David Hucklesby wrote:
> Otherwise, I guess I misunderstand the purpose of validation.
No, I don't think you have misunderstood _one_ of the purposes; to
check that all is "correctly typed" = "technically valid". As we all
should know: validity is easy to achieve - regardless of whether it
works or not.
However, there's another way to use validators, and that is to check
that we haven't _accidentally_ made a typing-error. The use of
IE-expressions is not accidental (I think), so the fact that they are
flagged as errors should not necessarily be seen as a validity problem.
IE-expressions clearly are *not* valid - wherever they are, but
they can solve a problem that validity can't - when used properly.
The use of non-valid code in order to work around a limitation in a
browser, is sometimes unavoidable. The practice can as such be placed in
the same garbage-bag as nonsensical but technically valid hacks, which
are also unavoidable at times.
Pity the validators can not check if something makes sense or not, as
that would make them a lot more useful. After all; validators are there
to help us make our designs work - and nothing else, IMO. Otherwise the
validators may become more of a limitation than weak browsers, again IMO.
Separating the _known_ 'non-valid and nonsensical stuff' from the
'better stuff', makes sense, even if it doesn't improve on validity.
How one achieve such separation is a matter of available methods and
personal choice. The fact that some separation-methods "hide" code from
the validators, is a less important side-effect - at least to me.
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/