Daniel Holbert

I've got patches that should fix this issue, over in https://bugzilla.mozilla.org/show_bug.cgi?id=1267462 - hopefully this will be fixed in Firefox Nightly 60 within a few days.

For convenience, I've posted a pull request that I think addresses this in https://github.com/necolas/normalize.css/pull/780 . (Let me know if I'm missing something in my analysis, or if you've got any...

So the behavior in that test makes sense based on what the spec used to say. I'm not sure whether it still matches the current spec (it might or might...

Quoting myself: > That flex line gets stretched to precisely fill the flex container (per align-content:stretch). This is backed by this part of the flexbox spec, in the explicit definition...

er, sorry, I guess in this testcase, the flex container is in fact multi-line, not single-line. (it has `flex-flow: column wrap;`) So my quote about `stretch` doesn't apply after all,...

Yup, agreed with the final sentence. I filed https://bugzilla.mozilla.org/show_bug.cgi?id=1784135 on Firefox's current behavior (i.e. the fact that we "pass" this test with its current incorrect expectations).

It looks like a bunch of the neighboring `align-self` tests are based on the same flawed premise -- e.g. [css/css-flexbox/abspos/flex-abspos-staticpos-align-self-002.html](https://github.com/web-platform-tests/wpt/blob/master/css/css-flexbox/abspos/flex-abspos-staticpos-align-self-002.html) They have `align-content` unset, so they're getting `normal` i.e. `stretch`...

They're asserting the old static-position behavior (align the box just as if it were the sole flex item in the flex container, and use that as the static position). Note...

> It looks like a bunch of the neighboring `align-self` tests are based on the same flawed premise -- e.g. [css/css-flexbox/abspos/flex-abspos-staticpos-align-self-002.html](https://github.com/web-platform-tests/wpt/blob/master/css/css-flexbox/abspos/flex-abspos-staticpos-align-self-002.html) Hmm, it looks like the abspos-staticpos-align-self WPT tests are...

Here's a screenshot of how my jsfiddle looks in a Firefox build with a fix for https://bugzilla.mozilla.org/show_bug.cgi?id=1784135 (aimed at keeping up with the test expectation changes in #35426):  The...