We do padding-right/top/bottom properly, but not padding-left.

The test looked fine to me *except* for the percentage test case. Peter: I think this is a bug in the CalcPaddingFor code, yes? If so, let me know and I'll reassign the bug to you.

Since you didn't answer my email, I'm going to assume it's your bug :-) Assign it back to me if I'm the culprit... thanks

There was a bug in the style context that caused the computation of percentage padding to fail in some cases. That's fixed now (as of my next checkin). The initial reflow however, generally fails to compute the correct size because the parent frame's width isn't set in time for the child to compute the percentage.

A note from peter: In cleaning up handling of percentage measurements, it became apparent that the code to calculate margins, borders & padding should get moved from style land into frame land. Step one was to add new methods to query for cached values that don't compute them if not found (return PR_FALSE instead). Next, Kipp'll remove references to the old calls (CalcMarginFor and the like) and add static helper functions to do the job when style can't provide enough info. Then the Calc methods will get removed. This will fix also bugs like #853. While I was there, I fixed the initial value of border widths (Bug 1051). Peter

% padding values now work

verified fixed with Feb 4 Seamonkey