- From: L. David Baron via GitHub <sysbot+gh@w3.org>
- Date: Thu, 23 Feb 2017 21:22:51 +0000
- To: public-css-archive@w3.org
In response to @FremyCompany's https://github.com/w3c/csswg-drafts/issues/1046#issuecomment-281551260 : I'm not sure how that spec statement is relevant; the parent div has 'width: max-content', so the containing block width should be the max-content width of the table, not the width of the browser window. But I think the spec I wrote a while back [explains how percentages should be handled](https://dbaron.org/css/intrinsic/#autotableintrinsic) (though also see the other parts of its [description of automatic table layout](https://dbaron.org/css/intrinsic/#autotable)). Part of the reason that the test didn't do what I expect, though, is that Gecko has only added unprefixed support for `max-content` in some contexts; for `width`, it needs to be `width: -moz-max-content`. The other part of the reason that Gecko didn't do what I expected is that tables actually have two different concepts of max-content width, specifically to avoid propagating the infinite max-content widths outside of the table (while still using them for the table's own sizing). I introduced this as part of the [original reflow refactoring landing](https://bugzilla.mozilla.org/show_bug.cgi?id=300030), so it might be a little hard to track down the history (which is in CVS, but was never imported to either git or mercurial). So it's actually possible that infinite max-content sizes aren't yet a general concept but are still local to only some areas. -- GitHub Notification of comment by dbaron Please view or discuss this issue at https://github.com/w3c/csswg-drafts/issues/1046#issuecomment-282125376 using your GitHub account
Received on Thursday, 23 February 2017 21:22:59 UTC