[jdev] Gaim and gnomemeeting using jabber

PUYDT Julien julien.puydt at laposte.net
Wed Dec 1 08:28:02 CST 2004


Le mercredi 01 décembre 2004 à 09:59 +0000, Richard Dobson a écrit :
> > You don't need to know the SIP uri to know that a client has one.
> 
> Exactly this is my whole point here, this guy seems to want to use a 
> voip-uri sent as a presence extension to indicate that he is in a call, as 
> others have already pointed out as well as me presence is not the correct 
> place for this, there is also the point that I was trying to make that why 
> do you even need the uri there in the first place, as you point out you dont 
> need the voip-uri to know the user is in a call so why even use it for this 
> purpose?

Notice that voip-presence and im-presence are not entirely unrelated:
while it is false to assume that voip-away implies im-away, being
im-away certainly means you're voip-away.

I really think that:
* using presence isn't as stupid as you think ;
* using jabber:x:oob to send the uri isn't against the standard ;
* it works _now_, and not in 10 years when people will have finished
debatting the issue ;
* it makes a very simple patch that has a chance to get upstream now ;

> > Once you want to intitiate your SIP conversation you retrieve the URI. The 
> > mechanism for this, well that will depend on your requirments. You could 
> > define a simple IQ request for an x:oob packet (this should appeal to you 
> > if you think security is important, since you could send an error if you 
> > don't want to share).
> 
> Good thats exactly the sort of thing I would want it to be.

Bad. You assume that deciding if the call is done or not is jabber's
problem: it's not. There's a voip-client there: it will ask the user for
confirmation. When someone has your voip-uri, you still have your word
to say.

Besides, aren't the presence packets only sent to contacts for which I
already said they were authorized to know about me?

> > But an argument could be made for putting it in  disco or vcard-temp, 
> > since that would enable offline retrival.
> 
> Yup, I also wasnt quite sure how you would advertise the voip-uri using 
> disco because I thought it was supposed to be for advertising capabilities, 
> but anyway neither will work because they do not get dynamically updated 
> (i.e. resent to people when something changes), so they are useless for 
> expressing voip presence.

Disco would be nice. But I don't remember seeing gaim sent disco request
to contacts. Only to the server.

> > What's the case for pubsub anyway? It's not dynamic..
> 
> Urm yes it is dynamic, i.e. when you update it it is resent to the people 
> who want to know about it.
> 
> > in fact, I don't  even *want* the uri to be updated, *unless* I put in a 
> > call.
> 
> As pointed out above you wouldnt even want it to be a uri, you would want it 
> to be something more generic to represent the voip presence i.e. "incall" or 
> "notincall".

No problem with sending the uri... the voip client isn't a web server:
it doesn't serve each and every request blindly. Please let the voip
client do its job.

> > Such a  message would be just as much "polluting" to the network as 
> > putting your  SIP uri in the presence.
> 
> No it wouldnt as using pubsub only the people who wanted to know about it 
> would be send the info (so its specifically not polluting the network), but 
> using a presence extension everyone would get it regardless of wether they 
> wanted it which is "polluting" the network.

Hmmm... as far as I remember, I had to allow my contacts to see me, and
they had to decide they wanted to see me... so telling them "I'm there"
isn't polluting the network, it's using it.

> > The only thing I can imagine you want this  dynamic is "I have it" or "I 
> > don't have it".
> 
> Pubsub is there so just like presence you can use it to alert people (just 
> the people who actually want to know about it) that you are in calls or not, 
> as this person seemed to want to do, from his previous emails he seemed to 
> want to be expressing his voip presence (i.e. wether he is in a call or not) 
> completely separately from what his IM presence says, now I could only see 
> this being realistically done using pubsub, since as others have pointed out 
> a presence extension for this is not the right way, also im not quite sure 
> how it could fit in with JEP-115 and still be inline with the spirit of it, 
> i.e. expressing the capability of being able to do voip is something quite 
> different from expressing that you are in a call or not.

Not completely separately. There's an implication, but not an
equivalence, between voip-presence and im-presence.

And yes, it could be nice to be able to separate:
* advertize voip is possible ;
* advertize the voip-uri ;
* advertize the voip-presence.

BUT I would like to do as much as possible *NOW* and not in ten years.

And as I already pointed out: I don't really want to _work on_ jabber,
but merely _use_ jabber. I would like some simple, standard and
_current_ mechanism to allow to call jabber-im contacts. Then provide a
patch for gaim (must be simple or they won't accept it), then for
gossip, then for ...

For the moment, most of the discussion was:
* uh, are you sure it's presence?
* no, you really should read all of the existing and obsolete protocols,
then propose one, fix it, get it through the standardization process.
* really, your remote contact should be asked if he's interested in
knowing about voip, then if he would accept calls from you, then for
each call if really he means he wants to get it through!

The goal is to have something that people will use! And will use _now_!

JP




More information about the JDev mailing list