Created attachment 66422 [details] Test case When an element has column-width and it needs to have its width set to the intrinsic width (using absolute pos, floats or width: intrinsic) the width is not correct. In the test case, all the columns should have white background.
While the exact behavior isn't described in any final spec yet, there's common sense and also the following ED to guide us: http://dev.w3.org/csswg/css-sizing/#multicol-intrinsic I think the particular test case in this bug report will be tricky to fix, since it's about finding "the max-content measure of a multi-column element with restrained-height columns". The spec suggests that we need to perform a layout pass to get that measure, and currently WebKit doesn't do that when calculating intrinsic widths. I have some changes in this area (but they currently only cover the cases where column-count is non-auto) that I'd like to submit, but since I probably won't manage to make the test case in this bug pass, I suppose I'd better open a new bug.
Bug 116677 reported, with patch.
No two browsers seem to agree on behavior here. Safari and Chrome are close (though Chrome clips more than Safari. Firefox is the most different.
<rdar://problem/96907406>
Is this will be fixed by bug 116677? As mentioned in below Chromium bug? https://bugs.chromium.org/p/chromium/issues/detail?id=54626