| mal | what do you need to know? | 01:18 |
|---|---|---|
| Mister_Magister | mal: nothing anymore sorry to bother but i was inch away about shooting you message "so mal… how good is your scenegraph knowledge" xd | 11:32 |
| Mister_Magister | mal: but different question, why does android leak memory so much? When i valgrind my app there's seemingly no memory leaks in my app and all of them come from sfos, mostly from android. Like rotating screen twice gives me 0.2MB more ram usage | 11:33 |
| *** rainemak__ is now known as rainemak | 11:50 | |
| mal | Mister_Magister: which android base, what kind of leaks do those appear to be based on valgrind, it is possible some issues could be because of libhybris and it might need fixing | 11:58 |
| Mister_Magister | mal: many bases like my phone is on 18.1 i think, even official phone has same thing | 11:59 |
| Mister_Magister | mal: i can run valgrind and send u the file | 11:59 |
| mal | I did notice some spam in valgrind while updating sdl to latest version but that happened only during start of the app not during rotation, maybe qt is different | 12:00 |
| Mister_Magister | I'm doing also bunch of opengl/scenegraph stuff so yeah more custom usecase | 12:50 |
| Mister_Magister | probably some edgecases | 12:50 |
| Mister_Magister | i'm always doing something custom aren't i | 14:18 |
| mornfall | i have an odd issue with appsupport… mpris-proxy seems to work fine, events show up in dbus, and next/prev/pause work in an android player, but not play (except in some rare circumstances, like right after a reboot, when it works) | 21:05 |
| mornfall | (though perhaps i am in the wrong place to ask about this?) | 21:05 |
| mornfall | and on an unrelated note, there's recently been some talk about jolla/sfos coming back from the dead once more… any more details on that? :-) | 21:08 |
| mal | mornfall: this has some info about that https://forum.sailfishos.org/t/sailfish-community-news-30th-november-jolla-reborn/17473 | 21:12 |
| mal | mornfall: how do you try to control pause/play? | 21:13 |
| mornfall | bluetooth headphones | 21:14 |
| mornfall | but the bluetooth part presumably works, given that the event shows up on dbus | 21:14 |
| mornfall | the player in question does show up on the lock screen and the play button there works, interestingly enough | 21:16 |
| mornfall | the lock screen button triggers this: destination=org.mpris.MediaPlayer2.alien.com.spotify.music serial=3046 path=/org/mpris/MediaPlayer2; interface=org.mpris.MediaPlayer2.Player; member=PlayPause | 21:18 |
| mornfall | while mpris (bluetooth headphones) trigger: destination=:1.157 serial=36 path=/org/mpris/MediaPlayer2; interface=org.mpris.MediaPlayer2.Player; member=Play | 21:19 |
| mal | mornfall: ok, do the headphones have just single button for pause/play and separate ones for previous and next? | 21:30 |
| mornfall | interesting, so PlayPause works (starts playback), but not Play, with otherwise identical message | 21:30 |
| mornfall | mal: there's a single play/pause/accept call control | 21:31 |
| mornfall | and yes, prev/next are separate | 21:31 |
| mornfall | guess if i can convince mpris-proxy to send PlayPause instead of Play on that control, it'd work | 21:31 |
| mal | ok, I have had some issues with that also sometimes, not sure what was causing those, last time I tested it worked (but I used development build in my test, I should test on a 4.5.0 device also) | 21:32 |
| mornfall | oh! it perhaps doesn't actually use mpris-proxy… it's AVRCP | 21:32 |
| mornfall | no, it does use mpris-proxy… no events if i stop it | 21:33 |
| attah | Is it a general thing or an app-specific thing? For Sailfish apps i did at least have to handle PlayPause explicitly | 21:36 |
| attah | It wouldn't be impossible that some one app here or there has gotten it wrong, especially if other commands are the norm | 21:36 |
| mornfall | well this is reverse – both Pause and PlayPause work ok, but not Play; the app in question is appsupport (alien), which makes things more complicated | 21:37 |
| attah | well... not bothered to hook up the specific ones if the combined is all they ever see natively | 21:38 |
| attah | appsupport could be mapping them straight through is what i'm suggesting | 21:39 |
| mornfall | the funny part is that it _sometimes_ works, but i don't know if that's because mpris-proxy sometimes sends PlayPause, or because appsupport sometimes recognizes Play | 21:39 |
| attah | very strange... one would have assumed the whole thing to work intermittently, not just certain commands | 21:40 |
| attah | maybe if the playback status doesn't show something sensible, there is a fallback of PlayPause? | 21:40 |
| mornfall | restarting appsupport makes Play work (i am just running dbus-send at this point) | 21:43 |
| mal | ok, so maybe there is something that gets out of sync in appsupport | 21:44 |
| mornfall | so i have tracked it down to closing and reopening spotify without stopping appsupport (that causes Play to break, though everything else keeps working on the new instance) | 21:50 |
| mal | hmm, interesting, I need to test that tomorrow | 21:51 |
| mornfall | mal: thanks; i'll stick around, feel free to ping me | 22:06 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!