[Standards-JIG] PEP and Pubsub Questions

Peter Saint-Andre stpeter at jabber.org
Fri Apr 7 20:53:58 UTC 2006

Hash: SHA1

Peter Saint-Andre wrote:
> Matt Tucker wrote:
>>> 2) Example 9 and the subsequent paragraph mention that the sender of
>>> notification messages should be the full JID of the user. How should the
>>> PEP service handle error messages that were sent back to the full JID?
>>> Should the PEP service intercept those messages and process only those
>>> with a known ID or otherwise forward them to the user? This question is
>>> related to http://www.jabber.org/jeps/tmp/jep-0060-1.8.html#impl-bounce.
> Good question. I have to think about the error handling. The publishing
> client would be PEP-aware since it did the publishing, but right now it
> doesn't have a way to tell the PEP service about any errors or to remove
> subscribers because of bounces. Hmm.

Hmm, this is icky. I see five options:

1. Force bounce processing onto the publishing client. This makes it
unecessarily harder to write clients.

2. Require the PEP service to watch for error stanzas intended for the
publishing client. This is resource-intensive.

3. Recommend that the PEP service periodically send test messages from
the publishing entity's bare JID. This seems like an awful hack.

4. Enable the publishing client to forward errors to the PEP service.
This is problematic since we don't have a way to forward stanzas.

5. Require the PEP service to send from the bare JID and use SHIM
headers to specify the publishing client. This goes against one of the
reasons for defining PEP in the first place.

/me ponders


- --
Peter Saint-Andre
Jabber Software Foundation

Version: GnuPG v1.4.1 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 3641 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://mail.jabber.org/pipermail/standards/attachments/20060407/f8dc696d/attachment.bin>

More information about the Standards mailing list