On 10/19/2011 10:56 PM, Kevin Smith wrote: > On Wed, Oct 19, 2011 at 4:51 PM, Sergey Dobrov <[email protected]> wrote: >> On 10/19/2011 10:25 PM, Kevin Smith wrote: >>> On Tue, Sep 27, 2011 at 5:35 PM, Sergey Dobrov <[email protected]> wrote: >>> The behaviour here is pretty much by design - the PEP defaults are >>> there for mutually shared information (e.g. extended presence) between >>> people with mutual presence subs. If you want a one-sided approach, >>> using manual subscriptions instead of the caps-based magic seems like >>> a better fit. >>> >>> /K >>> >> don't you think that such behavior is not consistent with simple >> presences? I mean, most of protocols that based on PEP are just >> extensions for simple presence mechanism and that's really unuseful that >> it's impossible to use them in a sheaf with subscriptions. A PEP service >> can't be implemented outside a jabber-server, so why we can't allow it >> to follow "probe" presences? > > Because sending probe presences isn't the same as sending available > presences - and if it was, it'd be a presence leak. > > /K > Why is it a leak? The presence will be sent only if we want to know someone's status, so why we can't give our caps?
-- With best regards, Sergey Dobrov, XMPP Developer and JRuDevels.org founder.
