> > This becomes complicated (for the case in question) when you start
> > adding pluggable new URI schemes to Mozilla. The "sanity-checking" code
> > would have to be updated for each scheme.
>
> I don't see why that is a problem. "Basic" sanity checking for
> conformance to the generic URI syntax (RFC 2396) can happen in a
> centralized location, but anything beyond that is--and *should*
> be--scheme-specific.
The problem is not checking the standards, but how to interpret ones which
are non-standard. Asking new scheme authors to implement random
"suggestions" for broken URLs for their scheme is a bad idea.
<rant>Look! This idea sucks! If we start popping up dialog boxes asking
how we should interpret invalid HTML, where's it going to end?</rant>
<retires from argument>
Gerv
