Re: [newprep] WG Review: Stringprep after IDNA2008 WG (newprep)

[Sam Hartman <[e-mail]>]

>>>>> "Alexey" == Alexey Melnikov <[e-mail]> writes:

    Alexey> Sam Hartman wrote:
    >>>>>>> "Jiankang" == Jiankang YAO <[e-mail]> writes:
    >>>>>>> 
    >>>>>>> 
    Jiankang> If there are many things we must do, we(WGs) normally
    Jiankang> prioritize the things.  sometimes, the easier one first;
    Jiankang> sometimes, the difficult one first.
    >> Sure.  That's fine for the WG to do.  I don't think it is good to
    >> do in the charter without some fairness criteria.  All items
    >> brought up by the time external review of the charter concluded
    >> seems like a reasonable fairness criteria.  Putting the cutoff
    >> before that seems unreasonable.
    >> 
    >> Obviously, the WG can internally prioritize (and change its
    >> priorities) within its normal administrative processes.
    >> 
    Jiankang> If peter's list is not ok for you, could you kindly give
    Jiankang> us your list?
    >> 
    >> The list in the charter plus:
    >> 
    >> 1) Considering Kerberos implications for SASLPREP revisions
    >> 
    >> 
    Alexey> Sam, I think this is granted. I don't think this needs to be
    Alexey> written into the charter, especially considering that there
    Alexey> is a good deal of overlap between people active in Kerberos
    Alexey> and SASL WGs.

I'm totally happy with that interpretation.

    >> 2) Considering RFC 4282.
    >> 
    >> 
    Alexey> Do you consider RFC 4282 to be a stringprep profile? My
    Alexey> quick scan of the document (in particular Section 2.4) is
    Alexey> not conclusive.

It's my read that section 2.4 invokes all of toascii from IDNA 2003,

which definitely implies stringprep under the covers.  However, I think
the spec is kind of broken both for IDNA2003 and IDNA2008; if you take a
look at the proposed erata aspects of the AAA community agree.  The
solution is probably not going to be a stringprep profile, although the
solution is probably not going to be any more complicated than something
like XMPP's solution.  However, I definitely think the AAA community
needs to have some internationalization experts to talk to in order to
have a chance of success.