libera/#maemo-leste/ Sunday, 2021-10-03

uvosnew iio-sensor-proxy should bring compass support00:06
Wizzupis there userspace that uses it?00:07
Wizzup(for testing)00:07
Wizzup(also fun for news update I guess)00:07
uvosWizzup: uh gnome mapps i think00:07
uvosWizzup: i also ported http://uvos.xyz/maserati/2020-09-16-132114_540x960_scrot.png00:09
uvosbut i need to package this00:09
Wizzup40400:10
uvosWizzup: i also ported http://uvos.xyz/maserati/screenshots/2020-09-16-132114_540x960_scrot.png00:10
uvossry00:10
Wizzupah cool00:10
Wizzupwhat app is this?00:10
uvos(this is something i wrote)00:10
Wizzupok00:10
WizzupI upgraded the kernel on my d3 and now it doesn't boot00:42
Wizzupwill attach serial..00:42
Wizzup(tomorrow)00:42
uvoshmm ok :\00:50
Wizzupcould do with some screenshots https://leste.maemo.org/Wireguard00:58
Wizzupparazyd: https://pastebin.com/raw/a34H9sNx 'update-rc.d foo disable' does the right thing01:55
WizzupI suggest we add that to the postinst and prerm of our {tor,openvpn}-network-applet packages01:56
Wizzupor postrm if it exists02:15
parazydWizzup: ACK09:44
Wizzupfreemangordon: so we agree: qt5 for conversations for now, right11:53
Wizzup(I think it's a good idea, just checking)11:56
lelMerlijnWajer assigned an issue: https://github.com/maemo-leste/bugtracker/issues/576 (wireguard/tor/openvpn status: add option to hide unless in use)11:59
lelMerlijnWajer opened an issue: https://github.com/maemo-leste/bugtracker/issues/576 (wireguard/tor/openvpn status: add option to hide unless in use)11:59
WizzupI'm going to rename 'tor-maemo' (just as user text) to something else12:30
Wizzuplike Tor support for Maemo12:30
freemangordonWizzup: yes, I think it is a good idea13:46
freemangordonmaybe we shall check what libraries conversations on fremantle use and see if there wouldn;t be some issue useng them from qt513:46
Wizzupfreemangordon: I worry about osso abook and rtcom14:01
Wizzupbut I think both will eventually work14:01
Wizzupfreemangordon: also we want to look at yappari I think14:01
Wizzupfreemangordon: it has the theming and everything14:02
Wizzupfreemangordon: e.g. screenshots here https://talk.maemo.org/showthread.php?t=8460514:03
Wizzupfreemangordon: looks like yappari can also use osso-abook14:07
Wizzupreally seems like a good starting point to me14:07
siceloyes it does (@osso-abook) - was required since that's how WA works14:32
Wizzupyeah I see it14:33
WizzupI never really got to use yappari much, got banned after like 2 days14:33
siceloi was a heavy user. when the bans intensified, i was 'recycling' up to 3 numbers per day :-p14:35
Wizzupheh14:35
Wizzupso I am thinking we take a lot of yappari's UI, but start with removing whatsapp-specific stuff (most) and add some basic abstraction for protocols14:37
Wizzupand then add basic telepathy / rtcom for xmpp or sms14:38
bencohoh14:38
bencohinteresting14:38
bencohI thought you'd go the path (take empathy / whatever and add missing features)14:39
Wizzupempathy is not qt14:40
Wizzupand it's also much further from what we want14:40
WizzupI think14:40
Wizzupit's also worth looking into, for sure14:40
Wizzupthe wiki really needs some love / organising14:41
Wizzupbencoh: got more thoughts on the matter?14:55
Wizzupit seems easier to me to hook up tp stuff to some cut down version of yappari14:55
Wizzupthere seems to be libempathy15:08
Wizzup# python message.py ring.txt +316<removed> tptest15:20
Wizzupworks :-)15:20
Wizzupuvos: what are your ideas regarding sphone and sms?15:37
bencohWizzup: not really, everytime I thought about that I wondered how we would add media support to empathy or rtcom15:41
bencohand pidgin never really seemed to do the work either15:41
Wizzupbencoh: I think telepathy supports tubes that are this, aren't they?15:42
Wizzupstill reading up on the terminology15:43
Wizzupor they are just another file transfer protocol perhaps15:43
bencohdata transfer yeah15:44
bencohconsidering what I just read, it might be a start, though15:46
WizzupI am not expecting to be able to do anything we want to do it15:47
Wizzupbut I expect that we'll just expand it to fit our needs if others haven't yet15:47
Wizzupbe that for mms or signal audio messages or whatever15:47
Wizzupuvos: wondering if it makes sense to make the sms stuff a separate application for sphone, or not, because I'm planning to work on tp support for sphone, but I won't work on the sms/text side of it, since we will want that to be a separate thing I think15:48
Wizzupuvos: does your bionic still boot on latest -devel16:32
freemangordonwhy sms should be separate from conversations?17:18
Wizzupfreemangordon: not separate from conversations17:18
Wizzupseparate from sphone17:18
freemangordonah, ok :)17:19
Wizzupfreemangordon: comments on my comments wrt yappari as base?17:19
Wizzupthoughts on comments*17:19
freemangordonnever used yappari so no idea how close it is as UX17:20
Wizzupfreemangordon: https://www.google.com/search?q=yappari+maemo17:21
Wizzup(mostly: messages overview and per-contact/chat)17:21
Wizzupseems to (re)use fremantle html stuff, or at least mimick it very well17:21
freemangordonI mean UX, not UI17:24
freemangordonbut yeah, lets see17:25
Wizzupok17:25
freemangordonsounds sane to me17:25
freemangordonat least it is a start17:25
Wizzupand it's qt, and has osso-abook and stuff17:27
freemangordonmhm17:27
freemangordonoh, https://github.com/agamez/yappari17:27
Wizzupyes17:27
freemangordonand Scorpius is there :)17:27
freemangordonif we ca pull him :p17:27
freemangordonWizzup: why did glib failed though?17:28
Wizzupfreemangordon: what do you mean there?17:28
Wizzupfreemangordon: no, the newpid stuff baffles me, but the others went through I think17:28
freemangordonbut, they are not in the repos17:28
Wizzupoh the -repos jobs failed17:29
freemangordonin the VM glib upgraded with the one from debian17:29
freemangordonmhm17:29
WizzupCannot put file 'libglib2.0-udeb_2.58.3-2+deb10u3+leste1+2m7_amd64.udeb' of 'glib2.0_2.58.3-2+deb10u3+leste1+2m7_amd64.changes' into component 'main',17:29
Wizzupas it is not listed in UDebComponents of 'beowulf-devel'!17:29
Wizzupparazyd: ^?17:29
parazydDelete everything udeb from glib/debian/17:30
parazydWe don't build udebs as we don't have debian-installer17:30
freemangordonsec17:30
freemangordondone, lets see17:36
freemangordonI've spinned a new build too17:37
Wizzupwon't fix the newpid thing I bet17:47
freemangordonwhatever it is17:47
freemangordonwe have a test failure?17:47
Wizzupsource linux namespace thing17:47
WizzupI don't think it's the test that fails17:48
Wizzupit's just using the command that fails17:48
Wizzupdidn't look too hard though17:48
siceloon my droid 4, when i received an incoming call with sphone, i could hear the other party, but they couldn't hear me. just reporting. maybe next time i'll have time to investigate further.18:57
WizzupI think I have the same problem19:00
bencohaudio routing issue?19:19
siceloyes19:25
WizzupI slightly modified this script to not exit after sending and just linger, and it can also nicely receive messages. It doesn't interfere with phone calls it looks like (empathy does) https://github.com/freedesktop/telepathy-python/blob/master/examples/message.py20:33
WizzupBut just running the script, or really, using a ring connection seems not to interfere with sphone, so that's good news20:33
freemangordonuh, just hit abook contact merger code :)20:59
freemangordonwill take a while20:59
Wizzupfreemangordon: hehe, no way we can stub that out for now?21:00
freemangordonwell, we can, but what's the point?21:00
freemangordonI have already put g_assert(0) on couple of places21:01
WizzupI think in ~2 weeks I will start on telepathy chat ui, and eventually that could use osso-abook, but it might not care so much about contact merging at the start21:01
Wizzupbut there's not really a rush21:01
freemangordonI guess by that time I'll have that ready21:02
Wizzup:)21:02
Wizzupeven better21:02
freemangordoncontact merger is about 3000 LOC, so shouldn;t take more than that21:02
freemangordonmore that 2 weeks that is21:02
Wizzupok21:03
freemangordonthe good think is that I think i am approaching finishing of that21:03
WizzupI don't know how much is left before e.g. modest can link against it21:03
freemangordonI know:21:03
freemangordonsec21:04
freemangordonWizzup: https://pastebin.com/wqZGMGDk21:05
freemangordonI already did osso_abook_temporary_contact_dialog_new code, but in turn it needs contact merger :)21:06
Wizzupfreemangordon: ah, I see21:06
Wizzupcool21:06
freemangordonthe other missing thing is OssoABookSelfContact21:11
freemangordonbut that should be easy21:14
Wizzupok21:14
WizzupI'm looking a bit into telepathy but trying to stop myself from actually doing something with yappari still because I need to do the news post and audio first :-)21:14
freemangordonand glib :p21:16
WizzupI was kind of counting that not being my problem but yeah :P21:17
freemangordonok, lets pester parazyd then I am fine either way :p21:17
freemangordonor, give me a hint where to look and I can look at it21:18
WizzupI am not sure. Could try to build it on the pinephone and probably see it not fail there21:18
Wizzupmaybe we need a new kernel for arm64, or otherwise figure out wth newpid does that makes things fail21:18
freemangordonwhat i snewpid, man newpid on my ubuntu says nothing21:19
Wizzupit's related to namespaces and I think it makes a new pid namespace21:19
freemangordonthe same on leste21:19
freemangordonNo manual entry for newpid21:19
Wizzupit's also installable as a package21:19
freemangordonhmm, ok21:19
WizzupNEWPID(1)                                                                                  NEWPID(1)21:19
WizzupNAME newpid - launch a command in a new PID namespace21:19
Wizzupso it's possible that perhaps our arm64 kernel does not support separate pid namespaces21:20
WizzupI'd have to ook21:20
freemangordonbut why it fails for glib only?21:20
Wizzupnobody else makes use of it21:20
Wizzupprobably21:20
Wizzuphm we do have CONFIG_PID_NS=y21:21
freemangordoncould you login to that machine and try to newpid something?21:22
Wizzupdoing atm21:22
Wizzupso much for now my problem ;)21:22
Wizzupseems to work21:22
freemangordonany idea what 135 is?21:22
Wizzupmaybe it is autopkgtest that fails then21:23
Wizzupbut it should not exit with 135 https://manpages.debian.org/testing/autopkgtest/autopkgtest.1.en.html21:24
freemangordondo you have enough free space?21:26
Wizzup/dev/etherd/e0.1   96G   19G   72G  21% /21:26
WizzupI could try to build it on the pinephone later21:28
Wizzup(it's armhf)21:28
freemangordonsee https://github.com/Codeception/robo-paracept/issues/2821:28
WizzupI saw that but wasn't sure if it was relevant21:28
freemangordonthey guess 135 is 128+721:28
freemangordonI think it is21:28
freemangordonbecause there is no 135 signal, no?21:29
Wizzupmaybe yeah21:29
freemangordonthis also suggests bus error:21:30
freemangordonhttps://groups.google.com/g/wkhtmltopdf-general/c/mvkQrGk_lng21:30
Wizzupok, but then some test triggers that perhaps21:31
WizzupI don't know, the log is not informative enough21:31
freemangordonI wonder if we can somehow strace or get a coredump21:31
freemangordonyeah, coredump sould help21:31
freemangordonit builds in chroot or in container?21:32
WizzupI am not sure how jenkins does it, it might be more than a chroot21:32
Wizzupin any case they get auto cleaned21:32
Wizzupand yes, we can get strace by changing how debian invokes the tests21:32
freemangordonthere is an option to not auto-clean21:32
WizzupI have no idea, that's all parazyd territory21:32
Wizzupsry21:33
freemangordonuser script /var/cache/pbuilder/build/cow.27699/tmp/hooks/B20autopkgtest starting21:35
freemangordonI guess those hooks are under our control21:35
freemangordonhttps://github.com/mika/jenkins-debian-glue/blob/master/pbuilder-hookdir/B20autopkgtest21:36
lelMerlijnWajer assigned an issue: https://github.com/maemo-leste/bugtracker/issues/565 (Conversations/SMS/IM user interface)21:36
Wizzupfreemangordon: oh I see21:36
Wizzupparazyd: yeah stracing that would be helpful21:37
Wizzupsomething like 'strace -f -o /tmp/log.txt $*' and then cat /tmp/log.txt21:37
freemangordonor, how to set that ADT to 'skip'21:37
WizzupI think you can disable tests in the debian pkg21:37
freemangordonvia gbp.conf I would guess21:37
Wizzupbut last time we suggested that you were quite against skipping the tests :P21:37
Wizzuphence me not suggesting it this time21:37
Wizzuppretty sure we had this same problem with glib in ascii21:37
freemangordongbp.conf is per package21:38
WizzupI think debian/rules works too21:38
freemangordonhow's that?21:38
Wizzupyou can disable tests there21:38
Wizzuphttps://iomem.com/archives/18-Avoiding-tests-when-building-Debian-packages.html21:38
Wizzupfreemangordon: btw do you agree https://github.com/maemo-leste/bugtracker/issues/565#issuecomment-933012505 ?21:39
WizzupAlong the way we will probably also need to look at the conversations stuff in ida a bit, but that's ok21:40
Wizzupas long as we don't need to do the whole thing21:40
freemangordonI think conversations on fremantle use rtcom-accounts21:40
Wizzuphm?21:40
freemangordon:nod:21:40
Wizzupto configure accounts you mean?21:40
freemangordonnot sure what for21:40
freemangordonyes, we'll have to look in IDA21:40
freemangordonto get the idea21:40
WizzupI found it's quite simple to at least configure a ring tel account21:41
Wizzupso at least for most of the telepathy interaction we can for now assume the account exists21:41
freemangordonfrom the shell, yeah21:41
Wizzupalso via code I think21:41
freemangordonactually we should have that preconfigured in leste21:41
Wizzupyes, something should21:41
Wizzupbut the tel acc uses a phone no I think21:41
freemangordonhmm, wait, installing telepathy-ring creates an account21:41
Wizzupdoes it?21:42
Wizzupk21:42
WizzupI thought I made it with empathy on my machine21:42
freemangordon/tel/ring0 or something21:42
freemangordonlemme check21:42
freemangordonring/tel/account021:43
freemangordonalso, osso-abook has that hardcoded21:43
freemangordonon fremantle it is different, but still constant21:43
Wizzupok21:44
Wizzupwell, either way, sounds like it can be used for initial conversations21:44
freemangordonlook at _osso_abook_eventlogger_remove()21:44
Wizzupyeah I don't know how to use eventlogger at all yet21:45
WizzupI plan to copy over my n900 db and see if I can see my messages21:45
Wizzup(eventually)21:45
freemangordonI guess we'll need to understand how fremantle conversations read the DB. I guess through eventlogger21:46
Wizzupyes, it does21:47
WizzupI did check that in IDA a while ago21:47
WizzupI guess I could push my renaming work for that somewhere21:47
freemangordonuvos will like that code for osso-abook:21:52
freemangordonpriv->modest_accounts_notify = gconf_client_notify_add(gc, "/apps/modest/accounts", (GConfClientNotifyFunc)cb, self, 0, 0);21:52
freemangordon*from21:53
freemangordon:)21:54
Wizzuphm?21:56
freemangordonosso-abook uses modest mail accounts21:56
Wizzupaha21:56
freemangordonfor 'me' account21:56
Wizzuphaving modest read my email account every 5 minutes however will probably make me install spamassasin on my server :-D21:57
freemangordon:D22:01
sicelothe eventlogger db is a simple sqlite db with easy to understand structure, thankfully22:38
freemangordonWizzup: hmm https://github.com/mika/jenkins-debian-glue/blob/master/pbuilder-hookdir/B20autopkgtest#L2022:39
freemangordonbut lets see what parazyd will say first22:40
Wizzupsicelo: yes22:40
Wizzupfreemangordon: you can also nuke that, there are various ways not to run tests, but is that the solution you're looking for?22:40
freemangordonno, I guess we want to see why it fails22:42
Wizzupok22:47

Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!