[Standards-JIG] Re: JEP-45: Discovering occupants whose presences are not being broadcasted

Peter Saint-Andre stpeter at jabber.org
Thu Jan 13 20:10:52 UTC 2005


In article <cpq88f$m7j$1 at sea.gmane.org>,
 "Gaston Dombiak" <dombiak_gaston at hotmail.com> wrote:

> According to the JEP it's possible to query a room for its items. This will 
> return a list of existing occupants if that information is publicly 
> available. My question is how do you define "is publicly available"?

By "publicly available" in this context means "the room owner has 
approved of sharing this information", presumably via a room config 
option. Some implementations may assume that it is OK to share this 
information if the room is a Public Room (i.e., not a Hidden Room).

> IMO, if the room is configured so that some roles are not being broadcasted 
> (i.e. room may have invisible users) then the returned list shouldn't 
> include the occupants whose presences are not being broadcasted. Is my 
> interpretation of the JEP correct? Should the JEP include a note for this 
> case? 

Some implementations may allow "invisible users" (I call them "ghosts"). 
This is a bit of a grey area in the JEP, and intentionally so. For 
example, some implementations may consider all registered members to be 
"in the room" but never send presence from them until they join. So far, 
we have left such functionality up to the implementation, and have not 
talked about it in the JEP. I don't particularly want to add text on 
this point, since it will simply cause people to start believing in 
(MUC) ghosts despite the fact that they are not mentioned anywhere else 
in the spec. That said, if a room has "ghosts" then I don't think they 
should be returned in the list of (active) occupants; but that's more a 
personal opinion than a formal recommendation.

/psa




More information about the Standards mailing list