Having a profile that leaves case mapping decisions open and a profile
that maps (in a separate spec) is definitely worse than having a profile
that doesn't map and a profile that maps in the same central location.
It's still two profiles, but they are both in a readily usable state.
The only caveat may be "but what variant of case mapping exactly"? If
the WG has discussed this already, then we should be fine.
Regards, Martin.
On 2014/11/24 04:52, Peter Saint-Andre - &yet wrote:
In draft-ietf-precis-saslprepbis we have a profile for usernames, which
differs from the profile for localparts of JabberIDs defined in
draft-ietf-xmpp-6122bis in only one respect: the localpart profile
specifies case mapping whereas the username profile leaves decisions
about case mapping vs. case preservation up to the using application.
I am wondering whether it would make sense to define two username
profiles in draft-ietf-precis-saslprepbis: one that uses case mapping
(say, Usernamecasemapped) and one that uses case preservation (say,
Usernamecasepreserved). This seems cleaner to me, and would avoid the
need to define a separate profile in the XMPP specs (which might not be
where future protocol designers would look for a username construct).
Thoughts?
Peter
_______________________________________________
precis mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/precis
