Re: [css3-exclusions] Shapes depend on which box?

On Tue, Nov 27, 2012 at 1:59 PM, Alan Stearns <stearns@adobe.com> wrote:
> On 11/27/12 1:51 PM, "Tab Atkins Jr." <jackalmage@gmail.com> wrote:
>>On Tue, Nov 27, 2012 at 1:37 PM, Alan Stearns <stearns@adobe.com> wrote:
>>> The current draft of the spec says that the coordinate system and
>>>resolved
>>> percentages for declared shapes uses the border box of the element. I am
>>> thinking it might make more sense to use the content box of the element.
>>> As it stands, specifying a 100% width and height rectangle to
>>>shape-inside
>>> can change how its inline content is laid out (depending on the border
>>>and
>>> padding). If we change the coordinate system and percentages to use the
>>> content box, then a 100% width and height rectangle for shape-inside
>>> changes nothing, and modifications to percentages are relative to what
>>> you'd get without defining a shape-inside.
>>
>>Your rectangle argument is convincing.  This sounds fine to me.
>>
>>However, people might actually want border-box sizing.  Have you given
>>though to adding an optional <box> value to the properties, defaulting
>>to "content-box"?
>
> Hmm - what if we used the value of box-sizing?

That makes sense to me.  Then we can punt on manual control until it's
proven we need it.

>>> As for shape-outside, the current definition says that a 100% width and
>>> height rectangle for shape-outside on a float would shrink the float
>>>area
>>> from the margin box to the border box. Making the change would further
>>> shrink the float area to the content box, which isn't any less confusing
>>> than before. I'm assuming a single, consistent definition of how lengths
>>> and percentages work with shapes is preferable to having separate
>>> definitions for shape-inside and shape-outside (particularly when you're
>>> using the same shape for both).
>>
>>Yes, consistency is probably best, so you can easily just give the
>>same values to both.
>
> (just FYI) You can just specify the shape once with shape-outside - by
> default shape-inside takes on the computed value of shape-outside.

Yes, that's why I agree that consistency is best.  ^_^

~TJ

Received on Tuesday, 27 November 2012 22:35:54 UTC