Tuesday, 2023-06-20

malakrosi: maybe something like this is missing https://github.com/sonyxperiadev/device-sony-common/pull/891/files but of course different content and paths, not sure how to get those values though00:17
malthere appears to be verizon firmware files on the device00:18
akrosiwell, ordered an xperia 10 III. I'll see if I can get VoLTE working; if not, it's not too big of a deal as I've wanted an excuse to change carrier for a while now02:50
akrosihas anyone tested emergency alerts? is there any system in place for their delivery? specifically referring to the Wireless Emergency Alert system in the US that makes phones screech whenever a child goes missing or there's a tornado nearby02:51
*** amccarthy is now known as Guest354303:33
*** amccarthy_ is now known as amccarthy03:33
dcalisteGood morning pvuorela, thanks a lot for the various reviews these last days.07:03
pvuoreladcaliste: good morning. quite a bit of code there to go through :)07:03
dcalisteIndeed, so let's start with pending questions in #60: should we support an 'after' argument in the listing routines ?07:06
dcalisteWith the backend, there is no calendar stuff, so all functions are returning lists of Incidences. There are the 'incidences(nbuid, uid)' and the 'incidences(qdate, qdate)'. But also the 'deletedIncidence(nbuid)'.07:08
dcalisteThe question is thus, should we add a 'deletedIncidences(qdate, qdate)' ? Or even a 'created(qdate, qdate)' and a 'modified(qdate, qdate)' for instance ?07:09
dcalisteSuch routines are currently used by some sync plugins (the Google one and the EAS one I guess).07:09
dcalisteBut, it's possible not to use them, as in the CalDAV sync plugin.07:10
dcalisteIn the Google plugin, at the moment, it is even done in a redundant way, since we list in the same routine (at the same time), all incidences, and modified incidences since.07:11
pvuorelai wouldn't think they should much affect the code except existing? one thing could be that if they are removed, i could rather have sync plugins remove the usage and port to new storage api separately than everything at the same time.07:12
henkI’m trying to install "Maliit Presage Plugin" via chum-gui but it gives some error message I don’t understand: »the installed libpresage-someversion requires 'presage-data', but this requirement cannot be provided«. the package "Presage Data" is installed, though. any advice how to get more info to figure out what’s wrong?07:12
dcalisteThis I agree, indeed. That's why, I kept sqlite accessors in the backend to keep support for ExtendedStorage.07:13
dcalisteLike that, rework of the API through CalendarStorage and MultiCalendarStorage can go on, while Buteo sync plugins are still buildable.07:14
pvuoreladcaliste: then the porting would probably go removing the "after" usage but still using ExtendedStorage. and after that from ExtendedStorage to other storage type.07:14
dcalisteExactly. That's my plan.07:14
dcalisteAfter everything settle a bit in mKCal (and ExtendedStorage is still in), I'll do a PR in Google Buteo plugin to remove after usage but still using ExtendedStorage.07:15
pvuorelaalso, we'll need to do eas for those.07:16
*** rdr_ is now known as rdr07:16
dcalisteIf everything goes fine, there will be a second PR to drop ExtendedStorage usage. If it's creating issues, there will still be time to add 'after' support in the new API.07:16
dcalisteSadly this plan implies more work from your side to adjust EAS plugin also…07:17
dcalisteIf it's too much of a burden, we can keep the 'after' argument for sure (maybe in a separated routine).07:18
pvuorelai'm still thinking there could be some guessing whether it's not useful in the end. if comparing calendar data to remote content, it could still be an optimization to skip comparisons on guaranteed in-sync events.07:21
dcalisteIndeed, it's an optimisation. But do you call 'allIncidences()' in EAS pkugin ? Because calling it kills any optimisation since it's loading in memory all incidences anyway. So listing one or two deleted incidences in addition will make no difference.07:26
dcalisteSee current status of Google plugin for instance in googlecalendarsyncadaptor.cpp#1688.07:27
pvuorelano allIncidences() in eas. but besides optimization on mkcal, it can be an optimization also on the sync side, maybe even more of that.07:29
dcalisteOk, so in EAS, you're doing it in an optimized way, just listing the difference (with the risk of missing something depending how the sync date is set). To keep it like that, I'll add routines in the backend with qdate arguments.07:31
pvuorelathe eas gets modified etc lists based on the after parameter. didn't yet figure out how easily it could be switched to compare everything, quite a bunch of code on that side too.07:33
dcalisteYes, I understand. So let's be a bit conservative in that matter. I'll add support routines in the backend and propagate them in the new CalendarStorage class also.07:35
pvuorelaalright, thanks.07:35
dcalisteAbout the alarm handling, I've created a dedicated PR since it's not relying on any SQLite changes, see mkcal#65.07:39
dcalisteAs I tried to explain in the PR message, the rational came from the necessity to add the Notebook uid argument and also to have a way to get the instances() of a recurring event.07:39
dcalisteBut this ends up in a partial rework on how alarms are handled, moving to a per series instead of a per incidence treatment.07:40
dcalisteThis avoid to treat cases like, if an exception is modified, check that its parent is also, or add it to the list of modified inciences.07:41
dcalisteLike that, you can call the setupAlarm() routine on the added + modified + deleted list, without paying attention of recurring and exceptions.07:42
pvuorelayeah. i'll keep on checking that. needs a bit of extra care since it's both moving and adjusting the code :)07:43
dcalisteYeh :/07:43
dcalisteAt least, PR#60 and PR#61 are becoming more and more like simple addition of new classes.07:44
pvuorelathat's good indeed.07:45
dcalisteBy the way, more a strategical question: in a while, after all of this API changes settle down, what do you think of the possibility to add KCalendarCore+mKCal as allowed libraries in harbour ?07:48
pvuorelahm, not totally out of the question if we get the api on satisfactory level and we're sure enough of it. might also need some work on the permissions, but let's see.07:51
pvuorelathen for another thing, i'm not having a long weekend off for midsummer and then on july there's the vacation time coming up. we could have a meeting still next week, but maybe some break after that.07:52
dcalisteOk, thanks for the answer on harbour question. Looks good. About summer break, sure. Myself, I'll be off the first week of July, and then, the last of July and the first of August. We can have a meeting next week and then enjoy vacation time !07:56
pvuorelasounds like a plan :)07:56
dcalisteOk, thanks pvuorela for the discussion this morning. I'll add the missing routines in mkcal#60 today or in the coming days.08:05
*** ilpianista is now known as Guest358310:13
akrosihad a thought about VoLTE on Verizon on the Xperia 10 III. The newer supported Xperias don't support CDMA, and the last test recorded on the provider chart at https://forum.sailfishos.org/t/volte-provider-support/11501 is in mid-2022, before the sunsetting of Verizon's 3G network. At the time, initiating a VoLTE call may have required negotiating through the CDMA 3G network... which the Xperia doesn't support.14:57
akrosiAll speculation, but I'm interested in seeing if I can coax it into working without changing carriers :) is VoLTE available without buying a license now, or is it still in beta?14:58
henkI’m trying to install "Maliit Presage Plugin" via chum-gui but it gives some error message I don’t understand: »the installed libpresage-someversion requires 'presage-data', but this requirement cannot be provided«. the package "Presage Data" is installed, though. any advice how to get more info to figure out what’s wrong?19:10
maltry refreshing the repo19:17
henkmal: done, no change19:23
malis the package installed?19:31
henkmal: no, doesn’t show as installed in chum-gui at least19:35
malI meant from commandline19:36
henkhmm, I never know what the right tool is … pkcon? zypper? rpm directly?19:37
malpkcon or zypper19:37
henkhm, haven’t found the right pkcon command to see that yet but this seems interesting: there are two maliit-presage-plugin packages, for different architectures it seems: http://ix.io/4yHL19:40
henk»pkcon get-packages | grep presage«: http://ix.io/4yHM19:41
henkno, this particular package is not installed but other presage packages are. I thought the whole list with 'presage' might be interesting as well for this issue, so I didn’t grep for that particular package but a bit wider …19:42
malso installing for example "Presage English US" from gui app doesn't work?19:43
henkmal: incorrect, that package is installed, as is the one for german. installing "Maliit Presage Plugin" (or "maliit-plugin-presage-1.2.1-1.2.1.jolla.aarch64", I guess) fails.19:45
henkI guess I should try with pkcon?19:45
henkinstalling presage-data: http://ix.io/4yHO19:47
henkFatal error: presage-data-2.0.0-1.2.1.jolla.aarch64 is already installed19:47
malno idea why that happens20:31
henkhttp://ix.io/4yHT ¯\_(ツ)_/¯ oh well, that worked20:35
attahsomething-something vendors20:37
malcould be20:38
attahhenk: what you mistakenly called architectures above20:39
henkit seems to me like the "maliit" and the "keyboard-presage" packages belong together and the "presage", "presage-data" and "presage-lang" packages and their "provides" and "requires" collide …20:41
henkoh, the former are version 1.2, the latter version 2.0. maybe20:43
henkthe pkcon error messages are crap o_O20:43
henk»Fatal error: the to be installed presage-2.0.0-1.2.1.jolla.aarch64 requires 'libpresage1', but this requirement cannot be provided«20:44
henkanyway: now the presage keyboard variants show up in the settings but selecting them does not cause them to be available in 'messages' … when I enable only presage variants, I get an empty field where the keyboard should be o_O20:47
henkwhen I enable one non-presage keyboard, that’s the only one available20:48
Mister_MagisterLet's play a game! Anyone who guesses first why sfos is not able to decode dts gets absolutely nothing!20:53
Mister_Magisterthat's right! Licenses!20:54
henkmal: it works for you, right? could you run 'pkcon get-packages | grep presage'?20:56
malhenk: only keyboard-presage-en_US, maliit-plugin-presage and presage-lang-en_US21:34
henkweird, almost same here, but also with de_DE variants …21:50
malI also have the finnish versions installed but that was not relevant so I skipped it21:55
maland you have the keyboard enabled in the settings?21:56
maland the normal keyboards if you want those21:56
b100dian[m]Mister_Magister: device trees?21:57
Mister_Magisterb100dian[m]: what21:57
Mister_Magisterno boooo21:57
Mister_Magisternobody does hw audio decoding21:57
b100dian[m]Ah, gotcha21:58
Mister_Magisterlol was doing something in gstreamer and found solution in… lipstick?21:59
Mister_Magisterwhat a coincidence21:59
henkmal: yeah. when I enable only the presage variant, I get only a blank blue area where the keyboard should be. when I enable a "normal" and a presage, I only get the "normal" variant and switching by holding the spacekey doesn’t show a selection. when I enable german and english "normal", as well as a presage variant, I only get the two "normal" keyboards for selection.22:11
malhenk: wondering how you manage to break things22:53

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