The test in layout/reftests/flexbox/lexbox-position-fixed-2-ref.xhtml has an incorrect expectation -- the fixed-position item should not affect the layout of the flex items, and should be positioned the same way as the following flex item. Quoting http://www.w3.org/TR/css3-flexbox/#abspos-items : "An absolutely-positioned flex item does not participate in flex layout beyond the reordering step. [...]" " In the main axis, If there is a subsequent in-flow flex item on the same flex line, the static position is the outer main-start edge of that flex item. " I'll attach a screenshot of how this should look.
Thanks! I'll leave this bug on fixing the test (and probably some of the other -abspos and -fixed tests) is wrong, and I filed bug 874718 on fixing the implementation.
FWIW, that the spec has changed since this bug was filed, so the static position described in comment 0 ("should be positioned the same way as the following flex item") is no longer correct. As of the current ED, the static position of an abspos child is the position that the child item *would* have if it were the sole item in the flex container, based on the justify-content and align-self property-values.
OS: Linux → All
Hardware: x86_64 → All
Version: unspecified → Trunk
Resolving as FIXED by bug 1269045 -- specifically, this change: https://hg.mozilla.org/mozilla-central/rev/707b2ab5879d ...which updated this testcase & reference case to be correct per the latest spec text. Here's a "live" version of this updated test & reference, as of that change: https://hg.mozilla.org/mozilla-central/raw-file/707b2ab5879d/layout/reftests/flexbox/flexbox-position-fixed-2.xhtml https://hg.mozilla.org/mozilla-central/raw-file/707b2ab5879d/layout/reftests/flexbox/flexbox-position-fixed-2-ref.xhtml The testcase & reference now match, in the latest Firefox Nightly (52) as well as Chrome Dev Edition 56.
You need to log in before you can comment on or make changes to this bug.