[Psi-devel] RC1 vs. suspend

[James Chaldecott]

I've looked at the sources & can't immediately see why it wouldn't be
working.

The detection of the PC going in/out of hibernation is in:

/trunk/src/tools/systemwatch/*

The handlers for those signals are in

/trunk/src/psicon.cpp

Looks OK to my untrained eye, though:

void PsiCon::doSleep()
{
        setGlobalStatus(Status("",tr("Computer went to sleep"),0,false));
}

void PsiCon::doWakeup()
{
        // TODO: Restore the status from before the log out. Make it
an (hidden) option for people with a bad wireless connection.
        //setGlobalStatus(Status());

        foreach(PsiAccount* account, d->contactList->enabledAccounts()) {
                if (account->userAccount().opt_reconn) {
                        // Should we do this when the network comes up ?
                        account->setStatus(Status("", "",
account->userAccount().priority));
                }
        }
}



On 22/06/07, James Chaldecott <james at swythan.co.uk> wrote:
>
> Hmm...
>
> I've now found this thread [1] that suggests it wasn't tuned off on
> purpose.
>
> Does it still work on OSX?
>
> James
>
> [1] http://thread.gmane.org/gmane.network.jabber.psi.devel/5875
>
> On 22/06/07, James Chaldecott <james at swythan.co.uk > wrote:
> >
> > It was me (and yes, I sort of watch the lists still).
> >
> > I've got a feeling that someone disabled this at some point, but I
> > haven't checked, yet.
> >
> > The reason I thought it was necessary was that (I think) my ADSL router
> > (with NAT) was still keeping the connection live when my PC went into
> > hibernation (at least there was some very long time-out somewhere) .
> > Packets from the server would then seem to black-hole on my side of the
> > network, but the server wouldn't drop the connection.
> >
> > The result was that (at least on some servers) I would appear to be
> > online for quite a while after the PC went into hibernation, and any
> > messages sent to me would never be delivered.
> >
> > I chimed in the the main jabber-dev list a couple of times to point out
> > this as a use-case that requires end-to-end ACK support in XMPP, but gave up
> > as I've not really got a voice there (I'm just a lurker).
> >
> > James
> >
> > On 20/06/07, Remko Tronçon < remko at el-tramo.be> wrote:
> > >
> > > > Make me a debug build please, I can test this.  As I recall it was
> > > > LonelyPixel who created this code.  Are you listening Yves?
> > >
> > > According to the sources, it's James Chaldecott.
> > >
> > > cheers,
> > > Remko
> > > _______________________________________________
> > > psi-devel mailing list
> > > psi-devel at lists.affinix.com
> > > http://lists.affinix.com/listinfo.cgi/psi-devel-affinix.com
> > >
> >
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.affinix.com/pipermail/psi-devel-affinix.com/attachments/20070622/d77da4a5/attachment.html