Saturday, 2019-11-23

deathmistmorning! kimmoli: spiiroin: any clue about my gesture detection MCE signalling behavior?
deathmistthe signal I'm talking about is (e.g. I get it for "power-key-menu" but not any gestures) like supposed to with the dbus connection details at
rinigushow can I setup CI builds for tama? r0kk3rz , my account name is rinigus on gitlab. would you mind to make tama project and let me manage it?09:08
T42<程彩磊 %lastname%> (Photo, 539x622) 😍please contact有需要请联系:  @angyey09:27
spiiroindeathmist: can you pastebin output from "evdev_trace -I" and "mcetool --status"; also which gesture is the one that is acting up? (nb you can configure 20 gestures, but have only 10 dbus actions to use for those... did not cross my mind when making the latest changes)09:44
deathmistspiiroin: "evdev_trace -I" "mcetool --status" I still didn't try latest MCE since I only care about same gestures as kimmoli has had for years. all gestures over id 10 seem perform same actions as double tap currently, which "works", but also unlocks instead of just unblanking :/09:51
deathmistothers don't seem to do anything and none vibrate when detected as I specify in the config for example09:54
deathmistfor now I enable all the gestures manually with "for g in $(ls -1 /proc/touchpanel/*_enable | egrep -v "sleep|tp"); do echo 1 > $g; done" till I change the gestured implementation10:05
spiiroindeathmist: what can be seen in mcetool output: gestures 5-10 end up being "do nothing" - does not match that config file (?) of yours, can you elaborate on what you've done with it?10:18
spiiroinoh, there was a title to it: so it is installed as /etc/mce/90-something.conf?10:19
deathmistspiiroin: currently
deathmistgestures 11-17 are unimportant at best but removing them from the file entirely doesn't help either (they still register as unlock actions too?!)10:23
spiiroindeathmist: it is by design that anything that can't be configured is treated as doubletap -> for your mce version gestures > 1010:24
deathmisthm interesting, but it also doesn't remap my doubletap gesture (4) action to "vibrate,unblank,dbus4" and instead is "unblank,tkunlock"10:25
spiiroindeathmist: well, these are just the defaults at "priority 90" - if you have file like 91-foo.conf that does something to doubletap / have manually changed doubletap config -> it will override what you have in that 90 file10:27
spiiroindeathmist: try: mcetool --reset-settings=/system/osso/dsm/powerkey10:29
spiiroinit should clear out any runtime/dynamic changes you might have10:29
deathmistspiiroin: ah, there we go!! thank you, but my config file is shown as the last one when running "ls -1 /etc/mce/"10:30
deathmistguess it'll work when placed there on the rootfs on a fresh install10:30
spiiroinalso noticed: if your defaults have "bogus" data - like those "dbus13" actions -> the defaults vs delta to defaults handling works in bit odd ways10:31
deathmistI removed them for now from the config file, could add them back anytime later maybe when 3.2.1 releases perhaps(?)10:32
spiirointhe dbusNN with NN > 10 are not going to be in 3.2.1 either ...10:32
spiiroinas I was trying to say earlier: increased the number of gestures you can configure, but forgot to increase the number of dbus actions that you can use for that purpose10:33
deathmistI'm fine with it. the gestures kimmoli had too are fine for me since the others are just "extras" imo and especially lines don't detect very accurately :p  I'll continue working on further when it's possible10:35
T42<singhnsk> Hi, can anybody helo me with these errors:11:53
T42<singhnsk> `[426/426] including vendor/qcom/opensource/dataservices/ ...11:53
T42<singhnsk> art/libart_fake/ error: "art-gtest-AbstractMethod (JAVA_LIBRARIES android-arm64) missing apache-xml (JAVA_LIBRARIES android-arm64)"11:53
T42<singhnsk> You can set ALLOW_MISSING_DEPENDENCIES=true in your environment if this is intentional, but that may defer real problems until later in the build.11:53
T42<singhnsk> art/libart_fake/ error: "art-gtest-AllFields (JAVA_LIBRARIES android-arm64) missing apache-xml (JAVA_LIBRARIES android-arm64)"11:53
T42<singhnsk> You can set ALLOW_MISSING_DEPENDENCIES=true in your environment if this is intentional, but that may defer real problems until later in the build.11:53
T42<singhnsk> art/libart_fake/ error: "art-gtest-DefaultMethods (JAVA_LIBRARIES android-arm64) missing apache-xml (JAVA_LIBRARIES android-arm64)"11:53
T42<singhnsk> You can set ALLOW_MISSING_DEPENDENCIES=true in your environment if this is intentional, but that may defer real problems until later in the build.`11:53
T42<singhnsk> I have tried including export `ALLOW_MISSING_DEPENDENCIES=true` but the outcome is not changing.11:56
deathmist@singhnsk please post more than 3 lines always in a pastebin including the command line you ran. I can't help if you don't give any info. what base? what device? what does your device manifest look like?11:57
T42<singhnsk> @deathmist [@singhnsk please post more than 3 lines always …], Sorry, I am new. I will push the full to a pastebin and share again. The errors are quite long and I am sure there's something wrong with the files. I just do not know what it is and hence have no clue on the next step to take.12:13
T42<singhnsk> I am trying to builkd for a Nokia 7.2 while making use of the code used for the Unofficial LOS 16.12:13
T42<singhnsk> I am following the porting documentation as provided by Sailfish. And I am doing this as a hobby. Just would like to reach some outcome before I give up :)) I do not want anbiody to give up his precious time, but I will appreciate some assistance.12:13
T42<singhnsk> Paste is here:
deathmist@singhnsk =>
T42<kisekinopureya> @singhnsk please read pinned post, tag instead of replying12:14
deathmist16.0 is kind of rough in terms of documentation out there, it's mostly in logs currently and you also need to apply patches before building hybris-hal with "hybris-patches/ --mb"12:17
T42<singhnsk> Thank you, deathmist, I see some daylight now from the earlier dark stage.12:18
T42<singhnsk> And noted, @kisekinopureya. I will take care of those.12:18
T42<singhnsk> @deathmist Do I still need to copy the 3 files from the OnePlus 5 git to mine if I do not have a OP5?12:28
T42<singhnsk> And if yes, where do they go? I do not have a directory like this: `hybris/droid-configs/sparse/usr/libexec/droid-hybris/system/etc/init`. Should I create it?12:28
deathmistyou'll get to it later, you can ignore it for now until you get to HADK chapter 7.112:28
deathmistand yes you need the 3 files later, they provide selinux_stubs patched *servicemanager init files and disables some services that cause issues (or are useless) regardless of your device brand etc12:31
T42<singhnsk> Got that.12:41
T42<singhnsk> For now, it moved a step further and I have a new error. Not sure if its fine to keep asking, but can you give me a clue? It looks far less messy now 😁12:43
T42<singhnsk> Also, how does SFOS manage the A/B devices. Since it clearly mentioned there that it couldn't find boot partiton. My device uses A/B and I have it the proper path of boot_a and boot_b12:47
deathmist@singhnsk not sure about A/B, but people have gotten it to work. not sure about that libawk error, but it also seems like it's using a prebuilt kernel instead of the sources you cloned which is really bad as your changes to kernel source won't matter12:56
deathmistdo `grep -rnw $ANDROID_ROOT/device/ -e "TARGET_PREBUILT_KERNEL"` and comment / remove the line in your * file12:56
T42<adampigg> Atm, we are assuming installed on A13:02
Thaodanwe ignore the a/b stuff mostly it's Android specific13:03
Thaodanyou flash the kernel on the active boot partition or on both.13:03
T42<elros34> jusa: ping any chance for dbus-daemon-launch-helper location fix in 3.2.1.x or do I have to create PR about that?13:16
T42<singhnsk> @deathmist I think that's because kernel spruce wasn't released by Nokia. The porter used the prebuild kernel mostly.13:38
T42<singhnsk> Moreover, it seems I had some issues with my repo. I do not get that warning anymore since I pointed to the correct sdm_660 kernel. However, still stuck with same error.13:38
deathmistif you don't have kernel source, you sadly cannot continue further. you need to make defconfig changes to allow disabling SELinux and fix a bunch of major issues that prevent systemd from even starting during init13:38
deathmist here are some with a quick search, but you'd probably need one for LineageOS for it to not simply boot to fastboot or whatever immediately13:42
T42<singhnsk> @deathmist I think it's good to use the SDM_660 sources from Nokia 6.1 Plus. He used those to boot the LOS on the 7.2. So, I'm still a little hopeful 😁13:42
T42<Arilaviel> Wew14:49
T42<singhnsk> Hi @deathmist or somebody else, can you please check this new log?15:50
T42<singhnsk> Everything was going well and then it stopped with no error description. Does anybody have a clue?15:50
malvknecht: I tried again the video snapshot PR, fails on xperia 1015:56
deathmist@singhnsk show your fixup mountpoints, you've made an error as per the failed hybris-boot part15:57
T42<singhnsk> @deathmist The sectin related to my device is here -
T42<singhnsk> Do you need the full file?15:59
T42<singhnsk> *section15:59
deathmist@singhnsk you had some double spaces near the end of boot_b for example, use these ones instead
vknechtmal, hmm, since it use pretty much "standard" Qt stuff like qt_imageFromVideoFrame and QVideoProbe, I guess it's due to differences in Qt <-> Gstreamer or droidmedia support for this device ?16:05
vknechtor maybe with 3.2 ? didn't try on 3.2 yet16:06
T42<singhnsk> @deathmist thank you. I will overwrite and try rebuilding again :)16:06
piggzmal: so, the build system doesnt seem to use the compiler in the target, which is a pita16:13
kimmolideathmist: did you got everything sorted out? been busy with you know life16:14
T42<singhnsk> @deathmist It was going well and now I get `fatal error: 'update_engine/dbus-constants.h' file not found`. I have checked that dir and there is no such file indeed. But is is being called in ``. Should I remove that include? Or is there anywhere I can find this file?16:28
T42<singhnsk> The LOS source does not have it. But the same file on the source has it in the #include.  Not sure what to do.16:31
T42<elros34> @singhnsk your fixup looks wrong, you need to choose one slot16:47
T42<singhnsk> @elros34 So, I should not mention the slot_b? I kinda copied another fixum there from A/B devices. So..16:48
T42<elros34> which device?. Every block/bootdevice/by-name/dsp in fstab or *.rc file  will be replaced by mmcblk0p3116:50
T42<elros34> unlikly you have _a/_b in these files16:50
deathmistkimmoli: for sure! I improved the plugin a bit and the daemon can detect them properly now too, but seems to activate twice (not seem in logs) so it skips tracks 2 times and flash doesn't work (also switched to moden D-Bus alternative instead of messing with sysfs) :D16:51
T42<singhnsk> @elros34 So, how should I specify the slots? Because that's how the filesystem on these A/B devices is made na?16:52
T42<singhnsk> Or should I just use the _a ones as the exact name. Say boot instead of boot_a? And avoid using boot_b entirely?16:52
deathmist@singhnsk so the fixups aren't for your device? maybe try out my script as that'll generate all of them for you (run it in platform SDK env) :p  haven't had others test it yet either really16:54
T42<singhnsk> @deathmist I took them from my device only. I just used the formatting from an existing A/B device in the fixup file.16:55
T42<elros34> @singhnsk u need to use: block/bootdevice/by-name/system mmcblk0p15. Take a look at your fstab16:55
T42<singhnsk> @elros34 alright. Wonder why the official fixup file has the A/B mentioned for some devices 😢16:56
T42<elros34> I am not sure but you will know immediately whether your fixup are correct if you check generated mount units when building droid-hal16:59
T42<singhnsk> This was the partition structure as I fetched from my device.17:01
T42<Akatsu %lastname%> is there any fix on how to resolve lack of /dev/proporties ?17:41
T42<elros34> afaik this should be created by droid-hal-init unless it fails to start for you17:43
T42<elros34> it's also possible that file is opened and removed so you can't see it but it's still used17:47
riniguskimmoli: as I switched to xz2 as a primary device, I will not be able to contribute with the work on onyx. we can try to find someone on xda - maybe there are people wishing to make some work for the port.18:58
Mister_Magisterrinigus: you mean… what?19:16
Mister_Magisternot supporting phones you don't have on main?19:17
rinigusMister_Magister: yes. that's exactly what I mean. this will allow me to continue develop maps apps19:18
riniguseveryone has their preference...19:18
Mister_Magisterpicture above says everything19:18
T42<Arilaviel> Hi19:19
T42<Arilaviel> ;)19:19
Mister_Magisterrinigus: half joking though don't take this as an offence :P19:20
rinigusMister_Magister: none taken!19:20
Mister_Magisterrinigus: can i help you with maps in any way?19:20
Mister_Magisteryou literally won the market19:20
rinigusMister_Magister: its a huge stack, to be honest, if we talk about offline as well. depends on what you like. there is a work on databases, styling maps, better search engine, gui, splitting server to gui and dbus service, optimization of maps import. just to name few19:23
Mister_Magisterk so i can't help xd19:23
rinigus... issues are usually filed under projects, sometimes overlap19:23
Mister_Magistergui could be better, i was told i'm good at doing those19:23
rinigus:) I hope with expansion to other mobile linux platforms, we may get some developers working on it.19:24
T42<Akatsu %lastname%> if i have device using shims and it has export LD_SHIM_LIBS inside of init  suppoort it?19:41
vknecht@Akatsu %lastname%: you could expand a bit on the context of this question (eg. which android base, what are you trying to fix), and also check for this env var in irc logs20:30
kimmolirinigus: i need to update my build setup anyways, and maybe find some time20:30
T42崔明筐 %lastname% was added by: 崔明筐 %lastname%21:09
riniguskimmoli: great!21:51
rinigusmal: just tested with sdcard on xz2. it is detected as mmcblk0. so, we have to use custom nemo-qml-plugin-systemsettings for now21:52
rinigusr0kk3rz: just a reminder - would you mind to give me access to sailfishos-porters-ci, project tama-ci?22:02
T42<faenil> rinigus: I was planning to start working on a port of sfos for xz2 compact. Did you steal all the fun already? 😁22:17

Generated by 2.17.1 by Marius Gedminas - find it at!