Re: [css3-ui] Review comments

On 07/14/2011 01:49 AM, Tab Atkins Jr. wrote:
> Section 4.1.1
> -------------
>
> * I recommend stating explicitly that when these pseudoclasses match
> is determined by the host language, and the definitions in the spec
> are merely informative descriptions of each.

If a host language doesn't specify, I think the UA should be allowed
to interpret Selectors onto it. So while the spec should have clear
hooks for a document language spec to be more precise, the prose in
UI should nonetheless be normative.

> * I recommend referencing HTML as the informative reference instead of
> XForms, as HTML is the language against which these are most likely to
> be used.  The relevant link is
> <http://www.whatwg.org/specs/web-apps/current-work/complete/links.html#pseudo-classes>
> (or the similar link to the W3C version of the spec).

I recommend linking to both. Why not, it's informative.

> * Given that we deprecated the system colors, should we deprecate/drop
> the system fonts as well?

Probably a good idea. I'm sure jdaggett would cheer that, too. :)

> * You can do fallback images now simply by using the image() function,
> defined in Image Values.  There are also many more types of images
> than just<url>.  I recommend replacing this with just<image>.  (If
> you'd like to avoid a dependency on Image Values, define<image>  to be
> equal to<url>  for now, with a note that other specs may expand its
> definition.)

I recommend following css3-background's lead here:
   http://www.w3.org/TR/css3-background/#value-types

> * The 'inherit' value is globally defined for all properties, and the
> preferred style appears to be to omit it in property definitions.

Agreed, we should be consistent across specs. It causes confusion otherwise.
(And yes, there have been posts a www-style about confusion on this exact
issue. It is not a theoretical concern.)

~fantasai

Received on Tuesday, 21 February 2012 15:29:23 UTC