> Deep agreement with everyone who's noted
> that getting a warning about background-color: transparent
> is a crock.
So you're saying you agree with anyone who agrees with you
on this matter.
> If the standard says something's kosher, no warnings
> should be given.
There is no standard on CSS. Even if there were, I don't think it would
take position in religious issues.
> An accessibility checker might be a place to put this sort
> of thing.
Perhaps, and you are free to write such a checker.
> But not a validator.
The CSS Validator isn't really a validator. It performs a collection
of checks, and issues some error messages (about things that seem
to violate some requirement in something that has to do as an excuse
for a specification of CSS du jour) and some warnings (about things
that are typically oversights or just bad practice).
> We who care about valid
> coding like whistle-clean validation results.
You can write a program that whistles whatever you like. If you regard
formal conformance to a quasi-specification as a measure of quality,
you're completely wrong, but that's your privilege.
You can switch off the warnings if you do not want to get warned
about oversights and mistakes that don't make your style sheet
technically wrong. What is the remaining problem?
> Somewhere in the validator bureaucracy there's a
> decision-maker who's acting in an arbitrary manner.
There's no bureaucracy there. Whether that's a good thing is
debatable. (Bureaucracy is one of the greatest advances of civilization
and development, though it has some rather nasty side effects.)