> If the CSS doesn't display properly, then it's truly invalid.  But, 
> if the CSS works on the screen, that's the true test of validity.

There's a problem with such a "validity statement", in that the browser
"behind the screen" may be (and in this particular case in fact is)
broken. If we rely entirely on what browsers do for deciding what's
valid and what isn't, then we may have to live with all the weak and
broken browsers for a long time, because their engine-bugs _can't_
really be corrected once enough sites are relying on same bugs.

This is in fact why it's now so hard (nearly impossible) to create a
standard-compliant version of IE/win, as billions of sites are based on
the engine-bugs in existing versions, and a large percentage of those
sites will break in a truly corrected version of IE.
Some sort of "standard compliance opt-in" is one of the alternative ways
this dilemma can be "solved", which probably means the next versions of
IE/win will not be truly standard-compliant.



What's rendered in a browser, any browser, says nothing about validity -
only about standard-compliance in that particular version of that
particular browser.

Web standards OTOH, _are_ standards that the browsers _should_ comply
with, and the word "valid" can only be used when something - a browser
or site - is measured up against these standards. The fact that no web
standard is perfect or complete and no browser flawless, has as little
to do with anything as the color and shape of "valid" icons. Both
standards and browsers need constant improvement - and always will until
we no longer need them.

Whether new browser-versions can't comply with standards because the
browsers are badly made to begin with, or because site-authors have
already locked them down by relying on their old bugs, doesn't make
much of a difference. They don't comply.
The result will be continued weak standard-support and/or "opt-in
branching", which will make it harder for all site-authors in the future.

I don't think "making web design harder" is what we'd like to be
remembered for a few years from now, so we should look a bit ahead and
rather not legitimate misuse of browser-bugs _whenever_ that can be
avoided. It is only when we don't really have a choice that "bug-riding"
_can_ be legitimated - to some degree, but lack of choices doesn't make
such practical solutions "valid" - no matter what ends up on screens.

regards
        Georg
-- 
http://www.gunlaug.no
______________________________________________________________________
css-discuss [EMAIL PROTECTED]
http://www.css-discuss.org/mailman/listinfo/css-d
List wiki/FAQ -- http://css-discuss.incutio.com/
List policies -- http://css-discuss.org/policies.html
Supported by evolt.org -- http://www.evolt.org/help_support_evolt/

Reply via email to