Re: Property interactions and computed style

On Thu, Jul 12, 2012 at 8:44 PM, L. David Baron <> wrote:

> On Thursday 2012-07-12 20:22 -0600, Glenn Adams wrote:
> > On Thu, Jul 12, 2012 at 6:55 PM, L. David Baron <>
> wrote:
> >
> > > We probably actually need to start adding "Used Value"
> > > lines that describe what getComputedStyle returns.)
> > >
> >
> > I was thinking the same thing today. We could have a blanket rule that
> the
> > used value is the computed value except for those properties that
> > explicitly specify a "Used Value" line in their definition.
> That would be a bad rule to introduce unless we go through every
> spec first.  Otherwise we'd end up specifying things that disagreed
> with existing interoperability, and that would lead some
> implementers to actually change their implementations and break that
> interoperability.

Sure. Clearly it would not work unless such a review occurred and used
value specified when it differed from computed value. However, if one is
going to do this, then it would either need to be done in all instances or
a default rule applied, provided there were enough cases where it is
useful. My expectation is that there are many more properties where used
value always matches computed value than the other way around.

Received on Friday, 13 July 2012 03:59:51 UTC