788358 - Switch XUL c++ code to check for min-width / min-height being "auto" instead of "0" as the "min-width not set" value

Status: RESOLVED WONTFIX

(Core :: Layout, defect)

Description

[Daniel Holbert [:dholbert]]

The main patches on bug 763689 just landed (switching the default computed value of min-height & min-width over to "auto" from "0px") -- filing this bug to follow up on this comment with some XUL issues that NeilAway brought up, which I mentioned in bug 763689 comment #12:
> NeilAway says that XUL assigns a special significance to the (old default)
> min-width / min-height value of 0.  It treats 0 as something min-content-ish
> instead of actually treating it as 0.  This lets XUL solve the same sort of
> problem that css3-flexbox is trying to solve here.  And then XUL content can
> specify "min-width: 0%" (with a %) if it _really_ wants a min-width of 0.
> 
> So, in light of that:
>  (a) We need to make sure that the new default (auto) still triggers the
> min-content-ish behavior in XUL, like the old default (0) did.
>  (b) We might want to fix XUL so that "min-width: 0" actually means 0 now
> instead of "min-content-ish"

Also: (c) We can probably remove any special cases for "min-width:0" in XUL C++ code.  (or, convert them to be a special-case for "auto" instead, depending on the situation)

((c) is sort of an extension of (a))

Comment 1

[Daniel Holbert [:dholbert]]

The CSSWG will likely revert "min-width:auto" / "min-height:auto", so this is will likely become RESOLVED|INVALID once that goes through & we remove support for these values.

For reference:
  http://lists.w3.org/Archives/Public/www-style/2013Mar/0137.html

Comment 2

[Daniel Holbert [:dholbert]]

per comment 1, this is sort of INVALID currently, though I believe the CSSWG was considering (at some point) reintroducing min-width:auto. So I'll leave this open for now.

Comment 3

[Daniel Holbert [:dholbert]]

(Yeah, min-width:auto is indeed coming back (which is happening in bug 984711).) So, disregard comment 1.)

Comment 4

[Daniel Holbert [:dholbert]]

(In reply to Daniel Holbert [:dholbert] from comment #2)
> per comment 1, this is sort of INVALID currently, though I believe the CSSWG
> was considering (at some point) reintroducing min-width:auto. So I'll leave
> this open for now.

They did indeed reintroduce it (and we added them back in bug 984711).

I think this bug is WONTFIX, though.  Specific notes on the action items from comment 0:

(In reply to Daniel Holbert [:dholbert] from comment #0)
> > So, in light of that:
> >  (a) We need to make sure that the new default (auto) still triggers the
> > min-content-ish behavior in XUL, like the old default (0) did.

Apparently this turned out to be fine without any changes.  Probably because we resolve "auto" to "0" in all the right places.

> >  (b) We might want to fix XUL so that "min-width: 0" actually means 0 now
> > instead of "min-content-ish"

Nope, we're not really bothering with improvements to XUL now.

> Also: (c) We can probably remove any special cases for "min-width:0" in XUL
> C++ code.  (or, convert them to be a special-case for "auto" instead,
> depending on the situation)

Nope, we're not really bothering with improvements to XUL now.