CSS 2.1 specifies that an absolutely positioned object may have a relatively positioned inline element as its containing block, but in such cases the equation for determining the bounds change from the ordanary. Currently no browsers that I have tested fully implement the specified behavior.
Created attachment 8648 [details] patch
Typo in the comments in two places: "relativly" should be "relatively" Should we be worried about the fact that this doesn't match other browsers?
I think we should.
I would be happy to make a quirks mode variation, but I am really unsure what we want to emulate from the other browsers. Our current behavior does not match either Firefox or Opera (the only two I have access to), and I am not sure what we would want to mimic. I will of course fix the typos, but I would also love to know what a desired quirks mode behavior would be.
Created attachment 8683 [details] patch 2 Fixes the misspellings.
Comment on attachment 8683 [details] patch 2 r=me
(In reply to comment #6) > (From update of attachment 8683 [details] [edit]) > r=me There were no *-expected.txt results in this patch. In the future, it would be helpful to have these test results included in the patch to make sure a local build is working properly before commit. Thanks!
Committed revision 14727.
*** Bug 5722 has been marked as a duplicate of this bug. ***