[css-flexbox-1] Fragmentation rules around break propagation are confusingly-written

[fantasai]

web-platform-tests/wpt#7958

[fantasai]

OK, I made some clarifications to Fragmentation and to Flexbox, specifically, pulled out break propagation into its own section into css-break-3 and clarified that:

• out-of-flows don't propagate their break values up to their parents
• layout modes like Flexbox and Grid have more specific propagation rules which override the one in css-break-3
• Flexbox uses order-modified document order in determining first/last child

Agenda+ since this has normative implications

[gregwhitworth]

@fantasai With regards to flex change specifically - this is how I understood the previous spec text to work so no argument from me. I think the clarity provide by this change will make it much more approachable by anyone; so, it's a good improvement. Finally, the addition of the note regarding the computed value clarifies the confusion the test author had. Thanks for doing this.

[css-meeting-bot]

The Working Group just discussed Fragmentation rules around break propagation are confusingly-written, and agreed to the following:

• RESOLVED: Accept changes in https://github.com/w3c/csswg-drafts/issues/2614#issuecomment-385836043

The full IRC log of that discussion

<dael> Topic: Fragmentation rules around break propagation are confusingly-written
<dael> github: https://github.com//issues/2614#issuecomment-385836043
<rego> no she sent regrets
<dael> Rossen_: Summary: There were some comments made about fragmentation rules not being clear, esp with child/parent break propagation. Changes added to css break were intended to address this.
<dael> Rossen_: [reads 2st clarification] makes sense because out of flow items proagate to containing block.
<dael> Rossen_: 2nd is flex and grid layout have specific prop rules that override css 3 break. That was in tune to previous since we've extended fragmentation before and assume that overrides, but now that's explicit
<dael> Rossen_: Last is flexbox....i think grid...order is modified and prop is based on layout order, not doc order.
<dael> Rossen_: These are the changes. Issue was on the agenda to either accept or challenge the changes.
<dael> Rossen_: From the 3 changes, as I said, 2 are mostly editorial. 1st is out of flows don't prop their break...that can be seen as normative
<dael> Rossen_: Opinions? Questions? Reasons not to accept?
<dael> Rossen_: I'll take silence as no. Obj to accept changes in https://github.com//issues/2614#issuecomment-385836043 ?
<dbaron> I think it's probably fine, but I think out-of-flows might be a *little* interesting, maybe?
<dael> RESOLVED: Accept changes in https://github.com//issues/2614#issuecomment-385836043
<dael> dbaron: I think it's fine