| *** ubuntu is now known as Guest76906 | 04:27 | |
| rinigus | kimmoli: how do we enforce removal of usbstick-utils? looks like OTA didn't do it. also, I cannot find which RPM package enforces OTA addition/removal of the packages (when checking provides/required at OBS) | 05:42 |
|---|---|---|
| *** ChanServ sets mode: +v T4 | 05:52 | |
| kimmoli | gm | 05:57 |
| rinigus | gm | 06:00 |
| kimmoli | we did something earlier with taaem | 06:02 |
| kimmoli | need coffee. but does it need to be in spec? not in patterns? | 06:17 |
| rinigus | me too (coffee). when does it read patterns? I'll try again with full ota sequence, maybe I missed something | 06:25 |
| rinigus | kimmoli: nope, redoing didn't help. note that tristate and other onyx packages are not required by any other packages. so, I couldn't find the dependency that way either | 06:41 |
| kimmoli | if patterns are only oneway? | 06:42 |
| kimmoli | did sdutils got removed? | 06:46 |
| rinigus | i am not sure I had them. not installed right now, but I think I haven't seen removal messages either | 06:50 |
| kimmoli | i think sd-utils was removed on previous update when udisks2 was introduced. | 07:39 |
| rinigus | kimmoli: ok, that could be. I expect that then udisks2 was dependency somewhere and adding it to the patterns is not needed as such. but how to get rid of old usbstick - that's not clear | 07:40 |
| kimmoli | https://git.merproject.org/mer-core/udisks2/blob/master/rpm/udisks2.spec#L79 | 07:42 |
| rinigus | kimmoli: good find! | 07:42 |
| rinigus | kimmoli: we have also https://github.com/mer-hybris/droid-config-onyx/blob/master/rpm/droid-config-onyx.spec#L28- it has few obsoletes commented out | 07:45 |
| rinigus | ... should we use that? | 07:45 |
| rinigus | ... and drop obsoletes from patterns? | 07:46 |
| kimmoli | yes, remove from pattersn, and add to droid-config-onyx.spec | 07:46 |
| rinigus | will do | 07:46 |
| kimmoli | just obsoletes should be fine afaik | 07:47 |
| rinigus | kimmoli: PR submitted | 07:52 |
| kimmoli | merged and tagged | 07:57 |
| kimmoli | i'll copypac when it finishes with patterns | 08:00 |
| rinigus | I'll try when its ready by OTA from devel. will report | 08:01 |
| kimmoli | looks silent... | 08:09 |
| kimmoli | do devel ota, check that the package got removed, i'll copypac then to testing | 08:10 |
| kimmoli | i think i'll try the CI built image | 08:10 |
| rinigus | kimmoli: looks like it removed usbstick, but silently (or even sneaky way). see https://ptpb.pw/hkFz | 08:13 |
| rinigus | kimmoli: will reboot and see if all seems fine | 08:14 |
| rinigus | kimmoli: re image - great idea :) | 08:14 |
| rinigus | ... image was built from devel though, not testing, if I remember chat over here correctly | 08:14 |
| rinigus | ... after reboot, seems to be fine as well. sdcard detected, anything else to check? | 08:16 |
| kimmoli | usb otg? | 08:17 |
| rinigus | don't have it => no testing | 08:17 |
| kimmoli | ok, i will | 08:17 |
| kimmoli | if i found my otg cable... | 08:18 |
| kimmoli | yeah, the .ks points to devel | 08:18 |
| kimmoli | i copypac now, and test ota from testing: | 08:19 |
| rinigus | sounds great! I'll move over to testing as well | 08:19 |
| kimmoli | https://gitlab.com/sailfishos-porters-ci/onyx-ci/-/jobs/103030089 | 08:37 |
| kimmoli | failed | 08:37 |
| kimmoli | typoed. | 08:37 |
| T4 | <adampigg> (Photo, 1280x720) http://149.202.119.142:9090/Wh0py5jk0E.png | 08:42 |
| T4 | <adampigg> (Photo, 1280x720) http://149.202.119.142:9090/lnNwQCegQB.png | 08:44 |
| T4 | <BirdZhang> đź‘Ť | 08:45 |
| T4 | <Ken_Ookami_Horo> (Sticker, 512x496) http://149.202.119.142:9090/72y1C1QNtm.png | 08:46 |
| kimmoli | rinigus: will do some more testing. but disabled the testing: builds for now. downloaded CI image, will test it after completed other tests | 09:22 |
| rinigus | kimmoli: I am still updating on testing. but disabling builds should be fine | 09:25 |
| *** vknecht_ is now known as vknecht | 10:34 | |
| elros1 | spiiroin: I can't build latest sensorfw on hybris-11: hybrisadaptor.cpp:282:49: error: 'const struct sensor_t' has no member named 'maxDelay' | 12:04 |
| elros1 | int maxDelay = (m_halSensorArray[i].maxDelay + 999) / 1000 | 12:04 |
| elros1 | looking at sensors.h and sensor_t does have minDelay but not maxDelay | 12:07 |
| mal | elros1: it seems that was added in cm.12.1 | 12:16 |
| mal | elros1: we need to add some ifdefs in the code | 12:16 |
| mal | elros1: build a bit older version until that is fixed | 12:18 |
| elros1 | yes, I am trying now | 12:18 |
| mal | elros1: looking at the code I think ifdef to one line might be enough | 12:20 |
| elros1 | mal: so I should just return 0 to: int maxDelay? | 12:24 |
| mal | elros1: no, this line is the problem https://git.merproject.org/mer-core/sensorfw/blob/master/core/hybrisadaptor.cpp#L282 | 12:28 |
| elros1 | yes, and this variable is used later so I am not sure what value it should have | 12:30 |
| elros1 | anyway 4fb14616beab35fd792b63f1c8c346ad92a30b4d builds ok | 12:30 |
| mal | elros1: you notice that the variable defaults to minDelay if there there if it's smaller than minDelay so you can just use 0 and it ends up as minDelay | 12:32 |
| spiiroin | elros1: yeah, some backwards compat problems possible - 1st fix was already done, then I had some obs issues and ... | 13:40 |
| spiiroin | I can look into it again tomorrow | 13:41 |
| elros1 | spiiroin: thx, no problem I built older version | 13:42 |
| Mister_Magister | jusa: ping | 14:19 |
| mal | spiiroin: probably #ifdef SENSORS_DEVICE_API_VERSION_1_3 around this line https://git.merproject.org/mer-core/sensorfw/blob/master/core/hybrisadaptor.cpp#L282 and just assignment of 0 in the #else | 15:00 |
| r0kk3rz | kimmoli: i think we could have a devel and a testing branch for ci | 16:10 |
| kimmoli | hmh. could. i usually OTA, but release only testing images | 16:12 |
| kimmoli | OTA devel* | 16:12 |
| r0kk3rz | yeah i guess it makes sense to point to testing for properly maintained devices | 16:14 |
| kimmoli | and imho the link on "pipeline passed" makes also more sense now... | 16:31 |
| r0kk3rz | yeah there is a way to link to the latest artifacts | 16:32 |
| kimmoli | i got spammed about failed job. not a passed one. | 16:32 |
| r0kk3rz | if you look at the link on the scorpion badge | 16:33 |
| r0kk3rz | http://gitlab.com/sailfishos-porters-ci/sailfishos-scorpion-ci/-/jobs/artifacts/master/browse?job=run-build | 16:33 |
| kimmoli | yes, that digs one step deeper than what i have on onyx now | 16:35 |
| kimmoli | is there access to bot to add webhooks from pipeline states? | 16:40 |
| r0kk3rz | no idea | 16:43 |
| r0kk3rz | looks like there is pipeline webhook events | 16:45 |
| kimmoli | yes there should be. now need a bot | 16:46 |
| r0kk3rz | a bot? for what? | 16:48 |
| kimmoli | to spam irc with those results ... like boss | 16:52 |
| r0kk3rz | heh | 16:54 |
| tbr | is that CI running on gitlab.com not on mer gitlab? | 17:04 |
| r0kk3rz | gitlab.com | 17:05 |
| *** ChanServ sets mode: +v T4 | 17:07 | |
| tbr | tough luck, no bot. merbot is only reachable from the inside as it is | 17:28 |
| tbr | IIRC I did even do some PoC for gitlab→merbot notifications | 17:28 |
| tbr | r0kk3rz: https://git.merproject.org/tbr/gitlab-irc-notifications if someone wants to set that up on mer infra, but with public URL | 17:39 |
| b0n0 | What package contains modifyrepo_c? | 17:48 |
| tbr | r0kk3rz: I guess it still runs as http://irc.in.merproject.org/push - would just need to be exposed to the public internet and https | 17:55 |
| tbr | ideally on a non-guessable URL | 18:02 |
| kimmoli | and maybe add secret token | 18:02 |
| tbr | that's the non-guessable URL part apparently | 18:02 |
| kimmoli | X-Gitlab-Token | 18:03 |
| tbr | ah | 18:03 |
| tbr | no clue if the mattermost integration sets that token | 18:04 |
| piggz | mal: spot the problem? https://github.com/qt/qtmultimedia/blob/5.6/src/plugins/gstreamer/camerabin/camerabinexposure.cpp#L257 | 18:10 |
| piggz | no vendor values are passed through | 18:11 |
| piggz | even if they were, would need to modify a layer in GST also to handle them, including gst-droid | 18:12 |
| piggz | maybe it would just be qt and gst-droid | 18:14 |
| mal | piggz: hmm, not sure how the vendor values should work | 18:19 |
| piggz | mal: i was thinking of just passing them stright through | 18:20 |
| kimmoli | its kinda good that devaamo input is low, can still stream netflix etc while uploading | 18:30 |
| b0n0 | When i try to insert wlan module it says 'couldnt insert wlan.ko operation not permitted'. Why this now? In 2.2.0 and previous i didnt have this error | 20:36 |
| r0kk3rz | what device? | 20:36 |
| b0n0 | A5 | 20:36 |
| b0n0 | Im the st333ve | 20:36 |
| r0kk3rz | did you ever publish your port? | 20:39 |
| b0n0 | Yes | 20:39 |
| b0n0 | 2.2.0 | 20:39 |
| mal | did you change anything in kernel before the new build? | 20:40 |
| b0n0 | No | 20:40 |
| b0n0 | I didnt touch anything | 20:41 |
| r0kk3rz | someone was asking for a port for the A5 not so long ago | 20:41 |
| b0n0 | I didnt even remember how to build sailfish, but luckily i created a script in the past :P | 20:41 |
| mal | you are using some systemd service the modprobe module? | 20:41 |
| b0n0 | r0kk3rz: who? | 20:41 |
| r0kk3rz | someone | 20:41 |
| b0n0 | mal: hmm, let me check | 20:41 |
| b0n0 | r0kk3rz: steeeve? | 20:41 |
| b0n0 | But i havent used that nick for months | 20:42 |
| r0kk3rz | no | 20:43 |
| r0kk3rz | id remember steeeeeve | 20:43 |
| b0n0 | Ah, rip steeeeeeeve | 20:43 |
| b0n0 | Anyway, my port is in my github, so, it's impossible to not finding it since im the only guy that keeps a5 alive | 20:44 |
| b0n0 | mal: yes i have it | 20:44 |
| b0n0 | https://gitlab.com/Marco01/droid-config-a5ultexx/blob/master/sparse/lib/systemd/system/wlan-module-load.service | 20:44 |
| mal | b0n0: just to be sure, what did you rebuild for the new release? | 20:45 |
| r0kk3rz | b0n0: is a build somewhere though? | 20:46 |
| b0n0 | mal: the same things i used to build in the past versions. I just run the same script every time i want to build sailfish | 20:47 |
| b0n0 | r0kk3rz: yes, it's a build: sailfish and lineage 14 zip | 20:47 |
| r0kk3rz | b0n0: where? | 20:49 |
| b0n0 | https://github.com/DeadSquirrel01/sailfishos-a5ultexx-releases | 20:49 |
| b0n0 | I think i will never push them to obs :/ | 20:50 |
| r0kk3rz | you should | 20:50 |
| mal | makes life much easier | 20:51 |
| b0n0 | Obs building is impossssible | 20:51 |
| mal | it's very simple | 20:51 |
| r0kk3rz | its really not | 20:51 |
| b0n0 | Hmm | 20:51 |
| r0kk3rz | unless you've done nasty things :) | 20:52 |
| b0n0 | I have a lot of nasty things | 20:52 |
| b0n0 | Some patches | 20:52 |
| b0n0 | Rip | 20:52 |
| mal | to which packages? | 20:52 |
| b0n0 | Ofono and something else i think | 20:53 |
| b0n0 | Need to check | 20:53 |
| mal | what doesn't prevent using OBS | 20:53 |
| mal | *that | 20:53 |
| mal | b0n0: with OBS you would be able to do OTA updates, not sure how many users your port has | 20:54 |
| mal | but would be nicer for users | 20:54 |
| r0kk3rz | no you can point things to your fork on OBS if you really need to | 20:54 |
| b0n0 | mal: not many i think lol | 20:55 |
| mal | but doesn't matter, it's also easier for you to update | 20:55 |
| b0n0 | In the past i tried that, then i filed | 20:56 |
| b0n0 | Failed | 20:56 |
| b0n0 | Didnt remember why | 20:56 |
| mal | you don't have the configs in your repo for updating kernel for example | 20:57 |
| mal | and updating requires OBS build probably | 20:58 |
| b0n0 | Yes, i know that obs is better but when i tried it i failed | 20:58 |
| b0n0 | I just dont remember why :/ | 20:58 |
| b0n0 | Maybe because patches are not allowed? | 20:59 |
| mal | there seems to be repo for you there already, you can always ask help debugging what went wrong | 20:59 |
| b0n0 | Yes i created the test repo | 21:00 |
| mal | b0n0: I still don't understand what patches you mean, if you have some patches package you can build it on OBS | 21:00 |
| r0kk3rz | your obs repo has nothing in it | 21:00 |
| b0n0 | This https://gist.githubusercontent.com/DeadSquirrel01/d992665ec1deae0297e0f68c500bd3d3/raw/3eda57d67884c406da2919cc1ca24c02d7dc9892/ofono_call-fix.patch | 21:01 |
| b0n0 | It's the patch from the s7 guy to make ril working on hardware/samsung librils | 21:01 |
| b0n0 | r0kk3rz: yeeeah | 21:01 |
| mal | b0n0: that is not a problem, you just fork the ofono repo, apply that patch and build on OBS | 21:02 |
| b0n0 | But then, id have to update my ofono fork. Cant that commit just be applied? I dont think it affects other devices | 21:03 |
| b0n0 | In upstream | 21:03 |
| mal | b0n0: has the person who made the patch asked for upstreaming? | 21:04 |
| mal | that needs some conditions to prevent it from being used on all devices, some config flag to enable it | 21:05 |
| b0n0 | I dont know if he asked that | 21:05 |
| b0n0 | I just saw&picked | 21:06 |
| r0kk3rz | updating the ofono fork would be much like updating everything else | 21:08 |
| r0kk3rz | pull the new code, wait for obs to churn | 21:09 |
| b0n0 | There's a flag that if is set in boardconfig, it will build samsung ril. So, it can be done something similar to the qcom bsp flag | 21:11 |
| mal | b0n0: anyway the patch is so simple that it's probably possible to get it in to ofono as a quirk | 21:15 |
| r0kk3rz | b0n0: is it the 2017 A5? | 21:16 |
| r0kk3rz | these guys were looking for a port - http://talk.maemo.org/showthread.php?t=100477 | 21:17 |
| r0kk3rz | ah its the 2015 one, i wonder what the difference is | 21:19 |
| b0n0 | The difference is so big that's theyre completely different phones :P | 21:20 |
| r0kk3rz | yes, i realised | 21:22 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
