|
Posted by Alan J. Flavell on June 7, 2006, 9:52 am
Please log in for more thread options
On Wed, 7 Jun 2006, David Stone wrote:
>
> > Not quite true. Since Mozilla-based browsers understand enough of
> > CSS2, you can do this via a stylesheet, as was recently discussed
> > http://groups.google.co.uk/group/alt.html/browse_thread/thread/e4b2a487ec8e02b
> > 4
>
> I looked at that. It's a bit more work than I care for with my
> particular table instance.
Seems quite a modest amount of work.
> Curiously enough, the background colour of a COLGROUP is one of the
> things that has supposedly been fixed, at least for the current
> build. (Well, as near as I can understand what bugzilla seems to be
> saying: https://bugzilla.mozilla.org/show_bug.cgi?id=4510)
But that's easy to fix: it only needs for the background colour to be
established on the col or colgroup element, and then allowed to "shine
through" all the elements (tr, td, whatever) which come later. And
that's their natural inclination, if their b.g colour is not
explicitly styled.
Whereas, getting HTML attributes of col and colgroup to percolate down
to the cells calls for some kind of action at a distance, and the
specification is very much at odds with the concepts of CSS, so you
have two different languages trying to pull in different directions.
It's frankly a mess, and I have some sympathy with the implementers
here.
> > > The next time someone tells me I should be standards-compliant,
> > > I think I shall just blow a big fat raspberry at them :P
> >
> > Nevertheless, it doesn't pay off in the long-term to keep relying
> > on browser bugs. In this case, you just need two sets of
> > specification-conforming settings, and you'll pretty much cover
> > the field.
>
> I don't see how I'm "relying on browser bugs"?
I don't see how I was suggesting that you *were*. You were talking
about blowing some *future* raspberry at the specifications, and I was
cautioning against the consequences of doing that.
ttfn
|