- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Wed, 10 Jul 2013 15:14:46 -0700
- To: Alan Stearns <stearns@adobe.com>
- Cc: "www-style@w3.org" <www-style@w3.org>
On Wed, Jul 10, 2013 at 2:42 PM, Alan Stearns <stearns@adobe.com> wrote: > Hey all, > > The Basic Shapes section had this sentence: > > --- > If the relevant box of the element > is dependent on auto sizing > (i.e., the element's 'width' or > 'height' property is 'auto'), > then the percentage values resolve to 0. > --- > > > This is not the case for how clip-mask uses basic shapes, so it needed to > at least move out of that section. > > This sentence was a punt added because of the mismatch between auto sizing > on content and a shape-inside affecting that content's composition. It's > not needed or useful for shape-outside because shape-outside doesn't > affect auto sizing of its element. Since we only have shape-outside in the > current draft, I've removed this sentence. Now percentages in shape > functions will use the 'relevant box' from the element with shape-outside. > > When we get to shape-inside, there will still be that mismatch to deal > with. But a possibly-wrong approximate percentage value will in some cases > be more useful than 0, which isn't ever a useful interpretation of a > percentage value. So I think this change improves the present situation > immensely and does not make the future situation any worse. I approve of this change for shape-outside. I disagree that a sometimes-wrong value is better than an always-useless value. The former is just asking for authors to accidentally depend on correct behavior, while the latter can't ever be depended on. ~TJ
Received on Wednesday, 10 July 2013 22:15:33 UTC