[Standards] certification etc.
jd.conley at coversant.net
Thu Mar 22 05:29:45 UTC 2007
I agree with Ian. It's also useful in combination with direct SSL (i.e. 5223) to limit round trips if you have a high latency connection.
I was also wondering about XEP-0077. In the current Basic profile it's recommended, which is fine. But I am seeing these examples on the list with it changing to REQUIRED. IBR is interesting, but it seems people use XMPP for the most part in the enterprise with central administration or a hosted service where they don't want people creating accounts through an automated fashion on their standard c2s ports. Though we obviously have this implemented already we'll probably be making it much less accesible in our client software (and only if the server sends out the stream feature). Anyway, I vote for keeping it RECOMMENDED.
Just my .02c.
From: standards-bounces at xmpp.org on behalf of Peter Saint-Andre
Sent: Wed 3/21/2007 3:50 PM
To: XMPP Extension Discussion List
Subject: Re: [Standards] certification etc.
Remko Tronçon wrote:
>> And Intermediate 2008 would be:
>> Everything in Basic 2008 AND
>> XEP-0077: In-Band Registration REQUIRED
>> XEP-0078: Non-SASL Authentication REQUIRED for servers; NOT
>> RECOMMENDED for
> The NOT RECOMMENDED can be dropped?
Ian Paterson has pointed out before that non-SASL authentication is
quite a useful protocol, in that it enables coders to quickly connect to
a server for testing purposes (wow, I'm connected, XMPP is easy, now I
can figure out how this SASL stuff works). So I think it would always be
good for server codebases to implement it (perhaps as an add-on module)
and for libraries to include support for it. But at some point (2008,
2009?) we may want to discourage production services from deploying it.
XMPP Standards Foundation
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 5103 bytes
Desc: not available
More information about the Standards