On 4/22/14, 9:39 AM, Adrian Farrel wrote:
Why isn't BCP18 an important reference?
Probably because, roughly speaking, BCP18 is to i18n as BCP61 is to
security. Plus much of BCP18 has been superseded by RFCs 3629, 4646,
5198, 6365, etc.
I like the idea of a BCP being superseded and still remaining as a BCP, that is
probably not helpful even if 4646 got stabbed to death by 5646.
What you are saying, I think, is essentially that RFC 2277 is not applicable to the space
of "Preparation and Comparison of Internationalized Strings in Application
Protocols".
So is RFC 2277 wrong, deprecated, or still applicable as written? And, in the
former two cases, is anyone doing anything about it?
RFC 2277 was written long ago (1998) and far away, when very few people
in the IETF paid attention to internationalization. As I see it, in
large measure RFC 2277 was intended as a call to do so, and provided a
few somewhat primitive guidelines regarding definitions (superseded by
RFC 6365), "charsets" [sic] (superseded by RFC 3629 and RFC 5198),
language tags (superseded by RFC 4646), etc. But I really don't see how
those early suggestions have much to do with the advanced work we're
doing on replacing stringprep. Heck, even the stringprep specification
from 2002 (RFC 3545), which this I-D will supersede if it is approved,
didn't reference RFC 2277.
Peter
_______________________________________________
precis mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/precis
