This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
Section: http://dev.w3.org/csswg/css3-flexbox/#flex-items # Flexbox layout algorithm operates on flexbox items, which are boxes that # satisfy at least one of the following criteria: I can't think of boxes that can simultaneously satisfy two criteria. Can we remove "at least" here? It might be confusing even if it is mathematically sound. # <!-- flexbox item: block-level replaced element --> # <iframe id="item7"> <iframe> is by default 'display: inline'. # <!-- flexbox item: inline-level replaced element --> # <img id="item7" style="display:inline"> I don't think 'style="display:inline"' is necessary here. It's not clear to me if that declaration would actually make things clearer or confuse people... # <!-- flexbox item: inline-level replaced element --> # <button id="item8">button</button> I know <button> is atomic, although whether it is a replaced element seems quite debatable...
# For example, a lone ‘table-cell’ box is fixed up by generating ‘table’, # ‘table-row-group’, and ‘table-row’ boxes around it. CSS 2.1 currently doesn't generate 'table-row-group' (because it is not detectable maybe?).
# <!-- not a flexbox item, because it's out-of-flow --> # <div id="not-an-item1.5" style="position: absolute;">block</div> # # <!-- flexbox item: block-level child --> # <div id="item2" style="display:table">table</div> Given that the placeholder for the absolutely positioned element generates an anonymous flexbox item, the next number should be item3, I think. # <!-- flexbox item: anonymous block around inline content --> # anonymous item 4 It's not clear to me if you really want to classify an anonymous flexbox item as an anonymous block, given that you didn't say so in # 3. Contiguous run of non-replaced inline children, wrapped into an anonymous # box and given that an atomic inline-level flexbox item is not a block. (That is, the annonymous flexblox item could be an anonymous inline-block too (at least conceptually since it doesn't matter).)
Fixed all your issues. I fixed the last by specifying "block box" - boxes need to have their type specified, even if it's undetectable currently.