T4 | <abhishek_0> gst-droid fails https://build.merproject.org/package/live_build_log/nemo:testing:hw:wingtech:wt88047/gst-droid/sailfishos_2.2.1.18/armv8el | 05:33 |
---|---|---|
*** ChanServ sets mode: +v T4 | 05:45 | |
entil | heya! | 06:18 |
entil | I think my fp2 broke down again, and I might as well email support, but I'm sure someone here knows if there's another way of debugging the sim modems than csd? | 06:19 |
entil | the phone rebooted itself and didn't as for a pin code, and there's nothing related to mobile connectivity anywhere, and csd says 0 modems found | 06:19 |
T4 | <abhishek_0> any idea about this anyone ? https://build.merproject.org/package/live_build_log/nemo:testing:hw:wingtech:wt88047/gst-droid/sailfishos_2.2.1.18/armv8el | 10:17 |
mal | @abhishek_0 https://build.merproject.org/project/meta/nemo:testing:hw:wingtech:wt88047 you have incorrect version of common repo as dependency | 10:18 |
mal | in the 2.2.1.18 target | 10:19 |
T4 | <abhishek_0> ah! thanks | 10:20 |
T4 | <abhishek_0> fixed | 10:20 |
mal | @abhishek_0 also you should usually disable build of the old release targets so those are not broken by updated packages | 10:22 |
T4 | <abhishek_0> okay, doing it | 10:22 |
mal | I always disable old target after I have released the version to public, see here https://build.merproject.org/project/show/nemo:testing:hw:fairphone:fp2-sibon | 10:23 |
T4 | <abhishek_0> done | 10:24 |
T4 | <abhishek_0> just pushed a new tag gitllab ci | 10:34 |
mal | I just updated my fp2 to 2.2.1.18 and seems to work fine | 10:36 |
wolfquin | hello | 10:46 |
T4 | <abhishek_0> r0kk3rz ping , you need to check the ci script | 10:50 |
r0kk3rz | do i? | 10:53 |
r0kk3rz | i expect its just the docker image being out of date | 10:53 |
T4 | <abhishek_0> yeah .. that, i was wondering if you have the dockerfile somewhere posted | 10:54 |
r0kk3rz | never had a dockerfile | 10:55 |
T4 | <abhishek_0> so you built the image yourself ? | 10:56 |
r0kk3rz | yeah all i did was tell docker to unpack the sdk tarball into a container | 10:59 |
T4 | <abhishek_0> i think it would be better to release docker images with release numbers | 11:04 |
r0kk3rz | since i did that theres a couple of people trying to package the sdk properly, so we should use those | 11:07 |
r0kk3rz | https://github.com/SfietKonstantin/docker-sailfishos-sdk | 11:07 |
T4 | <OhYash> is hybris-16.0 here yet? | 11:41 |
T4 | <Ken_Ookami_Horo> π€ | 11:42 |
r0kk3rz | nope | 11:43 |
mal | @OhYash no, not sure which device would have enough support so we could do the porting | 11:43 |
T4 | <Ken_Ookami_Horo> At first is there any hybris 15.1 build ? π | 11:43 |
r0kk3rz | thats been around for a while now | 11:44 |
T4 | <Ken_Ookami_Horo> π€ | 11:44 |
mal | which reminds me that rebasing should be done for 15.1 | 11:44 |
ohyash | r0kk3rz: ok. | 11:44 |
ohyash | mal: Yeah that too. And most Pie devices have a notch | 11:45 |
mal | what the hell is a notch, I have no idea what it is what it should do? | 11:45 |
T4 | <Ken_Ookami_Horo> And new Pie device may has worse restriction boot π | 11:46 |
ohyash | mal: That shit earpiece, camera are on the top mid of screen which iPhone X made famous. | 11:46 |
r0kk3rz | its a bit chunk out of the top of screen | 11:46 |
r0kk3rz | *big | 11:46 |
mal | so how does that affect usage? | 11:47 |
T4 | <Ken_Ookami_Horo> π€ | 11:47 |
ohyash | BTW, can android's project treble help non andorid OSes in anyway? | 11:47 |
r0kk3rz | it'll block the status bar | 11:47 |
r0kk3rz | because it still treats that area as actual screen | 11:47 |
mal | well that's a stupid idea | 11:48 |
ohyash | Its typicall useless. But gives more screen area to show off. | 11:48 |
r0kk3rz | it is a stupid idea | 11:48 |
r0kk3rz | but it seems stupid is trendy in the mobile world these days | 11:48 |
ohyash | On a side note, I'd love to see sailfish on some full-screen device like Vivo NEX | 11:48 |
T4 | <Ken_Ookami_Horo> So is there some device ported Sailfish OS successfully which has a notch ? | 11:48 |
T4 | <Ken_Ookami_Horo> π | 11:49 |
mal | how does android handle that, I thought android also has the top status bar and pull down menu | 11:49 |
r0kk3rz | i think Mister_Magister was working on a notched device | 11:49 |
r0kk3rz | mal: it reduces the status bar area to either side of the notch | 11:49 |
ohyash | mal: some workaround code that pushes stuff to the top edges | 11:49 |
mal | the first time I heard of a notch is after there were compains that the new iphone didn't have it or something, I didn't follow that much | 11:49 |
T4 | <Ken_Ookami_Horo> π | 11:50 |
r0kk3rz | but in general sfos will want lipstick patching | 11:50 |
r0kk3rz | and maybe not fullscreening apps | 11:50 |
ohyash | mal: the new iPhone had it. Now the yesterday's iPhone releases all the models have it. | 11:50 |
ohyash | yeah, SFOS has fullscreen devices | 11:50 |
ohyash | apps* | 11:50 |
ohyash | xD | 11:50 |
mal | ok, then it was some other device that didn't have it, but I wasn't interested enough to check what it actually is | 11:51 |
T4 | <samzn> Shouldn't be difficult | 11:51 |
mal | r0kk3rz: so in android app are no longer completely fullscreen if the device has a notch | 11:51 |
T4 | <samzn> Silica just has to implement header and footer | 11:51 |
mal | or was it so that top status bar is always visible on android | 11:52 |
kimmoli | just move the x0 to arbitrary value of notch depth | 11:52 |
mal | I don't remember how android looks like | 11:52 |
ohyash | top status bar is always visible on Android | 11:52 |
ohyash | and wow | 11:52 |
mal | kimmoli: yep, so basically just don't use x pixels from the top at all | 11:53 |
ohyash | except when you play games/videos. The notched status bar area is then blacked out | 11:53 |
r0kk3rz | mal: mal to illustrate - https://www.extremetech.com/wp-content/uploads/2018/03/zf5.jpg | 11:53 |
T4 | <Ken_Ookami_Horo> Emmmm | 11:53 |
r0kk3rz | the black area at the top of the screen is 'the notch' | 11:53 |
mal | r0kk3rz: so just as stupid as I thought | 11:54 |
r0kk3rz | indeed | 11:54 |
ohyash | I find Honor H8X Max's notch to be acceptable. This : https://cdn2.gsmarena.com/vv/pics/huawei/huawei-honor-8x-max-2.jpg | 11:54 |
kimmoli | that example has also fillet at corners | 11:54 |
mal | r0kk3rz: easiest way is to just limit the used display area in sailfish, that way we don't have to modify the status bar or anything like that | 11:54 |
mal | for some reason people have started to think that non-square display is a good idea | 11:55 |
mal | *non-rectangle | 11:55 |
kimmoli | https://image.shutterstock.com/image-photo/classic-vintage-retro-style-old-260nw-614643728.jpg | 11:55 |
T4 | <Ken_Ookami_Horo> Because App$e make it ?π | 11:55 |
mal | kimmoli: yep | 11:55 |
kimmoli | ohyash: with that small notch, pulldown menu says "Ab_ut" | 11:56 |
mal | and clock is not visible | 11:56 |
mal | or only partially visible | 11:56 |
ohyash | kimmoli: haha xD | 11:56 |
ohyash | and clock could have workaround of Hours and minutes on each side of notch. Notches themselves are workarounds | 11:57 |
mal | so when watching videos on such device the video is either cut from corners or only uses part of the display area | 11:58 |
ohyash | An official SFOS device with screen like this would be Puffect! : https://cdn2.gsmarena.com/vv/pics/vivo/vivo-nex-s2.jpg | 11:58 |
T4 | <Ken_Ookami_Horo> vivo ....... | 11:59 |
mal | I still find rounded corners of display area a bad idea | 11:59 |
ohyash | @Ken_Ookami_Horo, yes, can complain about that. | 11:59 |
T4 | <Ken_Ookami_Horo> π At first is it unlockable? | 12:00 |
ohyash | xD They boast about their in-screen fingerprint sensor for that | 12:00 |
mal | is it just me but I don't consider fingerprint a password, I think it's just a user id? | 12:02 |
T4 | <Ken_Ookami_Horo> It most probably proprietaryβ¦β¦ | 12:02 |
r0kk3rz | in-screen sounds like a really bad idea | 12:03 |
ohyash | DoubleTap FTW still | 12:04 |
ohyash | screw FP, and password even :D | 12:04 |
T4 | <Ken_Ookami_Horo> π | 12:04 |
mal | I don't use any password on my devices, well except sim pin code | 12:05 |
T4 | <meierrom> @Ken_Ookami_Horo [So is there some device ported Sailfish OS suc β¦], They are currently working on zenfone 5z, which comes with the notch. | 12:11 |
ohyash | sim pin code, woah | 12:11 |
T4 | <Ken_Ookami_Horo> π€ | 12:11 |
T4 | <meierrom> @mal [what the hell is a notch, I have no idea what β¦], There was a discussion on this just recently in the international sfos group. | 12:13 |
kimmoli | smells like telegram group... | 12:14 |
T4 | <Ken_Ookami_Horo> emmm | 12:14 |
T4 | <meierrom> @mal [so how does that affect usage?], The notch hides eg the date or shows it just partially. | 12:15 |
T4 | <meierrom> @mal [how does android handle that, I thought androi β¦], There is an additional setting in Android which adds a black bar at the top. The status bar then appears just below the notch in full. | 12:20 |
mal | so the easiest way in sailfish would be to also add such black bar at the top | 12:24 |
T4 | deathslayer123 was added by: deathslayer123 | 12:26 |
T4 | <meierrom> @kimmoli [smells like telegram group...], Yes, it is. | 12:29 |
T4 | <meierrom> @r0kk3rz [in-screen sounds like a really bad idea], Fp is even history in latest iPhones. They go face recognition and code as backup. | 12:31 |
r0kk3rz | gah telegram | 12:53 |
T4 | <abhishek_0> mal is there any way to debug gpu issues , performance is ery bad , i just played 1080p recorded video, it played at around 9 or 10 fps | 13:09 |
r0kk3rz | check aida64 | 13:12 |
mal | and check cpu usage in such case, and check that all media symlinks are present and so on | 13:18 |
wolfquin | hello | 13:38 |
T4 | <OhYash> @abhishek_0 [mal is there any way to debug gpu issues , per β¦], Ditto. YU Yureka in my case. | 13:38 |
wolfquin | i have an issue, is there a new command to enter chroot of ubuntu? i tried ubu-chroot but it says command not found, have i done a mistake? | 13:39 |
T4 | <abhishek_0> here is the logcat while playing the video https://paste.ubuntu.com/p/db2tVPn9kn/ | 13:41 |
T4 | <abhishek_0> cpu usage stays at 20%-30% while screen on | 13:42 |
elros1 | wolfquin: do you have android-tools-hadk installed as mentioned in hadk? | 13:43 |
wolfquin | OH! of course, i made a new void installation yesterday and forgot to install the tools, thought i had. | 13:43 |
wolfquin | thanks for pointing that out | 13:44 |
T4 | <abhishek_0> lipstick, sensorwrd, kworker processes take most cpu | 13:45 |
r0kk3rz | @abhishek_0: you're using gst-omx arent you? | 13:45 |
T4 | <abhishek_0> yeah | 13:45 |
r0kk3rz | then you probably want to look through the GST_DEBUG output | 13:47 |
T4 | <abhishek_0> gst-omx fixed camera for me , and few colors issue , due to broken gst-droid, idk what happened since then, | 13:47 |
T4 | <abhishek_0> give me sec | 13:47 |
r0kk3rz | which is weird, because the recording is done with droidmedia | 13:49 |
r0kk3rz | so if recording works... | 13:49 |
mal | still waitig information about the media symlinks | 13:52 |
T4 | <abhishek_0> r0kk3rz https://paste.ubuntu.com/p/B3Z5tCTdT6/ here is the gst-debug | 13:58 |
T4 | <abhishek_0> mal , media symlinks are fixed give me a sec | 13:58 |
T4 | <abhishek_0> here is the system/etc and /etc https://paste.ubuntu.com/p/xPqQH2PqXG/ | 14:00 |
mal | and the links are verified to be correct? | 14:06 |
T4 | <abhishek_0> yes | 14:12 |
T4 | <abhishek_0> here is my gst-omx.conf https://github.com/sailfishos-wt88047/droid-config-wt88047/blob/ebb983fe2844af33bf016ba8087f2d952c7a6603/sparse/etc/xdg/gstomx.conf | 14:12 |
r0kk3rz | why are you running that as root? | 14:34 |
T4 | <abhishek_0> r0kk3rz because root is the king :P ,, here https://paste.ubuntu.com/p/KXmfRyFrdt/ | 14:39 |
r0kk3rz | it seems to be dropping a lot of frames | 14:52 |
T4 | <abhishek_0> Exactly | 14:53 |
mal | @abhishek_0 that last logs looks bad, you have something really wrong if you get those permission denied messages | 15:13 |
T4 | <abhishek_0> mal afaik nothing at all, here are the device sources https://github.com/sailfishos-wt88047/ | 15:14 |
T4 | <abhishek_0> with all the commits | 15:15 |
mal | you have both gst-omx and gst-droid? | 15:16 |
T4 | <abhishek_0> yeah | 15:16 |
mal | hmm, or were both needed for gst-omx? | 15:16 |
T4 | <abhishek_0> you suggested to keep both | 15:16 |
mal | really? I have no memory of that | 15:18 |
T4 | <abhishek_0> you told me to keep gst-omx as long as it keeps my camera working | 15:20 |
mal | so is video performance a different if you use omx or gst-droid? | 15:22 |
mal | *any | 15:23 |
T4 | <abhishek_0> mal havent tried that , since gst-droid was broken 3 months ago when i was last active,, i should try it | 15:24 |
elros1 | hmm I guees it's not normal that I have flood of: UDEV [1720.555392] change /devices/platform/disp (platform) | 15:34 |
elros1 | when anything on screen change. I think I read about something similar but can't remember | 15:34 |
elros1 | looks like maguro had the same issue | 15:51 |
T4 | <adampigg> when anyone has upgraded to 2.2.1.18 using zypper ... were there more packages to downgrade than upgrade?? | 17:11 |
T4 | <adampigg> 387 up, 467 down here! | 17:11 |
mal | @adampigg I didn't notice anything like that, although I used version --dup | 17:13 |
r0kk3rz | downgrade? | 17:13 |
mal | @adampigg sometimes you can see downgrades because the version stayed the same but build number is smaller on the new release | 17:14 |
mal | build number always starts from 1 in new release target | 17:14 |
T4 | <adampigg> mal; that makes sense | 17:23 |
T4 | <adampigg> mal: i usually use zypper to know whats going on | 17:24 |
T4 | <adampigg> theyre functionally the same anyway arnt they? | 17:24 |
mal | quite the same I think | 17:24 |
T4 | <adampigg> balls, i need to free some space.... | 17:26 |
T4 | <abhishek_0> mofo OMX ... just removed that everything now just works fine | 17:31 |
mal | heh | 17:32 |
mal | so video playback is ok and so is camera? | 17:33 |
T4 | <abhishek_0> wait wait, just checked youtube , still have colour issues | 17:34 |
T4 | <abhishek_0> my phone has just gotten smoother, i dont want to go back to omx sighs | 17:36 |
elros1 | try to remove libI420colorconvert.so so gst-droid will use it's own implementation | 17:38 |
T4 | <adampigg> oh, thats something i can try too | 17:43 |
T4 | <adampigg> i guess you have the color problems in browser, but lls is ok? @abhishek_0 | 17:44 |
T4 | <abhishek_0> elros1 i dont have that lib under system/lib | 17:45 |
T4 | <abhishek_0> @adampigg [i guess you have the color problems in browser β¦], let me check this | 17:45 |
elros1 | so search it under vendor | 17:45 |
T4 | <abhishek_0> found it https://github.com/sailfishos-wt88047/proprietary_vendor_wingtech/blob/hybris-14.1/wt88047/proprietary/vendor/lib/libI420colorconvert.so | 17:50 |
*** ChanServ sets mode: +v T4 | 18:00 | |
T4 | <abhishek_0> @adampigg [i guess you have the color problems in browser β¦], just checked webcat works just fine, can you check the solution which elros provided ? since i am not in the position of setup-ing the build environment rn | 18:01 |
T4 | <adampigg> you dont need a build environment, just rename the file on the device ... i will when i have finished the update | 18:02 |
mal | @abhishek_0 remount vendor partition as readwrite and rename the file and remount again as readonly | 18:02 |
mal | @adampigg wondering if there would be some way to create a dummy lib which would cause the fallback to gst-droid method | 18:04 |
T4 | <abhishek_0> @adampigg [you dont need a build environment, just rename β¦], dude /vendor is a read only file system | 18:04 |
mal | @abhishek_0 mount -o remount,rw /vendor | 18:04 |
mal | oops | 18:04 |
mal | I think that should work | 18:05 |
T4 | <abhishek_0> no, i think since all the system files go into a system.img which gets mounted on the boot | 18:06 |
mal | you are mixing things | 18:07 |
mal | you rename the file whereever it happens to be | 18:07 |
elros1 | system is regular partition, just remount it with rw option | 18:08 |
T4 | <abhishek_0> are you guys sure https://paste.ubuntu.com/p/RMrNJhVK7c/ ? :P | 18:12 |
elros1 | yes, I am sure: df /vendor | 18:13 |
T4 | <abhishek_0> elros1 [root@Sailfish nemo]# df /vendor/ β¦ Filesystem 1K-blocks Used Available Use% Mounted on β¦ /dev/mmcblk0p23 1015704 584732 414588 59% /system | 18:14 |
elros1 | so if you didn't guess yet: mount -o remount,rw /system | 18:15 |
T4 | <abhishek_0> already did | 18:15 |
kimmoli | mount -o remount,rw /system | 18:15 |
T4 | <abhishek_0> lol | 18:15 |
mal | yep, vendor seems to be a symlink or something | 18:18 |
T4 | <adampigg> erm, where have all my android apps gone?! | 18:18 |
T4 | <abhishek_0> elros1 just mv that lib , now i have a black screen in while playing video in jolla browser | 18:19 |
elros1 | that is a progress :) | 18:20 |
T4 | <abhishek_0> whats the next step ? | 18:21 |
T4 | <abhishek_0> i should also add this to had-faq , if it is already not there | 18:21 |
T4 | <adampigg> surely jolla didnt ad something to lipstick to prevent ports using alien? | 18:23 |
T4 | <abhishek_0> wait what! you were using aliendalvik ? | 18:24 |
elros1 | next step is to figure out why gstdroidvdec fail | 18:24 |
T4 | <abhishek_0> elros1 , logcat ? gst_debug or anything else | 18:25 |
elros1 | probably gst debug: https://github.com/sailfishos/gst-droid/blob/8e19ceea625e0841e6db9e56f7dfe3d8f25936f0/gst/droidcodec/gstdroidvdec.c#L336 | 18:27 |
T4 | <abhishek_0> meanwhile i grabbed logcat https://paste.ubuntu.com/p/QFhjBGnMNW/ ,, let me get debug info on browser | 18:30 |
T4 | <adampigg> ok, drama over ... it just seems like all my apps were removed, and will need to reinstall :/ | 18:31 |
T4 | <abhishek_0> @adampigg [ok, drama over ... it just seems like all my a β¦], hey i might ping you ;) | 18:32 |
T4 | <adampigg> @abhishek_0 [hey i might ping you ;)], remember, alien is like fight club | 18:33 |
T4 | <abhishek_0> i know the first rule, i am a member | 18:33 |
T4 | <adampigg> I confirm the same problem with libI420colorconvert.so | 18:35 |
T4 | <adampigg> with the file there, there are color issues, and browser crashes a lot | 18:35 |
T4 | <adampigg> without it, browser doesnt crash, there is audio, but no video | 18:35 |
T4 | <abhishek_0> exactly | 18:36 |
T4 | <abhishek_0> here is the gst-debug of sailfish-browser https://paste.ubuntu.com/p/shkKQv7Gsh/ | 18:42 |
mal | it should be GST_DEBUG=droid*:5 | 18:42 |
T4 | <adampigg> maybe get a log with the library also | 18:44 |
T4 | <abhishek_0> here https://paste.ubuntu.com/p/4bw2hFZ5mS/ | 18:46 |
T4 | <abhishek_0> without library | 18:46 |
T4 | <adampigg> mal, still no compass joy | 18:46 |
mal | @adampigg wondering why only your device has issue | 18:48 |
T4 | <abhishek_0> compass is fine | 18:48 |
T4 | <abhishek_0> with library here https://paste.ubuntu.com/p/Zpr7tq76wV/ | 18:51 |
T4 | <abhishek_0> it segfaults | 18:51 |
mal | without the lib decoding seems to work but not sure why nothing is shown on display | 18:58 |
T4 | <abhishek_0> "error: Unknown codec colour format: 21" | 18:59 |
T4 | <abhishek_0> maybe | 18:59 |
elros1 | looks like it's OMX_COLOR_FormatYUV420SemiPlanar | 19:05 |
piggz | mal: where would compass debugging start? logcat? test_sensors? | 20:24 |
piggz | i guess: | 20:25 |
piggz | 09-15 20:25:13.212 2817 2817 E libsensor1: client_get_conn: qmi_client_get_service_list error -2 | 20:25 |
piggz | 09-15 20:25:13.212 2817 2817 E libsensor1: sensor1_write: Unable to create client connection | 20:25 |
piggz | 09-15 20:25:13.212 2817 2817 E qti_sensors_hal: AkmAlgo,enable: sensor1_write() error: 12 | 20:25 |
piggz | 09-15 20:25:13.212 2817 2817 E qti_sensors_hal: enable:akm_service->enable error:5 | 20:25 |
piggz | 09-15 20:25:13.212 2817 2817 E qti_sensors_hal: Activate the handle 54 is not successful | 20:25 |
T4 | deathslayer123 was removed by: deathslayer123 | 20:40 |
mal | piggz: not sure if that is bad | 20:41 |
piggz | mal: its the only thing that appeears in logcat when opening the compass app .... | 20:43 |
piggz | mal: does the compass sensor rely on any android services? | 20:45 |
mal | I think only the vendor sensor blobs should be enough | 20:47 |
mal | piggz: just to be sure, you haven't enabled the serviceservice manually? | 20:47 |
piggz | mal: ? | 20:48 |
mal | piggz: what sensor related processes do you have running | 20:49 |
piggz | mal: https://bpaste.net/show/c8d811b868f6 | 20:50 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!