Thursday, 2020-08-20

malbaloon: looking at the journal it seems something is quite wrong, it doesn't seem to have /system nor /vendor mounted, wondering if it found the correct fstab during build (fstab is not used in sailfish but instead build generates systemd mount services based on it)00:25
malalso other things like missing .rc files from /00:27
malwhich android base version did you use? meaning which hybris branch during repo init00:28
T42琴双梅 %lastname% was added by: 琴双梅 %lastname%07:27
T42<kabouik> Thanks @edp_17 for the detailed report. For me @elros34's qt hack works perfectly on Pro¹: coordinates are updated dynamically depending on the orientation of the device, with no need to relaunch qxcompositor and in both `-o portrait` and `-o landscape`. My only issue is with the cursor coordinates not updating live unless I hold clic07:53
T42<kabouik> What is surprising is even if `xev` does not see the coordinates changing live when I don't hold a button, the system still knows how much the mouse has moved since the last event. I would expect that if I don't hold click and move the mouse, the cursor will still be at the same position when I update it one second later by clicking,07:53
T42since xev does not report coordinate changes until that point. And yet the cursor updates and jumps to its new position as if coordinates were correctly monitored all along; the only difference is it jumps instead of moves progressively.07:53
T42<edp_17> @kabouik: Interesting that the coordinates not updating live. Can it be because you installed some other stuff into your container? You can double-check that if you try this in a new desktop installation.08:51
T42<kabouik> I don't think so, I understood why @elros34 said earlier that my lightdm/xephyr setup was irrelevant: you can test the cursor even without launching the container08:52
T42<kabouik> Just starting qxcompositor without anything inside it, you can see the cursor, since it depends just on qxcompositor and not the rootfs08:53
T42<kabouik> So if I have something that breaks the coordinates update, it's most likely on the SFOS side, not the LXC rootfs08:53
T42<edp_17> @elros34: What I spotted is that when I use your qxcompositor, it drains the battery rapidly. Even if the mouse is not connected to the device. Is it because it is constantly checking the cursor coordinates?08:53
T42<kabouik> I didn't use it in real conditions yet, just for testing, so I cannot test. Mine may not be checking coordinates continuously anyway. :>08:55
T42<kabouik> But I did observe a higher battery use on my phone since a week or two, which I couldn't even firmly correlate to LXC. I need to check if my battery max capacity suddenly dropped.08:55
T42<edp_17> @kabouik: Yeah it does make sense. I remember that I had the cursor when the screen was showing the 'No Xwayland window' screen.08:56
T42<kabouik> Yup, this is something I didn't understand initially when Elros was saying reverting the lightdm/xephyr setup on my test device was unnecessary08:57
T42<edp_17> @kabouik: When I started to use the lxc and left the container running with a desktop in a window, the battery life way ok. But since I use the qxcompositor, the battery goes flat over a night.08:57
T42<edp_17> Or, there can be another reason for the poor battery life! I installed conky and set up to refresh in every 2 seconds. I switch this off and re-test.08:59
T42<elros34> @kabouik please check @sailfishos irc when you will have time08:59
T42<kabouik> Oh thanks @elros34, I didn't see my mentions there, thought they were duplicates of my Telegram mentions09:00
T42<edp_17> By the way, the clock shows 1 hour less in the container (in the device=10am, in the container=9am) Is there a workaround to correct it?09:02
T42<kabouik> Probably some timectl magic, it works for me but I may have fixed it this way already months ago, I don't remember09:02
T42<kabouik> Not an LXC specific issue in any case, check Debian resources09:03
T42<edp_17> I know this is not related to lxc, I was just wondering whether you had a similar issue. I tried to set the time in command line within the container but it didn't let me.09:14
T42<kabouik> I didn't mean it in a bad way, sorry if that's what it looked like. I may have had this issue but really don't remember clearly because if I did, that was months ago and I'm not sure how I fixed it :\09:17
T42<edp_17> Don't worry, I didn't take it in a bad way. 😊 I'll keep trying to fix that but it is not a big issue so it is not urgent at all.09:26
T42<kabouik> Which DE/WM are you using?09:26
T42<edp_17> xfce410:09
T42<edp_17> @elros: In OBS, I have seen on other projects using "geoclue-providers-mlsdb". What is this? Do I need it for my i9100?11:18
T42<elros34> For sure you have it already installed form jolla repo: zypper se -s "pkg name"11:22
T42<elros34> not sure whether this is relevant anymore because of mozilla location services  issue11:25
T42<edp_17> Yeah, you are right. That package is installed on the device but as you said from the jolla repo not from my local one. Thanks. 👍11:37
T42<edp_17> My other question is that all packages are built in my i9100 repo except the droid-hal-version-i9100. The error message is: "nothing provides pulseaudio-modules-droid, nothing provides hybris-libsensorfw-qt5, nothing provides qt5-qpa-hwcomposer-plugin, nothing provides qtscenegraph-adaptation, nothing provides pulseaudio-modules-droid11:39
T42>= 11.1.66 needed by droid-config-i9100-pulseaudio-settings, have choice for i9100-bluez-configs needed by droid-config-i9100: droid-config-i9100-bluez5 droid-config-i9100-bluez4"11:39
T42<edp_17> How can I resolve this?11:39
T42<neochapay> Don't build bluez4 package11:40
T42<elros34> you mean obs? show url11:41
T42<edp_17> Yep, obs. Here:
T42<edp_17> I mean, here:
T42<elros34> so error is quiet clear: you are missing all this packages in obs11:45
T42<elros34> you have type in project config or it's some font rendering issue which had on obs. It should be "droid-config-i9100-bluez4" not "droid-config-i9100 -bluez4"11:46
T42<edp_17> So I need to create them. But why these packages are not in other devices' repos like hammerhead:
T42<elros34> the are: expand list11:47
T42<edp_17> 😳 Oh, It displayed only the first 10 packages. Bummer. Thanks.11:50
T42<edp_17> That helped, thanks. Now the only error is: "have choice for i9100-bluez-configs needed by droid-config-i9100: droid-config-i9100-bluez5 droid-config-i9100-bluez4"12:20
T42<elros34> read what I wrote earlier12:20
T42<edp_17> I don't know what rpm or spec to upload into what package. On the pc I have thses:
T42<edp_17> Oh, okay, I missed that!12:21
T42<edp_17> I have fixed that, thanks!12:28
T42<edp_17> And it failed:
T42<elros34> try to remove that /boot workaround from droid-hal. It's not needed anymore on obs12:29
T42<edp_17> I have removed that hack from rpm/dhd/ and I pushed the droid-hal-i9100 into github but in there the dhd is pointing to a non existing url:
T42<edp_17> How can I fix this and how can I trigger the OBS to rebuild the failed package?12:41
T42<elros34> maybe something like that: cd dhd; git checkout master12:44
T42<elros34> now you are on some not existing 4d60d90 commit12:44
T42<edp_17> It says: Already on 'master'12:45
T42<elros34> did you commit some changes in dhd?12:45
T42<elros34> submodule12:45
T42<edp_17> I think accidentally, yes I did.12:45
T42<elros34> try git reset --hard in dhd submodule12:48
T42<edp_17> It says: HEAD is now at 4d60d90 /boot hack is removed for OBS12:49
T42<edp_17> This is the last change that I committed.12:50
T42<edp_17> I have this:
T42<elros34> I don't know exact steps: either remove that commit and then git fetch; git pull  or remove dhd directory and clone it again12:52
T42<edp_17> I went back to the commit prior my one (2f853a69ce6f9a0f230154a9f94c30a2feb6aff7). Now the dhd is in this stage:
T42<edp_17> Should I push the droid-hal-i9100 again?12:55
T42<edp_17> Yeah, I pushed and it seems ok in github now.12:56
T42<edp_17> Now, how can I rebuild the failed package in OBS?12:57
T42<edp_17> Well, at least re-try. 😄12:57
T42<edp_17> Okay, I edited the _service file and saved it. This triggered the build. 👍12:58
T42<elros34> you dont rebuild droid-hal. build it locally and upload droid-hal rpms like in faq12:58
T42<edp_17> I meant all of those for, rebuilding droid-hal-version-i9100 in obs.12:59
T42<edp_17> Yeah, it failed again. I guess because I need to rebuild the local packages and upload into the obs, then the  droid-hal-version-i9100 will complete fine.13:00
T42<elros34> droid-hal-version should rebuild itself if any of it's dependencies (like droid-hal) change13:01
T42<edp_17> Okay. That /book hack was necessary for the image. If I need to create a new image, should I add that back into the rpm/dhd/ or this is not required anymore?13:05
T42<elros34> now it's only needed to chmod 755 /boot on fresh target os forget about it13:07
T42<elros34> so*13:07
T42<edp_17> Thanks. Is it enough to rebuild droid-hal rpms or I need to build droidmedia, gst-droid, and audioflingerglue too?13:08
T42<elros34> obviously only droid-hal13:08
T42<edp_17> Okay. The build has failed: and I needed to apply this
T42<elros34> looks like it's already merged13:18
T42<edp_17> I probably have older submodules. No worries, now the droid-hal rpms, are rebuilt. I upload them to OBS and will see how the droid-hal-version behaves. Thanks for your help.13:19
T42<edp_17> All 15 packages now are succeeded. 😄13:36
T42<elros34> now u need patterns13:37
T42<edp_17> Yeah.13:38
T42<edp_17> lbt: Can you "patternise" my project (nemo:devel:hw:samsung:i9100), please? Thanks!13:39
kaluber0kk3rz: can you make public for me?15:10
T42<adampigg> @edp_17 [lbt: Can you "patternise" my project (nemo:dev …], dont wait for patternising, switch to meta-packages15:46
rinigus@adampigg: one addition to your idea of a github project with ports support:
rinigusas usual, for some reason, there is no interest in merging it upstream17:01
riniguswhich reminds me, compass support was also improved and sent upstream. but those sensors PRs are under review...17:02
T42<adampigg> rinigius: i heard they may be doing that upstream.....17:05
T42<adampigg> anyway, i am heading to a hotel now, vacation time, catch you later!17:05
rinigus@adampigg: enjoy!17:06
T42<edp_17> @adampigg : Thanks for the advice! I'll try that. Have a good vacation!17:10
*** olesalscheider_ is now known as olesalscheider19:23
T42<edp_17> @ankaos [/system/bin/logcat -b radio output https://del …], Sorry but I cannot help with this because I am not good with radio. 😭 I would need some help to fîx issues with the radio on my port too.19:38
T42<ankaos> Call fix. But call sound not fix19:39

Generated by 2.17.1 by Marius Gedminas - find it at!