> > An implementation should only claim to support an attribute-token with no
> > attribute-namespace if it follows the behavior specified by a draft of the
> > C++ standard or of a technical specification produced by ISO/IEC
> > JTC1/SC22/WG21. An implementation should only claim to support an
> > attribute-token with an attribute-namespace if it follows the behavior
> > specified by the vendor identified by the attribute-namespace.

> > [OPEN QUESTION: Do we want to provide recommendations like this last
> > paragraph at all? If so, should we list the currently-know
> > attribute-namespaces? Having a centralized list of them is useful, and this
> > seems like a good place to maintain that list. Would it be in-scope for the
> > features SG to maintain a list of the known vendor extension attributes?]

> I don't think the last paragraph is needed, not that I'd object if someone
> really wanted it.

To me, this seems so obvious that I would really hope we wouldn't have to
say it.

> I don't think maintaining a list of known vendor extension attributes is in
> the charter of SG10. If one were maintained, would it be just the names,
> the names and the syntax, or the names, syntax, and semantics?

I agree with John: we don't want to keep track of all the vendor extension
attributes. But I think it would be great if we could maintain a table that
correlated attribute-namespace names with URLs at which vendors document
their list of extended attributes. (Assuming, of course, that vendors
actually use attribute-namespaces.)

Clark