Paul Hoffman <[EMAIL PROTECTED]> writes: > > Probably RFC 2026, but if not there, it is certainly in the folklore > of How Things Are Done.
That's unfortunate. A documented process is a requirement for open standards development, in the opinion of many <http://blogs.sun.com/dennisding/resource/Open%20Standard%20Definition.pdf> > > >It would probably be easier to add them in a separate document, huh? > > Maybe, but that would then confuse the issue by having two docs > people would expect to have read in order to do the format. Given the > very marginal value of Draft Standard, we should probably just revise > and recycle at Proposed rather than split into two. There is one mistake in the Relax NG (atom common attributes on author or something) and at least one section that is somewhat unclear around the external div in XHTML. The i18n attributes seem needed to display text without a guess based on xml:lang. Maybe we don't even need unicode-bidi. I don't think it would be smart to take other features. Robert Sayre
