Re: [css-ui-3][css-backgrounds-3][css-box][css-shape][css-round-display] Is the border box rounded by border-radius?

On 06/16/2015 02:25 PM, Florian Rivoal wrote:
>
>> On 16 Jun 2015, at 17:39, Alan Stearns <stearns@adobe.com> wrote:
>>
>> On 6/16/15, 8:30 AM, "Florian Rivoal" <florian@rivoal.net> wrote:
>>
>>> I was writing tests for css-ui-3, and ran into an ambiguity.
>>>
>>> About the cursor property, css-ui says this:
>>>
>>> "This property specifies the type of cursor to be
>>>   displayed for the pointing device when within the
>>>   element’s border-box."
>>
>> Perhaps it should say something like “when within the area clipped by the
>> element’s border edge” and use this definition:
>>
>> http://dev.w3.org/csswg/css-backgrounds/#corner-clipping
>
> Right. If "border box" isn't rounded, as various parts of various specs seem to assume without spelling it out, then adding a sentence and pointing to this section looks like the right thing to do.
>
> Maybe I should add a note anyway. Even if border-box is or will be defined to do the right thing, this seems hard enough to track down that it might be worth spelling out explicitly.
>
> But we probably need to establish a term (border shape? shaped border box?) that means "the not necessarily rectangular shape that follows the edge of the actual border, whatever it's shape, excluding border-image-offset". And while we're at it, tweak the phrasing to make sure that it still makes sense if the box is fragmented. css-box, if it was stable, would probably be a decent home for this, but since it's not backgrounds and borders is probably alright.
>
> Then once we have that term, both 'border-radius' and css-round-display's 'border-boundary' should clarify that they do affect that shape (and any other similar property, if I'm missing one), and css-shapes should probably refer to it rather than repeatedly paraphrase it.

I think you should just use the term "border edge", which is what border-radius
effectively redefines.

"Border box" sounds rectangular, and therefore should probably remain rectangular.

~fantasai

Received on Wednesday, 17 June 2015 15:48:43 UTC