[css-display] Should 'run-in flow-root' blockify to 'block' or 'flow-root'?

[Loirooriol]

I was drawing a diagram to understand the CSS Display transformations after the F2F resolutions and how a hypothetical ruby-root could fit in there.

I noticed that it's not clear how run-in flow-root should blockify, and you didn't discuss this in the F2F.

It was resolved that inline flow-root blockifies to block flow because inline flow-root is syntactically equivalent to inline-block, and inline-block must blockify to block because of backwards compatibility.

A run-in flow-root is basically an inline-block with some special tree munging. So it could make sense to let run-in flow-root be consistent with inline-block and also blockify to block flow.

The other possibility would be to keep it simple and only change the outer display type, i.e. blockify to block flow-root.

In practice, the difference should not matter much because I expect blockifications to trigger becoming a formatting context. But it will affect getComputedStyle, of course.

I don't have a strong opinion, but maybe I prefer consistency with inline flow-root.

[tabatkins]

Agenda+ to verify making "run-in flow-root" blockify to "block flow", to match the behavior of "inline flow-root" (aka "inline-block").

[frivoal]

+1 to the above change.

Incidentally, the full table make sense to me, and I'd like to use it as a basis to close #1457. Following up in that issue.

[astearns]

reopening for the agenda+

[css-meeting-bot]

The Working Group just discussed Blockifying 'run-in flow-root' to 'block' (for consistency with inline-block), and agreed to the following resolutions:

• RESOLVED: Run-in flow-root blockifies to block and not flow root

The full IRC log of that discussion

<dael_> Topic: Blockifying 'run-in flow-root' to 'block' (for consistency with inline-block)
<dael_> github: https://github.com//issues/1715
<dael_> florian: Was an alternative considered? I can't think of one.
<dael_> TabAtkins: It's the question in the title. Should it blockify to flow-root. It shouldn't.
<dael_> TabAtkins: An inline flow root blockifies to an ordinary block. This is for legacy reasonings. A run-in flow-root needs to blockify somehow. The general rule for run-in they're identical to inlines. Thus it shoudl blockify the same as inlines.
<dael_> TabAtkins: Alternative is if blockifies to flow-root which makes sense in abstract, but we'd loose the connection to inline.
<dael_> florian: Consistancy argument wins or theoretical purity
<dael_> astearns: Run in flow root blockifies to block and not flow root
<dael_> astearns: Obj?
<dael_> RESOLVED: Run-in flow-root blockifies to block and not flow root

[Loirooriol]

Just for reference, this is the updated table: