On Tue, 2019-04-02 at 08:00 -0400, Greg Troxel wrote: > > It's also free of google code. If you didn't add in microg, or > gapps, > then there will be no GCM/FCM support.
I did add gapps. > Are you really sure there is FCM push going on in this case? There is definitely an FCM push. Co-incidentally, I just filed https://github.com/BelledonneCommunications/linphone-android/issues/542 this morning which might be related. I am entirely open to the possibility that my FCM API requests don't have the right data in them to trigger the app to wake up, even from a "quit" if it's not already running. But I have also not seen anywhere documented what needs to be in an FCM API request, if there is specific needs. > As I understand it "background mode" is telling linphone to keep the > service running when the screen is off. My understanding is that with push, nothing should need to be kept running because Android with FCM will start Linphone if it's not already running. > I would expect that to either > keep tthe service or register for push. That seems reasonable. More or less what I said above I think. > I would expect it not to > register for push without that, as it's a request from the user to > get > calls when not running, and by implication a request not to be > registered when not running. But then there is also a separate per account setting of "Allow push notification". All in all it feels like there are too many options conflicting or overlapping each other. > Without background mode, linphone stops on exit for me, which is what > I > expect. Even with Background mode and Enable service notification, if I Quit linphone, it exits "hard" from everything I can tell. Even the service notification icon goes away, and it will not answer a call if one is made, even when I have push enabled for the account. But this goes back to perhaps my push "data" is not correct. > I have read somewwhere that there are issues with selecting and > deselecting push in that there are bugs in the linphone code. But > I'm > not clear on that. Interesting. > Enable service notification should be more robust against killing. But with push, I'm not supposed to even care about killing because the push should wake linphone up. That's my understanding at least. Cheers, b.
signature.asc
Description: This is a digitally signed message part
_______________________________________________ Linphone-users mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/linphone-users
