At 03:52 PM 10/22/00 +0200, Harald Alvestrand wrote:
What turned out in practice was that the most important part of RFC 1766 was
the registration procedure, including the references to authoritative sources
of tags (ISO 3166 and ISO 646).
I agree that this is an important part.
This clearly belongs in the BCP category, as do the registration procedures
for charsets and MIME types; there is no meaningful way there can be more than
one such document.
I concur that the registration procedure belongs in the BCP category.
The rest of the technical content, namely the syntax of tags (note that the
Content-Language: header itself is moved to another draft, which WILL be
standards-track) COULD have been moved to a THIRD document, standards-track,
but this did not seem like the Right Thing to do; also, if published at
Proposed, it would have required progression up the standards track before any
referring document could progress. Did not seem right either.
If the technical specification has significantly changed, yes,
the issuing at Proposed is quite appropriate. Issuing at BCP
allows for the technical specification to bypass the 3 step
maturity process.
Yes, this is a small technical specification, but what concerns
me is the precedence set or extended by allowing BCPs to contain
technical specifications used in standard track protocols.
Kurt