| Mister_Magister | man I changed nothing and now libhybris fails to build | 14:39 |
|---|---|---|
| Mister_Magister | what the heck | 14:40 |
| Mister_Magister | maaan now droid-hal doesn't want to build for god unknown reason | 17:22 |
| Mister_Magister | I come back to a device and what do I see? A patch isn't applied… why? I had this device built and everything how come? | 17:32 |
| Mister_Magister | all patches are not applied what the hell | 17:35 |
| Mister_Magister | i just ran grep and dhd building issue changed | 17:39 |
| mal | you ran repo sync again? | 17:45 |
| Mister_Magister | i'm not sure, i don't think so, its android 14 so maybe it didn't need patches? | 17:47 |
| Mister_Magister | mal: since you're here, which dhd commit should i have? master? | 17:47 |
| Mister_Magister | i don't remember much neither does my terminal history apparently | 17:49 |
| Mister_Magister | screw it i'll apply patches and see | 17:50 |
| Mister_Magister | but yeah seems like patches will resolve all my issues so i'll go with it | 17:55 |
| Mister_Magister | mal: did you see my report about tablet? | 17:55 |
| Mister_Magister | i'm just using evdevtablet plugin in the app directly and even made an app that uses them | 18:01 |
| Mister_Magister | too much work to implement it otherwise | 18:01 |
| Mister_Magister | probably busy :P | 18:03 |
| Mister_Magister | whaddyaknow everything built | 18:15 |
| mal | dhd can be latest, droid-config not | 18:25 |
| Mister_Magister | oki | 18:25 |
| mal | @NotKit I had a look at the ubports libhybris patches and wondered which issue this fixes https://gitlab.com/ubports/development/core/packaging/libhybris/-/blob/ubports/latest/debian/patches/0007-hooks-use-o-linker-on-Android-9.patch?ref_type=heads also decided to take couple of the other patches to be included in upstream | 20:14 |
| mal | also the wondering about that vndk env patch | 20:16 |
| T42 | <NotKit> regarding debian/patches/0007-hooks-use-o-linker-on-Android-9.patch, I'm not even sure at this point. It was a hotfix as some Android 9 devices broke with newer Q linker, possibly due to linker paths | 20:19 |
| T42 | <NotKit> without that change it will attempt to use Q linker for Pie, though historically Pie used Oreo linker in libhybris | 20:20 |
| Mister_Magister | mal: if you wanna hear funny bug, if I connect charger, almost all touch input is considered tablet input | 20:27 |
| Mister_Magister | >does it basically disable touch? yes | 20:28 |
| Mister_Magister | now all the bugs left are the ones to be fixed by… well probably mal | 20:34 |
| Mister_Magister | that being the weird suspend bug, animations being 2x speed, camera2 api video recording | 20:36 |
| mal | Mister_Magister: video recording can be made to work in a12+ without camera2 | 21:13 |
| mal | if you update droidmedia | 21:13 |
| mal | @NotKit any idea in which kind of situation is this needed https://gitlab.com/ubports/development/core/packaging/libhybris/-/blob/ubports/latest/debian/patches/0005-q-allow-to-set-VNDK-version-with-HYBRIS_VNDK_VERSION.patch?ref_type=heads | 21:16 |
| mal | Mister_Magister: I forgot which suspend bug you are talking about? the one about usb needed to be connected once? | 21:17 |
| Mister_Magister | yeye | 21:17 |
| Mister_Magister | mal: oh… but well, only aunts record with a tablet so i'm not in hurry xd | 21:18 |
| mal | Mister_Magister: wich android base are you talking about regarding video recording? | 21:18 |
| Mister_Magister | all 3 are 14 base | 21:18 |
| Mister_Magister | all other devices are quite old, this one is as new as we could muster xd | 21:19 |
| mal | ok, so if you just update droidmedia then video recording should work | 21:19 |
| Mister_Magister | hm okay | 21:19 |
| Mister_Magister | i'll try that tomorrow thanks | 21:19 |
| mal | in case you want to see what the fix was https://github.com/sailfishos/droidmedia/pull/131 | 21:20 |
| mal | so just update the latest master branch and things should be fine | 21:21 |
| T42 | <b100dian> mal: is playback in gstreamer for FP5 - where you said there were missing timestamps - in the works in droidmedia too? | 21:21 |
| mal | hmm | 21:22 |
| mal | not sure what you mean, there is timestamp issue in camera2 but there is some codec bug in fp5 which is why i disabled 264 hw codec for now | 21:23 |
| mal | *h264 hw codec | 21:23 |
| mal | which is interesting that it seems to be only that codec | 21:24 |
| T42 | <b100dian> mal: I mean the playback issues that are shared with sake, nagara and fp5 in gallery - which are independent of the playback issues from gecko-camera. I thought you diagnosed them as missing timestamps, was curious if you went further | 21:37 |
| mal | and it was only h264 also on those other devices? | 21:39 |
| T42 | <NotKit> mal: HYBRIS_VNDK_VERSION is probably artifact from something earlier. I can't find it being used at all when searching at UBPorts GitLab org | 21:39 |
| Mister_Magister | mal: thanks, any comment on the tablet stuff? I wonder if I got things right or not | 21:39 |
| T42 | <NotKit> should be dropped | 21:39 |
| mal | Mister_Magister: quite odd that it changes input type in that situation | 21:40 |
| Mister_Magister | i didn't mean that | 21:41 |
| Mister_Magister | i meant what changes are needed to get it to work, the huge block of text i dumped like yesterday | 21:41 |
| mal | @NotKit ok, I'll test how lh works with those apex path changes and if something else is needed also | 21:41 |
| mal | I think those won't help with the minimediaservice issue, that might need changes to linkerconfig | 21:41 |
| mal | or some path env var in the .rc file if possible | 21:42 |
| Mister_Magister | aw man I broke audio | 21:45 |
| T42 | <b100dian> mal: re h265 - if you mean video/avc=0 disabled in gstreamer conf then yes | 21:54 |
| Mister_Magister | avc is h264 | 21:55 |
| T42 | <b100dian> yeah I meant 264, was a typo. | 21:56 |
| * Mister_Magister fixing audio so that i can go to sleep watching videos | 21:57 | |
| mal | @b100dian I think the issue was not missing timestamps but out of order timestamps | 22:45 |
| mal | gstreamer doesn't like if frames come in wrong order from codec | 22:45 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!