On May 18, 2010, at 1:27 PM, Paul Davis wrote: > > > On Tue, May 18, 2010 at 4:15 PM, Paul Davis <p...@linuxaudiosystems.com> > wrote: > > > On Tue, May 18, 2010 at 3:55 PM, Colin Walters <walt...@verbum.org> wrote: > Oh, one thing I really need from OS X porters (and win32 for that > matter) is a solid recommendation for an application identifier > string. GApplication needs this for uniqueness, and GSettings > requires one too. > > Probably the biggest question here is whether the application > identifier is platform-independent (i.e. every app picks > org.example.MyApp), or we suggest people use a uuid on windows, etc. > > LaunchServices provides a unique app ID to the app as part of argv. Its > pretty ugly, actually. I think there are also ways to ask LaunchServices for > your ID, but I have yet to see or understand the reason why the app itself > needs it. None of this answers the question of what the "application > identifier" really means. > > Ah, forget all that nonsense. Its all true, but its about a different kind of > identifier. > > OS X has a somewhat deeper problem in that if LaunchServices isn't used to > start your app, then it has no identifier as far as the rest of the system is > concerned. Ie. if you have some way to start if from a command line, that > instance doesn't exist as far as the desktop is concerned. One might consider > this a non-issue, and I'm not necessarily going to disagree with you. > >
Unless you call "open" on it on the command line, which does fire up LaunchServices. But even if you just execute the binary from the command line, it will still have an icon in the dock and you can switch to it with Command-tab. Regards, John Ralls _______________________________________________ gtk-devel-list mailing list gtk-devel-list@gnome.org http://mail.gnome.org/mailman/listinfo/gtk-devel-list
