Re: [EAI] POP (draft-ietf-eai-rfc5721bis-05), SASLPrep, and Stringprep

[John C Klensin <[e-mail]>]

Jiankang,

My apologies.

If you considered the fourth paragraph carefully before making
the changes, I think the changes are reasonable.  My personal
preference, although not a strong one, is that we try to
future-proof these documents -- and raise the odds as much as
possible that an update to SASLPrep will explicitly update this
and other documents as needed.

I had assumed that you had not considered those issues (similar
things did occur with some changes in prior versions) and
erroneously assumed that you had not done so here.  Again, I
apologize for jumping to that conclusion and for any comments I
made as a result.

I take the point make by others that either the authors of such
a SASLprep revision or various people reviewing that spec will
spot this.  But there are a _lot_ of interconnections.  I'd
really prefer to avoid having to reopen this spec when a
revision to SASLprep occurs -- the difficulty getting reviews
and comments while this WG is active predicts that it might be
even more difficult if/when revisions are needed after the WG
has concluded its work.

    john


--On Tuesday, July 17, 2012 22:24 +0800 Jiankang Yao
<[e-mail]> wrote:

>> If we picked (2) with "... or one of its standards-track
>> successors", then I have some concern whether the following
>> paragraph (the 4th in section 3.2, about how to apply
>> SASLprep on different arguments) may trip or confuse readers.
>> >
> 
> I do not think that it will cause confusion.
> 
> The fourth paragraph of section said "When applying SASLprep
> [RFC4013], servers MUST reject UTF-8 user
>    names or passwords that contain a Unicode character listed
> in Section
>    2.3 of SASLprep [RFC4013]. "
> 
> That paragraph is for RFC4013- SASLprep ,  not for  its
> standards-track successors since the reference to SASLprep is
> rfc4013 not  its standards-track successors .
> So it will not cause confusion.
>...