option is to have the analysis pass actually construct the
subtree of frame construction data for inline frames (recursing into
kids that are themselves inlines, then propagating information about
block kids up that tree) and then pass that information to
ConstructInline so that it doesn't have to recompute all that information.
Need to think about that some more.
-Boris
_______________________________________________
dev-tech-layout mailing list
[email protected]
https://lists.mozilla.org/listinfo/dev-tech-layout
