Re: [css3-images] Images with an intrinsic aspect ratio and exactly one of an intrinsic height or width

On Wed, Mar 14, 2012 at 4:44 PM, L. David Baron <dbaron@dbaron.org> wrote:
> On Wednesday 2012-03-14 14:54 -0700, Tab Atkins Jr. wrote:
>> On Wed, Mar 14, 2012 at 2:04 PM, L. David Baron <dbaron@dbaron.org> wrote:
>> > The prose in http://dev.w3.org/csswg/css3-images/#default-sizing
>> > (5.3, Default Concrete Object Size Resolution) seems to assume that
>> > if an image has an intrinsic aspect ratio, it either has neither or
>> > both of an intrinsic width and height.
>>
>> That's not an assumption, it's a basic fact.  You can't have only two
>> of them; the third is automatically determined by the other two.
>> Should I make this more obvious somehow?
>>
>>
>> > However,
>> > http://www.w3.org/TR/2011/REC-SVG11-20110816/coords.html#IntrinsicSizing
>> > certainly defines cases where there's an intrinsic aspect ratio
>> > (from a viewBox) and only one of an intrinsic width or height.
>>
>> What cases are these?
>
> An SVG image that has a viewBox attribute and a non-percentage value
> for exactly one of the height or width attributes.
>
> It's reasonable to want to describe such an image as having both an
> intrinsic width and an intrinsic height, but if you want that to be
> the case, something should say so.  (I'd suggest putting it in the
> definition of "intrinsic dimensions" in 5.1 (Object-Sizing
> Terminology).)

I've clarified the "intrinsic dimensions" definition in response to
another of your bugs, and I think it satisfies what you're asking for
here now:

# An object may have no intrinsic dimensions (such as CSS gradients),
only one intrinsic dimension (SVG images designed to scale may have
only an aspect ratio), or all three intrinsic dimensions (all raster
images, for example).  <span class='note'>(Note: an object cannot have
only two intrinsic dimensions, as any two automatically imply the
third.)</span>

Is this acceptable?

~TJ

Received on Thursday, 15 March 2012 19:37:41 UTC