[Standards] Domain & JID Aliasing

Chris Mullins chris.mullins at coversant.net
Thu Feb 1 19:15:22 UTC 2007

Many email systems provide aliasing both of domain names, and user accounts. 

For example, via email, I can be reached at:
cmullins at coversant.net
cmullins at coversant.com
chris.mullins at coversant.net
chris.mullins at coversant.com

We're getting to the point in XMPP land where we need this same capability. For example, I currently give out "cmullins at coversant.net" as my XMPP address, but would prefer to give out "cmullins at coversant.com". I don't want two separate accounts, so domain aliasing is the obvious answer.

This brings up a few problems though:
1 - If a user subscribes to my presence as "cmullins at coversant.net", and I send him a message, what "From" address do I stamp on it? Do I use the original address that he subscribed to? It seems like I should. This means a server would need to keep track of the "alias" that was subscribed to, so that outgoing packets to that endpoint can have the proper from address.

2 - If I subscribe to someone else's presence, which of the alias accounts should the subscription come from? "The master" account seems the obvious answer, and we could dictate that all packets sent to an unknown remote contact come from the "master" account. This is how Exchange seems to do it. 

3 - Would we want to tickle the subscription management flow to specify a "change address" field? That is, perhaps I'm subscribed to "stpeter at jabber.org" as "cmullins at coversanet.net". It would be nice to send "<change newAddress='chrismullins at coversant.com'/> or a packet to that affect. 

4 - should there be a way for a client to "see" it's aliases? Right now, my email client can't do that (Outlook 2007), and it seems un-needed for an XMPP client as well.

I know Google Talk is currently doing some form of aliasing between gmail.com & googletalk.com, but I've not looked into the details on how it works. Any other servers offering this today? (XCP? OPN?)

Chris Mullins

More information about the Standards mailing list