[Standards-JIG] Re: jabber forwarding

Joe Hildebrand JHildebrand at jabber.com
Fri Feb 27 21:18:49 UTC 2004

As I'm currently on a plane, I don't know if anyone has responded to this

One of the issues with redirection is the from address, and the careful
trouble we've gone to in the XMPP world to ensure that from is difficult to
spoof.  SMTP doesn't suffer under this restriction.

A use case:

- sender at first.example sends to original at second.example
- original has set up a forward to real at third.example
- second.example connects w/dialback or S2S SASL TO third.example
- message is sent
- error occurs, since the from address is from first.example, not

Second.example could change the from address to be a local address, but it
would have to keep a mapping for each of these back to the original senders,
and all subsequent traffic would continue to go through the redirect server.

Hm... I guess that would work, if you have a relatively small number of
forwards.  You could add a JEP-33 replyto address of the original sender (so
clients could bypass the redirect if they could), and potentially send a
redirect back to the sender.

Joe Hildebrand


> -----Original Message-----
> From: Eeli Kaikkonen [mailto:eekaikko at mail.student.oulu.fi] 
> Sent: Thursday, February 26, 2004 5:34 PM
> To: standards-jig at jabber.org
> Subject: [Standards-JIG] Re: jabber forwarding
> (I found a bit old thread from the g.n.j.standards-jig 
> newsgroup. I Think it is still valid, and I have some 
> comments and questions.)
> Robert Norris wrote:
> > On Thu, Dec 11, 2003 at 02:50:06PM -0600, Peter Saint-Andre wrote:
> >> So, when are we going to define "jabber forwarding" -- the 
> ability to 
> >> have message sent to one of my accounts temporarily or permanently 
> >> forwarded to another of my accounts? Any thoughts on how 
> that should 
> >> work? Inquiring minds want to know.... ;-)
> > What other possible uses of this sort of thing can people think of?
> I belong to IKI. Here is info from http://www.iki.fi :
> "English summary: The Internet Users Forever IKI is a society 
> that provides its members, private individuals in Finland, 
> permanent iki.fi-addresses with e-mail and WWW forwarding 
> services (IKI does not host the web pages, it just forwards 
> the addresses).
>     This allows our members to keep the same personal 
> identity should the actual location or ISP of their e-mail or 
> www homepages change."
> Some people have recently discussed if IKI could provide a 
> jabber service.
> Natural way to do it would be some kind of forwarding, 
> because IKI tries to manage with as little resources as possible.
> As far as I have understood by looking at documentation a 
> www/http-like redirecting is not possible. I.e. a server 
> cannot say "this address X at jabber.somewhere.not is not here 
> in server jabber.somewhere.not, it is really 
> Y at jabber.somewhereelse.not". That would be permanent 
> forwarding to another account, if the requesting server then 
> would translate all of it's client's requests for 
> X at jabber.somewhere.not to Y at jabber.somewhereelse.not.
> (The idea of jabber is that client doesn't have to know about 
> details but servers should manage them, so the client 
> shouldn't know about forwarding - that would be unlike 
> www-redirecting even though it would be possible.)
> The closest equivalent which I could find was <redirect/> 
> error stanza from
> draft-ietf-xmpp-core-22 document, but it probably would 
> require all messages being redirected one by one when they 
> come - that would be like email forwarding. That would not 
> give any reason for using redirecting instead of the real 
> service if resources are considered. And additionally, if it 
> is server's purpose to redirect, then it is not an error at 
> all, and other servers do not need to know about redirecting. 
> What does that <redirect/> mean anyways?
> From Administration Guide I read that hardware or bandwidth 
> requirements are not big for running a real server. But 
> running a complete service would make administration harder - 
> you have to choose the services, update them regularly 
> (especially gateways for proprietary protocols, I think) and 
> users would be tied to your decisions. IKI's administration 
> is voluntary based so any extra work is bad.
> I would like to know if I have understood these things right. 
> Maybe the main point is if any kind of forwarding, relaying 
> or redirecting to lessen the load of the server, it's network 
> load and administrational work is possible according to 
> standards. The real-world implementation in server level is 
> of course the next step, but it does not belong here.
> Eeli Kaikkonen
> _______________________________________________
> Standards-JIG mailing list
> Standards-JIG at jabber.org
> https://jabberstudio.org/mailman/listinfo/standards-jig

More information about the Standards mailing list