Tuesday, 2018-12-18

*** feodoran_ is now known as feodoran03:31
*** jvd is now known as jvd_06:36
dcalisteHello chriadam, how are you ?07:56
chriadamhi, I'm well thanks!  how are you?07:56
dcalisteFine, thank you. I've got some minor questions if you have time (and I've answered your (valid) remark about my latest secret PR).07:57
chriadamgreat, I'll check that later tonight07:57
dcalisteFirst, how can I get debug info from a 'invoker …' command line ? Usually I do 'export Q_LOGGING_LEVEL="*.debug true"', or something like that, but when using invoker, I got nothing.07:58
chriadamusually invoker sets up the environment according to some conf file07:59
dcalisteI need it, because from cli, calling "gpg2 -s" or "jolla-email" and signing make the collection unlock request appears, but when jolla-email is called through invoker, I don't get it.07:59
chriadamunder /var/lib/environment/ I think08:00
dcalisteOk, I'm looking at it…08:02
dcalisteSecond question is about staging accepted MR in gerrit for messagingframework: https://codereview.qt-project.org/#/c/247158/ has a +2 and I've a button to stage it, but since another PR entered first, what is the prefered way to go?08:03
dcalisteStage it and this may create a merge commit, or rebase it anf force push it again ?08:03
chriadamI personally would rebase it if the other one has been merged in the meantime08:04
dcalisteOk, so let's go that way.08:04
chriadambut it may not matter.  I'm not sure how the gerrit machinery does CR merges - it may do a cherry pick anyway08:04
chriadamdcaliste: regarding the email signing end-to-end stuff, are we still waiting for anything on QMF side?08:06
chriadamI see that there are a couple of CRs up there still - are they necessary?08:06
chriadamor should we be able to merge the jolla-email + n-q-p-em + jolla-settings-accounts changes already?  jpetrell how do you see those?08:06
dcalisteTo send attachments, the one I'm speaking about is required. I've modified the IMAP strategy to download parts to ensure that signed email with attachments are not downloading things twice, it's https://codereview.qt-project.org/#/c/248214/08:07
jpetrellchriadam: asked earlier pvuorela to check those before merging, afaik he hasn't had time to do that yet. pekka did say few days back that we should get those things merged so it is on his mind.08:08
jpetrellI am ok merging08:08
jpetrellpekka currently knows the email domain the best08:08
dcalisteThese two MRs in QMF are to correct bugs about signed email with attachments, that I didn't test properly during the first part of my work, when I concentrate on having simple text mails already working.08:08
chriadamjpetrell: dcaliste: I guess there's no huge rush - if those upstream QMF patches are needed ot make attachment signing work properly, maybe we can just come back to this after Christmas, and get everything in then?08:10
chriadamI mean, I was hoping to get it in before Christmas, but probably best to integrate when attachments are ready too, I guess08:10
dcalisteAbout the Mer side of things, the QMF MR will have to be ajusted to get the latest commit from upstream, otherwise, the other parts are ready.08:10
dcalisteWe need the MR in GPGme https://git.merproject.org/mer-core/gpgme/merge_requests/4 to go in also to be able to sign emails with S/MIME without running a GPG agent first.08:12
dcalistechriadam: I've explained in detail how the patch adding the function callback is actually making the launch of a GPG agent on the fly working.08:13
chriadamjust saw that, thanks08:13
chriadamcan you add that as context to the commit message also?08:14
chriadamthen I'll merge/tag that one08:14
dcalistechriadam: we can also merge all MR from Mer side to be able to have more people test the UI flow and design and have early feedback. Then, when the attachment MR are accepted upstream, we just need to update the upstream commit in Mer QMF.08:14
dcalistechriadam: ok to update the commit message, sure.08:14
chriadamdcaliste: on mer side what remains?  was there something outstanding in n-q-p-email or mer qmf?08:16
chriadam(well and that gpgme one)08:16
chriadamah https://git.merproject.org/mer-core/nemo-qml-plugin-email/merge_requests/3308:17
dcalistein mer qmf: the packaging of the crypto plugin: https://git.merproject.org/mer-core/messagingframework/merge_requests/1808:17
dcalistein nemo-qml-plugin-email: https://git.merproject.org/mer-core/nemo-qml-plugin-email/merge_requests/33 to add QML bindings for signature status.08:18
chriadamthat packaging one LGTM.  I will do a smoke test/build tomorrow and then merge it / tag it, unless someone says otherwise08:19
dcalisteGreat, thanks.08:20
dcalistechriadam: I've rebased https://codereview.qt-project.org/#/c/247158/ and requires (again) a +2 to be staged.08:20
chriadamregarding the nqpemail qml binding for signature support, I think that one is fine but would like pvuorela to review first too.08:21
dcalisteSure (and thanks for the CR about attachment boundaries, I've pushed the stage button).08:21
dcalisteSo, just https://codereview.qt-project.org/248214 remains, but requires full review (also pvuorela, if you have time, thank you in advance).08:22
chriadamThursday I'll ping matt08:23
chriadamdcaliste: regarding the jolla-email (and jolla-settings-accounts) PRs, can you use JB#43783 as the bug number for that08:23
chriadame.g. in the commit message etc08:23
dcalistechriadam: I think I'll update Mer QMF packaging request for upstream to point on latest after the attachment boundary fix is in.08:24
chriadami'll hold off on that one for now then08:24
dcalisteI'll do it later today I think.08:24
dcalisteOk, to update the commit messages in jolla-email and setting.08:25
chriadamthanks very much08:25
chriadamok, so just to summarise:08:25
chriadam1) mer-qmf packaging commit, waiting on the boundary change from upstream but then I will merge/tag08:25
chriadam2) nemo-qml-plugin-email needs review from pvuorela, but LGTM and I think jpetrell also reviewed it a while ago08:26
chriadam3) gpgme - LGTM once the commit message is updated then I will merge/tag08:26
dcalisteExact. I would add also:08:26
chriadam4) jolla-settings-accounts + jolla-email -> just need the bug number updated to JB#43783, then pvuorela to give final review.  I won't merge that, but will leave to jpetrell or pvuorela I think.08:26
dcaliste5) add dirmngr in the public repo when publishing the SFOS release x+1.08:27
chriadamlbt: will that (5) happen automatically or is there something I have to do?  I guess in that qmf packaging PR we need to ensure that it has "Requires: dirmngr" in the .spec?08:28
dcalisteI've added a Requires: on it in Mer QMF packaging08:28
chriadamI think that's all that should be required, but let's wait for lbt to confirm08:28
chriadamor pketo if lbt is on vacation08:28
dcalisteBut it's not a BuildRequires: so just mentioning it in case ;) You may not notice it without actually installing QMF.08:29
dcalisteAbout the bug raised by lipstick-secuiry-ui, do you have any idea how we can proceed, it hanged on me again yesterday.08:29
dcalisteI've quickly looked in the commit messages of QtWayland without finding anything that may be related.08:30
chriadamdcaliste: I do not :-(08:30
chriadamjpetrell: is that one a candidate for maintenance task?08:30
dcalisteI'm a bit afraid to rebase Mer QtWayland on QtWayland upstream @ 5.6.3 and test on device.08:30
chriadamif so I can look into it over the next couple of days08:30
chriadamdcaliste: that won't work anyway, really08:31
chriadamupstream changed a lot08:31
chriadamhopefully we move to qt5.9 soon and then we can try again, in the meantime it's mostly just hacking wayland internals :-/08:31
dcalisteJust using "secrets-tools --get-user-input" repeatedly and pressing cancel to go quicker make the thing finally hangs.08:32
dcalisteBut I've no idea where to put debug printf to get a clue on why the server side is not responding…08:33
dcalisteWhen Googleing for the symbol found with gdb, I also just arrive on hybris egl, which confuses me a bit more. Well I guess QtWayland is using EGL implementation provided by libhybris.08:34
dcalisteBut then, where to debug ? Maybe, I need to look in Hybris more than QtWayland…08:35
chriadamdcaliste: frajo might be able to help a bit also08:35
chriadammy expectation is:08:35
chriadamthe compositor (lipstick) talks to clients (e.g. lipstick-security-ui) via wayland.  internally, it allocates egl surfaces via libhybris (-> android surfaceflinger)08:36
chriadamso lipstick-security-ui at low level has qtwayland waiting for a response from the compositor (lipstick)08:37
chriadambut ALL OF THAT may be totally wrong08:37
chriadambecause I have NFI about any of that stuff08:37
chriadamabranson may be able to shed some light ^08:37
pvuorelalots of highlight on my name here :) yea, i'm aiming to do a review round on the signing stuff.08:38
pvuorelafor gerrit, i think it just always cherry-picks commits.08:39
abransonsorry, no idea how the compositing works. it's on my list of things to find out at some point :)08:39
chriadampvuorela: no rush, if you don't get to it this week it's no huge problem, just would be nice to get this stuff merged before christmas.  thanks!08:40
chriadamabranson: me too ;-)08:40
pketodirmngr needs to be added in our internal OBS, an in the release filtering rules08:45
chriadampketo: oh!  thanks - is there something we need to do for that?  (it was previously added to the gitlab Maintainers file also)08:47
dcalistepvuorela: thanks, and ok for next time I'll stage even if not base on master (if it can be cherry picked without conflict of course).08:47
chriadamdcaliste: I created JB#44199 internally to track the lipstick security ui issue and asked for it to be prioritised.  with any luck I will be able to do some investigation later this week, but let's see08:48
pketochriadam: probably best if you file a bug about it, I'm not sure of all the steps08:48
chriadampketo: Ok I will do that, thanks08:49
dcalistechriadam: great, huge thanks. I'll dig into various sources myself to better understand how the composition is working and where it may fail.08:49
chriadamdcaliste: no problem.  thanks again for your hard work and patience.  greatly appreciated, and I'm sure lots of SFOS users will appreciate the signatures feature once it all gets integrated.08:51
dcalisteWell, I didn't loose my time, now I understand much better the process of file signing and feel more confident to use it. I don't like magic processes.08:52
chriadampketo: created JB#44200 for that one08:54
dcalistechriadam: I've just updated the commit message in GPGme to explain the rational of the patch.08:55
chriadamI will merge that one tomorrow (and the secrets one)08:56
dcalisteSure, thank you.08:56
chriadamok, I have to head home.  Thanks for the productive meeting.  If we don't chat on IRC before the weekend, I hope you have a great Christmas and New Year.  I work this week, then will be back in the office from the 2nd of January.08:57
chriadamI will track carefully those PRs this week, hopefully we can get everything merged, but if not, I guess the new year.08:57
dcalisteSure, I wish you a happy Xmas time too.08:58
dcalisteI'll be off from the 24th to the 3rd of January myself.08:58
*** Renault_ is now known as Renault13:25
*** Renault_ is now known as Renault16:06
*** Renault_ is now known as Renault16:54
GreatEmeraldI just updated my Jolla 1 to SFOS 3 (it was a bit rocky but it worked), and all seems good, except for one issue19:59
GreatEmeraldWhen I take a picture, it's shown in the camera app history, but disappears20:00
GreatEmeraldAs soon as I close the app, it's gone, and it's not shown in the gallery while the camera app is open either20:00
GreatEmeraldIn the camera app I can see its name, so it looks like the file physically exists somewhere20:01
r0kk3rzhow long ago did you update?20:01
GreatEmeraldHm, I think the files exist, but they aren't indexed20:02
r0kk3rzok tracker should be well and truly finished reindexing by now20:03
r0kk3rzwhere have you set the pictures to be stored?20:03
GreatEmeraldDefault location20:04
GreatEmeraldSo that's Pictures/Camera20:04
r0kk3rzdouble check the pictures are there using the file browser20:05
GreatEmeraldI used the terminal and I think ls shows them there20:06
GreatEmeraldI'll sftp that to my laptop20:06
malare the file sizes ok?20:06
malGreatEmerald: if the files are ok but not showing in gallery then install sailfish utilities and clear the media database from settings->utilities so it will refresh it using tracker, hopefully that will help20:08
GreatEmeraldmal: Yeap, the pictures are fine20:08
r0kk3rzor 'tracker -se' if you cant be bothered installing the utilities20:11
malyep, there is of course a manual way20:12
GreatEmeraldThat's easy20:12
GreatEmeraldr0kk3rz: -se is not a tracker command20:13
r0kk3rztracker-control maybe, its been a long time since ive prodded around with it20:13
GreatEmeraldr0kk3rz: No such command, do you mean `tracker reset`?20:17
r0kk3rzyeah basically reset and reindex20:20
r0kk3rzperhaps the management commands have changed over the years20:20
GreatEmeraldI did a soft reset20:25
GreatEmeraldNow tracker status kind of just sits there waiting20:25
GreatEmeraldOh, now it says complete20:32
GreatEmeraldSure enough, now the photos are there20:32
malGreatEmerald: do new photos also appear there?20:39
GreatEmeraldmal: Good question20:50
GreatEmeraldOne I would probably be able to answer if the Camera app would start, hm20:51
GreatEmeraldmal: Looks like the answer is no20:54
GreatEmeraldThe lockscreen camera app works20:54
malGreatEmerald: how did you update the device, what issues did you have when updating?21:10
GreatEmeraldmal: Using the updater in settings21:13
GreatEmeraldmal: I had the issue of it not finishing the update; the boot screen showed "Sailfish OS" at 100%, but wouldn't finish throughout the night21:13
GreatEmeraldSo I had to pull the battery, and after the reset it seemed to work just fine21:14
GreatEmeraldBut it may not, indeed21:14
malok, long press of power button would have rebooted it also21:14
GreatEmeraldI tried that21:14
GreatEmeraldIt didn't work21:14
GreatEmeraldI held it for like 10 seconds21:15
malor vol up + power button for a long time21:15
malit needs to be qutie long time21:15
GreatEmeraldI tried that too, but for not as long21:15
malcheck output of journalctl when you try to use camera app21:17
GreatEmeraldUh oh21:19
GreatEmeraldI/O errors in mmcblk121:20
GreatEmeraldThere's a lot of things about msm21:23
GreatEmeraldBut it looks like debug/info, not warnings21:24
GreatEmeraldThe SD card dying happens twice in a row for me :\ That's weird21:26
malI have always found sd cards to be quite easy to break21:45
r0kk3rzthere are a lot of dodgy ones around21:59
GreatEmeraldI buy ones with a life-long warranty22:10
GreatEmeraldWhich comes in handy more often than I'd like22:11
*** Renault_ is now known as Renault22:12
flav0ris there a good browser out yet ?22:12

Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!