| piggz | mal: time to get rid of r0kk3rz repo.... | 16:56 |
|---|---|---|
| mal | piggz: I really can't think of what else to try | 17:00 |
| eugenio | piggz, https://build.merproject.org/package/show/nemo:devel:hw:xiaomi:latte/mce-plugin-libhybris-nondroid | 17:20 |
| eugenio | should also push the patched sensorfw as it's also missing | 17:24 |
| piggz | eugenio: nice ... time to start....walked the dog! | 18:04 |
| piggz | eugenio: does the .spec need a Provides: mce-plugin-libhbris ? | 18:05 |
| eugenio | piggz, I guess modifying the pattern is enough (PR incoming shortly) | 18:06 |
| eugenio | btw, does this look good for you? https://github.com/sailfish-on-latte/droid-config-latte/pull/7 | 18:07 |
| piggz | eugenio: yes, thats fine | 18:07 |
| eugenio | ok, I'll merge it then :) | 18:07 |
| piggz | i had an idea for a device earliler ... need to see if someone else already invented it .... basically, my car stereo has ipod control, but no BT ... i suppose a bt receiver powered by usb already exists ... but, what about one which bridges ipod control with AVCRP, so, the car could send ipod controls, which would be mapped to AVCRP to send to a phone, and get back metadata and albums etc | 18:12 |
| eugenio | how does ipod control work? | 18:16 |
| mal | piggz: can I rename the latte devel target? | 18:18 |
| piggz | mal: feel free | 18:18 |
| piggz | eugenio: dunno, i guess there is some ipod control protocol | 18:18 |
| piggz | over usb | 18:18 |
| mal | piggz: renamed, remember to changes repo path on device | 18:20 |
| piggz | similar idea https://forum.xda-developers.com/showthread.php?t=2220108 | 18:25 |
| mal | piggz: do you have now all needed stuff in latte project, is r0kk3rz repo still needed? | 18:27 |
| piggz | mal: im doing it atm.... | 18:28 |
| mal | ok | 18:28 |
| eugenio | piggz, cool | 18:31 |
| eugenio | piggz, https://github.com/sailfish-on-latte/droid-config-latte/pull/8 | 18:32 |
| piggz | eugenio: ah cool, i was going to ass what was needed for them | 18:33 |
| piggz | eugenio: the -keepalive rpm Provides pulseaudio-modules-droid, do you prefer the more explicit version? | 18:34 |
| mal | piggz: libva missing | 18:34 |
| piggz | mal: on it....probably need a bunch of audio codecs too | 18:35 |
| eugenio | piggz, hmm not sure... I don't think that there is something out there that depends on pulseaudio-modules-droid (excluding the hw patterns, but we control those) | 18:36 |
| piggz | eugenio: i thought there was.... | 18:36 |
| piggz | eugenio: yes, droid-hal-version rewuires it | 18:38 |
| eugenio | ok, so I guess we should keep the Provides field | 18:39 |
| mal | piggz: https://build.merproject.org/package/show/nemo:devel:hw:xiaomi:latte/droid-hal-version-latte | 18:39 |
| eugenio | (as it's picked up from the .inc file) | 18:39 |
| eugenio | nice | 18:39 |
| piggz | mal: huh, it worked! | 18:39 |
| piggz | now ti fixe up gst | 18:39 |
| mal | piggz: so I was right again :) | 18:40 |
| eugenio | :D | 18:40 |
| piggz | mal: yes, but you dont know _why_ it failed | 18:40 |
| piggz | :P | 18:40 |
| piggz | the packages should have conflicted still | 18:40 |
| mal | piggz: mixing different targets probably caused some interesting versions to be installed | 18:41 |
| mal | piggz: seems you are only missing the main gtreamer package from there | 18:42 |
| piggz | mal: i need to roll back some plugins to 1.14.1.. | 18:42 |
| eugenio | piggz, should we depend on the non-hybris variants (mce-plugin-libhybris-nondroid, pulseaudio-modules-droid-keepalive) on -version if native_build is set? | 18:43 |
| eugenio | I think we should | 18:43 |
| piggz | im not sure .... if we use Provides: then we dont need to make any changes ? | 18:43 |
| mal | piggz: better do it for all then, mixing different versions is not good | 18:43 |
| eugenio | piggz, if native_build is set the packages aren't picked up at all (talking about your forked droid-hal-version repository) | 18:45 |
| eugenio | (partially true... the pulseaudio modules are always picked up while mce-plugin-libhybris isn't) | 18:46 |
| piggz | eugenio: that was just supposed to be temporary to get it to build | 18:46 |
| piggz | my intentation was to remove that check on mce | 18:46 |
| eugenio | ok | 18:46 |
| eugenio | ok, will add a Provides: field then! :) | 18:48 |
| piggz | mal: i wonder if I even need gst-plugins-good and bad in the repo? | 18:48 |
| piggz | eugenio: cool, i think that will make d-h-v changes smaller, so it can be accepted easier | 18:49 |
| mal | piggz: probably not if you use 1.14.1 of all | 18:49 |
| mal | piggz: so if you would only build gst-vaapi 1.14.1 there and take others from main repos | 18:50 |
| piggz | yes, thats what im thinking | 18:50 |
| piggz | mal: and gst-av, is that packaed in main? | 18:50 |
| mal | hmm, not sure | 18:51 |
| mal | I'll check | 18:51 |
| mal | piggz: not in main repos, it's in common but a very old version | 18:52 |
| piggz | ok, i'll keep that maybe | 18:53 |
| eugenio | piggz, ok now the package supplies the Provides field (already built in OBS) | 18:53 |
| eugenio | will keep the correct one on the pattern | 18:53 |
| mal | piggz: just build correct version | 18:53 |
| mal | now it's 1.14.4 | 18:54 |
| piggz | mal: no, its .1 now https://build.merproject.org/package/show/nemo:devel:hw:xiaomi:latte/gst-libav | 18:54 |
| mal | piggz: indeed, sorry | 18:55 |
| piggz | r0kk3rz: why does your repo contain libvorbis, libtheora, mpg123 etc? | 18:56 |
| piggz | ok, removing plugins goo/bad ... i was playing to good to enable the v4l2 plugin for the camera, but that doesnt work yet | 19:01 |
| piggz | eugenio: do you want to submit the patterns to get built ... then we can race to upgrade to 3.0.1! | 19:05 |
| mal | piggz: maybe you should ask for the project to be patternized | 19:08 |
| piggz | mal: probably ... lbt ? :) | 19:10 |
| piggz | what does patternizing even do? | 19:10 |
| T4 | <eugenio_g7> It does extract patterns from the rpms and puts them alongside the repository metadata | 19:12 |
| mal | piggz: generates patterns from rpms to the format mic build uses | 19:14 |
| piggz | mal: so, just needed for mic? a live upgrade should work ok? | 19:15 |
| T4 | <eugenio_g7> In the meantime we could put them manually, in the _pattern project (which should be created) | 19:15 |
| piggz | and, would be needed for building on r0kk3rz build server ? | 19:15 |
| mal | piggz: not sure if live update needs those | 19:15 |
| piggz | eugenio: @eugenio-g7, i merged config | 19:30 |
| mal | piggz: should you change the repo used in OBS for config package | 19:34 |
| piggz | mal: oh, yes | 19:34 |
| eugenio | heh, sorry about that :) | 19:35 |
| piggz | mal: fancy testing my d-h-v changes against a droid device, and if it works i will PR ? | 19:35 |
| piggz | eugenio: upgrading..... | 19:42 |
| eugenio | good luck! | 19:42 |
| piggz | worth a go! | 19:43 |
| piggz | eugenio: also, mal sent some config updates for font sizes | 19:44 |
| piggz | i will add to configs later | 19:44 |
| eugenio | yup saw that | 19:44 |
| eugenio | I think we also need to pick up the -large variant for the icon theme | 19:44 |
| eugenio | at least I think the tbj uses those | 19:44 |
| piggz | i still think it s a bit odd looking, for example, in telegrame, the padding on the cover view on my phone looks much better than tablet | 19:45 |
| eugenio | yup, here too | 19:45 |
| mal | piggz: old revision in config _service | 19:48 |
| eugenio | just updaed | 19:48 |
| eugenio | updated* | 19:48 |
| mal | heh | 19:48 |
| mal | piggz: I'll test the dhv changes at some point | 20:00 |
| mal | piggz: eugenio tbj does have z1.5 large installed, also z1.0 | 20:02 |
| mal | probably z1.0 is always installed | 20:02 |
| mal | piggz: I suggest trying the z1.5-large in patterns here https://github.com/sailfish-on-latte/droid-config-latte/blob/native/patterns/jolla-configuration-latte.yaml#L15 | 20:03 |
| piggz | mal: ok, thx | 20:04 |
| piggz | mal: eugenio: rebooting | 20:05 |
| piggz | eugenio: ah rats, no UI! | 20:06 |
| mal | piggz: does it go to bootloop or just fail otherwise | 20:09 |
| piggz | mal: no, it boot up, and everything is loaded, it must be the mesa/qt side | 20:10 |
| mal | piggz: check the repos on device | 20:10 |
| piggz | mal: hang on, found it... | 20:11 |
| mal | piggz: so it didn't use wrong version from r0kk3rz repos | 20:11 |
| piggz | i remember this hack i did... | 20:11 |
| mal | ah | 20:11 |
| piggz | mal: eugenio: Feb 09 20:12:39 Sailfish lipstick[4265]: [F] unknown:0 - EGL Error : Could not create the egl surface: error = 0x300b | 20:13 |
| mal | hmm, did you check repo list on device? | 20:13 |
| piggz | one moment... | 20:13 |
| piggz | mal: repo list looks ok | 20:14 |
| piggz | mal: ah, it hasnt picked the qt version from the repo | 20:17 |
| piggz | sfos version is +git8, we have git8 | 20:17 |
| piggz | git7 | 20:17 |
| piggz | eugenio: would you mind updating ;) | 20:18 |
| T4 | <eugenio_g7> Not at all, once the football match ends :P | 20:22 |
| piggz | eugenio: forcing a downgrade atm.... | 20:22 |
| piggz | mal: eugenio: gui is up | 20:23 |
| T4 | <adampigg> (Document) http://149.202.119.142:9090/eWZoSV0GB8.png | 20:26 |
| T4 | <adampigg> (Document) http://149.202.119.142:9090/EnVVMp53lv.png | 20:27 |
| mal | piggz: cool | 20:31 |
| piggz | mal: yup | 20:32 |
| mal | piggz: in the future I suggest tagging config repo (and -version also) for nicer package versions but that is not relevant yet | 20:33 |
| piggz | mal: i know that, im almost an experieinced porter ;) | 20:33 |
| piggz | when i do that, ill also setup webhooks | 20:34 |
| mal | piggz: most porters don't about nice version numbers or clean commit histories | 20:34 |
| mal | piggz: actually your commit history for latte looks quite good :) | 20:35 |
| piggz | mal: eugenio: the hack that is needed, is moving /usr/lib/qt5/plugins/feedback/* out of the way, becuase of a missing symbol in the droid-vibrator plugin | 20:35 |
| piggz | thoughts? | 20:35 |
| mal | piggz: which symbol? | 20:36 |
| piggz | Feb 09 20:06:17 Sailfish lipstick[3815]: /usr/bin/lipstick: symbol lookup error: /usr/lib/qt5/plugins/feedback/libqtfeedback_droid-vibrator.so: undefined symbol: _ZN7ProfileC1EP7QObject | 20:36 |
| piggz | tha one | 20:36 |
| mal | piggz: where does it even the droid-vibrator, that is not the repo? | 20:37 |
| mal | +get | 20:37 |
| piggz | mal: its the native one | 20:37 |
| mal | piggz: but not in OBS project | 20:38 |
| mal | piggz: it's missing the whole package | 20:38 |
| piggz | maybe it was a leftove.. | 20:38 |
| mal | probably | 20:38 |
| mal | both vibra packages are missing from OBS | 20:39 |
| mal | piggz: hmm, where does it get those, version package needs and installs that | 20:40 |
| piggz | mal: no, hang on... | 20:40 |
| piggz | its a different package | 20:41 |
| piggz | mal: qt5-feedback-haptics-native-vibrator-0.1.0-1.2.1.jolla.i486 | 20:41 |
| piggz | the actual qt plugin | 20:41 |
| piggz | maybe becuase our base qt doenst match | 20:41 |
| mal | maybe | 20:42 |
| mal | piggz: I didn't know native vibra packages are in common repos | 20:45 |
| mal | I mean in jolla repo | 20:47 |
| piggz | mal: nope, its weird, built a version on obs, and it still has the same symbol error | 21:27 |
| piggz | mal: symbol is Profile::Profile(QObject*) | 21:28 |
| piggz | i guess from libprofile-qt5 | 21:30 |
| mal | piggz: hmm, odd | 21:32 |
| piggz | mal: it is | 21:34 |
| piggz | mal: should libprofile be lsited here? https://paste.gnome.org/po9z0vpk7 | 21:37 |
| piggz | mal: hmm, https://github.com/mer-hybris/qt5-feedback-haptics-droid-vibrator/blob/master/droid-vibrator.pro#L30 | 21:39 |
| mal | piggz: it's listed on fp2 | 21:40 |
| piggz | mal: i thought it should be | 21:40 |
| mal | piggz: so you'll have fix that | 21:42 |
| piggz | mal: atm im not seeing why it wouldnt be linked | 21:43 |
| mal | piggz: me neither | 21:44 |
| piggz | mal: easier fix..... | 21:47 |
| piggz | device does not have vibration! | 21:47 |
| piggz | mal: any way to force the package not to be installed? add something to the pattern? | 22:01 |
| mal | piggz: which package? | 22:04 |
| piggz | mal: qt5-feedback-haptics-native-vibrator | 22:04 |
| mal | commenting it out from patters and versio repos | 22:05 |
| piggz | oh nice, bluetooth-rfkill-event works with brcm-patchram-plus | 22:07 |
| piggz | i can scrap my custom service | 22:09 |
| eugenio | hmmmm the hw pattern already has qt5-feedback-haptics-native-vibrator commented https://github.com/sailfish-on-latte/droid-config-latte/blob/native/patterns/jolla-hw-adaptation-latte.yaml#L33 | 22:26 |
| eugenio | piggz, is there a reason why your droid-hal-version-latte has have_vibrator_native 1? Mine doesn't | 22:28 |
| eugenio | do we even have a vibrator on the device? | 22:28 |
| mal | a good question | 22:30 |
| T4 | <adampigg> eugenio,, we dont | 22:37 |
| eugenio | piggz, I have manually update the patterns on the repository, I think we could even try a mic build now (once I update qt) | 22:37 |
| eugenio | @adampigg: so have_vibrator_native can be set back to 0? | 22:38 |
| eugenio | updated* | 22:38 |
| mal | eugenio: I think not defining any vibrator defines in spec is the way, 0 probably also works | 22:39 |
| T4 | <adampigg> yes | 22:40 |
| T4 | <adampigg> i think i assumed | 22:40 |
| eugenio | ok, on my own droid-hal-version I have have_vibrator set to 0 and it worked fine (that was before the native vibrator has been introduced, but I guess it doesn't matter) | 22:40 |
| mal | eugenio: both ways work | 22:41 |
| mal | the conditions in .inc check that if any of the defines is 1 | 22:42 |
| mal | if no defined exist then it's the same as 0 | 22:42 |
| eugenio | ok | 22:45 |
| eugenio | in the meantime, updated qt is building | 22:45 |
| eugenio | @adampigg: took the liberty to disable vibrator on a separate branch in droid-hal-version-latte, https://github.com/sailfish-on-latte/droid-hal-version-latte/tree/piggz-native-build | 22:52 |
| eugenio | OBS build incoming... | 22:53 |
| T4 | <adampigg> i need to make bt fixes, but they can come latre | 23:01 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
