Sam Ruby wrote:
> [snip]
> 1) You do realize that there is a [EMAIL PROTECTED]
> mailing list, don't you. :-)
> [snip]
Yes, but I felt that this was more of a spec issue than of any single
impl. Specifically, I am concerned that other implementations might get
the wrong idea by looking at how the validator acts and emulating it's
behavior with regards to unknown attributes.
> [snip]
> follow this (atom-syntax) mailing list are, and would be confused. In
> such cases, simple (it either allowed by an existing spec, or it is not;
> if allowed, it either generally works interoperably or it doesn't) is
> better.
>
Or it is generally allowed by the spec but there are a number of subtle
and not-so-subtle issues that could impact interoperability. I
understand the need for simplicity, but I think it is equally important
to educate implementors about the various subtleties inherent in the
spec. For instance, there is a common misconception that "foreign
markup" only covers stuff outside the Atom namespace; however, Section
6.2 makes it very clear that that's not the case.
- James
