Tuesday, 2020-11-03

*** zbenjamin_ is now known as zbenjamin02:04
*** frinring_ is now known as frinring03:02
*** nyov is now known as Guest6365204:02
dcalisteHello chriadam, thank you for the different merges today. I've seen also that your Qt6 upgrade of QtPim has been validated, congrats !07:58
chriadamhi dcaliste, yes I finally got those integrated haha07:59
chriadamnow I can start to upstream some of our changes07:59
chriadamhow has your week been?07:59
dcalisteIt was a long fight with the CI ;)07:59
chriadamyes, it was haha07:59
dcalisteWeek was fine, thanks. Yesterday, I've seen this https://forum.sailfishos.org/t/calendar-bug-recurring-event-every-year-gets-treated-as-last-sunday-every-month/3316 and it's obviously misbehaving.08:00
dcalisteLooking into it, I think it's an issue in Silica.08:01
dcalisteYesterday, I asked pvuorela for write access to submit a correcting MR, but he was busy I guess.08:01
chriadamsilica bug, interesting08:01
chriadamI look forward to seeing the PR once pvuorela can sort out access08:02
dcalisteI think the problem comes from the fact that in the dialog with the selection, only visible items are shown (good), but the index that is emitted is the index taking into account only the visible items, while everywhere else, like in the _updateCurrent method, the index are counted on the model items without taking into account the visibbility.08:03
dcalisteThe patch is simple : in the dialog, instead of emitting the index with the visibility, we should emit an index without the visibility.08:04
chriadammakes sense08:04
dcalisteThe bug araises when you have enough items to show the selection dialog and when you have invisible items in the middle, and select items after the hidden ones.08:05
dcalisteI don't think it's that common case.08:05
chriadamyes, the ordering requirement makes it interesting08:06
dcalisteAbout the CalDAV syncing issue that Bree noticed, do you have any info on reproducing it ? Is it linked to Nextcloud specifically ?08:07
chriadamno info about how to reproduce the issue, yet, unfortunately.  Bree was going to create a bug with the info, but haven't seen it yet.08:09
dcalisteOk, if it's linked specifically to Nextcloud, I need to find if there is a free trial service somewhere, it would help to solve the issue.08:10
dcalisteIt would be bad to miss the next release...08:10
chriadamI use a docker image for nextcloud testing08:11
chriadamI'll see if I can pastebin it, sec08:11
chriadamI can't find my docker file currently08:16
chriadamI'll have to get it when I'm back in the office tomorrow08:17
chriadamsorry about that08:17
dcalisteNo problem, I'm totally ignorant about docker, but that would be good if I can start from your config. Thanks.08:18
dcalisteAnyway, I need to learn about it so that's a good opportunity. I've already switched the SDK to docker base one, and I need to use it from time to time at work.08:19
chriadamI really appreciate the help.  the issue which Bree had was able to be reproduced with master, she said, so it wasn't a regression08:21
chriadamhence why I merged that PR08:21
chriadamso I'm not exactly sure what / how to repro it08:21
chriadambut yeah, I will provide the docker file asap08:21
dcalisteIt seems to be linked to Nextcloud, because I cannot see it with OpenXchange. If I can reproduce it with a docker, that would be cool indeed.08:22
chriadamin regards to your other outstanding PRs, are we down to just the sailfish-components-accounts one, and the jolla-calendar one, currently?  oh, there's also the buteo-syncfw one that I was meant to merge a while ago08:23
chriadamam I forgetting any?08:23
dcalisteNo, that's the state of affair indeed ! About the buteo-syncfw, it's the one adding extended log capabilities per item, right ?08:24
dcalisteI think, I need to write the using code in CalDAV to see if the API is good enough and I didn't forget anything.08:25
chriadamyes, it's that one.  should I wait until you use it in the caldav plugin, just in case some other additions/changes are needed?08:25
dcalisteYes, I think we can delay it's integration up to the moment, the CalDAV part is ready.08:26
chriadamprobably for the best08:27
chriadamI think flypig is going to take a deeper look at the setEnabled() s-c-a PR08:27
chriadamaside from that, the jolla-calendar one is fine except for some strings, right?  maybe we should just merge that one... flypig?08:27
dcalisteNow that you merged the KCalCore bits, I'm going to publish also the changes in mKCal to inherit from MemoryCalendar and save a good amount of lines (even let say delegate maiuntenance to upstream KCalendar).08:27
dcalisteI've seen the notes from flypig on the sailfish-account MR. I'm going to answer later today in the MR to keep track of the discussion.08:29
flypigYeah, I only just took a look at the setEnabled() changes this morning and I think I need to think more on it. Feel free to tell me I'm wrong on both points dcaliste.08:29
Kabouik(Stupid question: is a SFOS meeting ongoing or just the channel is busy today?)08:29
chriadamwe meet with dcaliste at this time.  other contributors are welcome to join in, also08:30
dcalisteLooking quickly to your remarks, flypig, this morning before our meeting, I think I already checked on some while creating the MR, but I need to check again and answer properly, so later today I guess.08:30
KabouikThanks! I can't right now, other meeting in real life right now, but I'll read later08:30
chriadam(dcaliste is so prolific with his contributions that I wanted a regular session to ensure we stay on top of it all)08:30
flypigdcaliste, there's no rush for your reply, I expected you'd likely thought about it already.08:31
flypigFor the jolla-calendar PR: I'd argue for merging it already, but it'd be good to have pvuorela's blessing.08:31
dcalisteabout the jolla-calendar MR, sure, no problem.08:32
pvuoreladcaliste: what was that access thing? think i might have missed something.08:35
dcalisteIt's about https://forum.sailfishos.org/t/calendar-bug-recurring-event-every-year-gets-treated-as-last-sunday-every-month/3316, I think it's a silica bug and I prepared a MR for dicussion in silica but I have only READ access at the moment.08:36
dcalisteI commented about it yesterday around noon on this channel, but you may be busy at that time pvuorela.08:37
pvuoreladcaliste: ah, there. will handle.08:38
dcalisteThank you pvuorela !08:38
chriadamwas there anything else for us to cover this week?  I haven't yet fully digested the kcalcore upgrade information, unfortunately, so I don't have any input on that yet, or where we can help.  But I will have some availability / capacity free to help with that activity over the next couple of weeks.08:42
dcalisteI think for the migration of kcalcore, you can compile my new repo https://git.sailfishos.org/dcaliste/kf5-calendarcore08:44
chriadamoh wow, thanks08:44
dcalisteThen, you can take EAS or buteo social and see what it is changing for it compiles.08:44
chriadamI will hopefully get a chance to take a look at that next week, then :-)08:45
dcalisteSee https://git.sailfishos.org/mer-core/mkcal/merge_requests/16/diffs for instqnce what needs to be changed in the .pro file to compile with the upstream version.08:46
dcalisteIt will give you an better idea on the changes it requires to at least compile.08:46
dcalisteTesting and have it running is another story...08:46
dcalisteOh, I think the mKCal MR I'm pointing to is out of date, based on early tests for migration.08:47
dcalisteI've a better one on disk, but still not quite perfect, I need to polish it more (and make all the tests pass).08:47
dcalisteBut it's a complicated matter, because many things changed between KDateTime and QDateTime that requires to rewrite the tests completely but trying to keep the general idea of them.08:48
chriadamouch, yeah08:50
chriadamIf there is some nice way to divide up the work into chunks, I can definitely take some of those chunks08:50
dcalisteThanks, well looking for other repos I'm not very accustomed to is a good help indeed.08:52
chriadam:-)  buteo-social / sailfish-eas sounds like a good place for me to start08:52
dcalisteAnd doing a listing of all repo requiring KCalCore at the moment in their build requirements to check if we don't miss any is also a good help, because I cannot do that easily.08:53
chriadamthat's a very good point08:54
dcalisteI know pvuorela often did some grepping in all repos for me previously, maybe it's something you can do easily also ?08:54
chriadamyep, the grepping definitely08:55
chriadamI can also ask mkosola if there's some fancy dependency graph tool available from the build side..08:55
chriadambut let's see08:55
chriadamok, well, if nothing else to discuss, maybe we shall adjourn until next week?08:56
chriadamthanks again for your huge efforts and your patience with me being slow to review and merge!08:56
dcalisteYeh, my work for this week may be :08:56
dcaliste- open the mKCal MR on ExtendedCalendar rework,08:57
dcaliste- begin to implement per item logging now that the failure branch is in (it will reuse part of the mechanisms used for failure reporting),08:57
dcaliste- push the silica MR for discussions,08:58
dcaliste- try to improve the mKCal rebase on upstream (no promise).08:58
chriadamthanks very much :-)08:58
dcaliste- and also maybe try to setup a docker for Nextcloud testing.08:58
chriadamI'll send you the docker file tomorrow08:59
chriadamI will keep an eye out for the silica PR too :-)08:59
chriadamexcellent :-)  Have a fantastic week!08:59
chriadamthanks also flypig and pvuorela08:59
* chriadam -> home, gnight08:59
dcalisteYou too, thank you for the discussions and the reviews.08:59
flypigThanks guys, and thanks as always dcaliste :)09:01
dcalisteflypig, if you have any questions on your Buteo social MR, don't hesitate to ping here or on Gitlab when you have time to come back to it.09:02
flypigAh, yes, thank you dcaliste. I'm about halfway through making the changes. I'll push an update (hopefully this week, but I've got sidetracked with other stuff). Your comments were super-helpful. Thank you.09:03
dcalisteLooking at what you did, I wonder (or dream let say) if we could have a proper deserialisation absttraction on KCal::Incidences, because most of the sync-logic code looks very similar between CalDAV and Google one.09:05
dcalisteOf course it's more smoke thinking than actual work to do, because there are many subtle variations I guess, but many issues are identical between the two and bugs also it seems.09:06
flypigThat would be nice :) It's hard to say whether there might be sync edge cases that differ between the two.09:06
flypigIt would be a thing of beauty though!09:06
flypigIt would also help to avoid duplicating fixes. Right now I'm really duplicating the fixes you already did for caldav.09:08
dcalisteThat's what I notice looking also looking at your MR, but there no easy way not to duplicate at the moment :/ I tried several times earlier to divide NotebookSyncAgent in CalDAV into several classes to separate sync-logic from actual storage / incidence peculiarities but with no success. Always ended up with a spageti-like code.09:12
dcalisteBut well, looking at the Google sync implementation, it may provide me some new ideas, we never know.09:13
pvuoreladcaliste: adjusted silica branch permissions.09:42
dcalisteThank you pvuorela, it works.09:44
gmcugh, fell for it again.. qt 5.7 does not have endsWith on js strings .....09:50
gmcdcaliste: if you want, I can create an account for you on my nextcloud instance for testing09:51
dcalistegmc: that would be great indeed !09:51
samsep10lHey people! Anyone know what does dsme mean?16:34
samsep10lFor what is it16:35
samsep10lPeople do I need NFS support? Porting to Xiaomi Mi 8(dipper16:43
x2ssamsep10l: depends. Do you want to use NFS?17:14
samsep10lx2s: No actually, I am asking because I am not sure if it is needed by system.17:51
samsep10lI heard about NFS for the first time17:51
x2sJust one of the oldest network filesystems: https://en.wikipedia.org/wiki/Network_File_System18:10
x2sthe system wont need it to run18:11
attahWould anyone here in the fishyverse be interested in cellular protocol logging?18:53
fledermausattah: what level of detail?20:26
attahfledermaus: RRC procedures20:27
fledermausattah: fascinating. I have no requirement for such but I'm always interested in drilling down into lower layers of stuff.20:32
fledermausdoes sfos expose that to userland?20:35
attahmore or less, as root you can persuade /dev/diag to give it out20:36
attahThere are several tools that do this to slightly different extents... i hacked in some stuff to "QCSuper" to allow the program to work with a log forwarder you start manually on the phone (as opposed to having it started through adb)20:41
attah"This branch is 2 commits ahead of P1sec:master."20:41
attahAnd i hacked in dualsim log format so my Xperia 10 produces something useful20:42
attahThe better solution is really https://github.com/fgsect/scat, but i haven't been arsed to convince it to work with straight /dev/diag20:43
attahsupposedly it is enough, and perhaps more useful, to get the usbmoded mode right... but wasn't able to do that either20:43
spiiroinsamsep10l: dsme = Device State Management Entity; orchestrates shutdown/reboot, keeps taps on temperatures, hw and sw watchdogs, aligned timers, ...20:56

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