Re: [css-flexbox] percentage heights of flex item descendants

Hi Tab,

On Tue, Aug 19, 2014 at 8:38 PM, Tab Atkins Jr. <jackalmage@gmail.com>
wrote:

> On Tue, Aug 19, 2014 at 4:46 PM, Christian Biesinger
> <cbiesinger@google.com> wrote:
> > Hello,
> >
> > so I saw that the flex box spec requires that a descendant of a flex item
> > with percentage height gets resolved against the flex item's actual
> height:
> > http://dev.w3.org/csswg/css-flexbox/#algo-stretch
> >
> > I was wondering why that is? This is inconsistent with every other use of
> > percentages, and requires multi-pass layout (ie. is slow).
>
> I'd have to dig for the archeology of the change, but it was meant to
> directly address author feedback on it being confusing.  Note that
> there's no need to do multi-pass layout if the flexbox's cross size is
> definite and it's single-line, as you know the size of the line
> automatically then.
>
> I find it surprising that this behavior is less confusing. Why do authors
want this behavior in this one place in the platform? Why only for stretch?

Anyway, point taken.

thanks,
-christian