[standards-jig] JIDs

Iain Shigeoka iainshigeoka at yahoo.com
Mon Feb 4 19:45:11 UTC 2002

On 2/2/02 5:53 AM, "Matthias Wimmer" <jabber at matthias-wimmer.de> wrote:

> Is there any clearer documentation of what characters are allowed within
> JIDs than http://docs.jabber.org/proto/html/identity.html?
> The above document just reference to RFC2396. But I don't think all our
> used JIDs conform with this RFC. Think about the two SMTP gateway
> implementations we have. They use "user%smtp-domain at gateway-domain" as
> JID. This JID doesn't conform the BNF grammar in appendix A. The problem
> with it is the "%". This character is (according to the given syntax)
> only allowed as first of three characters in an "escaped" (the other two
> characters have to be hex digits.

I think the document is the best source for information and that the
referenced RFC is the correct standard for conforming JIDs.

> Should the "%" character be allowed within the user part of a JID or
> not? If it's allowed we can't refer to RFC2396.

Ironically,  I think that the reality of the situation is that many things
just don't comply with the standard.  :)  This is a combination of people
not reading the spec, no compliance tests, and people operating under
Postel's robustness principle ("be conservative in what you send, liberal in
what you accept").  In fact, the gateway usage of the % in the username is
sort of a de facto standard as most of us like to be consistent with the
established patterns in existing gateways.

So you are correct in both the standard and the fact that many things are
non-compliant.  Which raises an interesting standards-jig question: should
we try and crack down on these violations, change the standards to fit the
implementations, or continue with the status quo?


Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com

More information about the Standards mailing list