> On July 5, 2011, 7:41 p.m., David Edmundson wrote: > > presence.cpp, line 45 > > <http://git.reviewboard.kde.org/r/101858/diff/1/?file=26121#file26121line45> > > > > This isn't really ref counting. > > > > It will say how many of this type of object are in the system, not > > count how many pointers there are to this one instance. (actual ref > > counting) > > > > Anyway, I'd be very surprised if plasma worked this way. I'm pretty > > sure if an other plasmoid starts using it, it will increase the ref count > > of the dataengine and NOT call the constructor again. > > > > I would suggest looking where the parent class has it's ref counter > > increased. > > > > (arguably this will work to tell if the dataengine exists or not.. > > which is probably actually all that's needed in reality.) > > > > tl;dr pretty sure this won't work. Happy to be proved wrong if you have > > evidence otherwise. > > > > > > Martin Klapetek wrote: > You're right about the refCount, it is not an actual refCount as that > variable was initially used for refCount and then I changed it to static var, > but left the name like this. In the end it doesn't really matter, it works as > it should, really ;) It works in such way, that the dbus interface exposes a > method for returning this refCount var (whatever it really is) only if the > dataengine is created, when the dataengine is not in use (ie. deleted), the > dbus method does not exist, so in the end it serves its purpose. I can change > the variable to something like int isActive, but I think it's enough for a > temp workaround.
Change the name and I'll approve it. Actually - why have this variable at all? If you get a dbus response it's active. If you don't - and get a dbus error, then it isn't. You may even simply be able to see if the dbus address exists and not need any methods in the address at all. - David ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://git.reviewboard.kde.org/r/101858/#review4409 ----------------------------------------------------------- On July 5, 2011, 7:15 p.m., Martin Klapetek wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://git.reviewboard.kde.org/r/101858/ > ----------------------------------------------------------- > > (Updated July 5, 2011, 7:15 p.m.) > > > Review request for Telepathy. > > > Summary > ------- > > This simple patch adds a simple dbus adaptor to the dataengine, so it can be > checked if the plasmoid is in use (the dataengine is created only with the > plasmoid, when not needed, it is deleted). Next part will be checking from > contact list for this dbus entry/method. It is meant as a temp workaround for > bug 270675, once the upstream has proper features, we will remove it. > > > This addresses bug 270675. > http://bugs.kde.org/show_bug.cgi?id=270675 > > > Diffs > ----- > > CMakeLists.txt 69b839a > presence.h 9b199c0 > presence.cpp 34fa5ea > > Diff: http://git.reviewboard.kde.org/r/101858/diff > > > Testing > ------- > > Added the plasmoid, dbus interface was created, removed the plasmoid, dbus > interface gets removed (though the path still stays, but the method is not > valid anymore). > > > Thanks, > > Martin > >
_______________________________________________ KDE-Telepathy mailing list [email protected] https://mail.kde.org/mailman/listinfo/kde-telepathy
