08:00:40 #startmeeting Sailfish OS, open source, collaboration -- 17th February 2022 08:00:40 Meeting started Thu Feb 17 08:00:40 2022 UTC. The chair is flypig. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:00:40 Useful Commands: #action #agreed #help #info #idea #link #topic. 08:00:49 #info Meeting information and agenda can be found here: 08:00:54 #link https://forum.sailfishos.org/t/community-meeting-on-irc-17th-february-2022/10181 08:01:03 I am the meeting's chairperson today, and will be doing my best to keep time and order. Please respect the timings and stay away from the tiger. 08:01:08 #topic Brief introduction (5 min). Please prefix your name/handle with #info 08:01:12 #info Otto Mäkelä, community 08:01:32 #info David Llewellyn-Jones, sailor@jolla 08:01:39 #info pherjung, community 08:01:45 #info attah, community 08:02:07 #info Ville Nummela, Sailor@Jolla 08:02:20 #info fridlmue, community 08:02:27 #info Pami Ketolainen, Sailor@Jolla 08:04:30 Nice to see everyone. It's quite balanced today. 08:05:10 4-3 to community \o/ 08:05:20 Next time... ;) 08:05:27 #info Nikita/TheKit, Sailor@Jolla :) 08:05:36 Ahaa! 08:06:31 Oh well, can't complain about more particiaption 08:06:42 Very true! 08:06:48 Let's move on to our first question. Good that you're here attah, since it's from you! 08:06:51 #topic Calculation of minimum SFOS version in Store/Harbour (15 min -- asked by attah) 08:07:00 #info Continued from the discussion on the forum. 08:07:04 #link https://forum.sailfishos.org/t/apps-that-havent-been-ported-to-aarch64-store/7331/67?u=attah 08:07:11 #info Adding aarch64 packages limits the minimum SFOS version to 3.4 for the other architectures too. 08:07:16 #info Ideally the minimum versions should be specific to each architecture determined by applying whatever magic to the individual packages. 08:07:21 #info From last time the question was brought up it sounded like that was hard to do with the current Harbour architecture. 08:07:28 #info So i would like to ask if it is possible to not count the aarch64 package in current checks - since getting 3.4 as minimum version is just as wrong as 2.0 anyway. 08:07:32 #info Many nice apps are missing for aarch64 due to this. 08:07:56 Does that cover everything attah? If so, I can give the answer we prepared. 08:08:14 And some time yesterday(?) Wordle for aarch64 popped up in store, so something moved? 08:08:19 Yes, go ahead 08:08:25 Yes, something moved :) 08:08:27 #info Thank you for raising this attah. There's been a lot of earlier discussion about this internally, and it was useful to be prompted again to review it. 08:08:34 #info The Hourbour Masters have taken another look at the backend Store code, and have found a way to accommodate all of the requirements. 08:08:40 #info The changes have been deployed, so you can now already upload aarch64 binaries without blocking armv7hl binaries with support for Sailfish OS releases earlier than 3.4.0. 08:08:43 Wohoo! 08:08:47 #info If you run into any difficulties with the version detection, please let us know. 08:08:52 #info This is also a good opportunity to remind users that only the latest Sailfish OS release (4.3.0 as of today) is supported and receives updates from Jolla. 08:08:55 #info Older releases may be vulnerable to security issues which have been fixed in the latest release. 08:09:48 We also have Keto here, who made this happen, in case you have any questions. 08:10:43 Do you have some insight into how the heuristics work that you can share? 08:11:00 It now picks the oldest version compatible with any of the different architecture rpms uploaded 08:11:44 Nice, good to know - should work fine for almost all cases 08:12:45 With that i'm happy to surrender the rest of my allotted time unless anyone else has something on the topic 08:12:57 It is in theory not the correct way to do it, applying different versions to different architectures does not fit easily in other places in the model 08:13:23 And in practice this should not cause any issues 08:14:22 If an app is built against the latest SDK, it probably gets locked to something relatively recent anyway? 08:14:38 I agree... my preliminary suggestion was to do that unless version_magic(aarch64) came out to anything more than 3.4, but indeed this should work fine too 08:14:58 3.4 seems to be what you get from recent SDKs, via glibc i guess? 08:15:14 Okay, interesting, that's not so recent. 08:15:44 Do note that the compatibility there is purely checked based on the Requires that the rpm defines, so it does not guarantee that things actually work 08:15:46 Perfectly fine by/for me... but having limitations just because "computer says no" is also not nice 08:16:13 True - but at least in my case it does work 08:16:15 Keto, the developer has to do their due diligence? 08:16:34 yes 08:17:20 for example qml side things don't automatically get versioned dependencies 08:17:56 Right, there's no easy way around that. 08:19:20 Okay, well, going with attah's suggestion, let's move to the next topic a little early if there are no objections. We can always discuss further at the end. 08:19:33 Thanks for raising this attah, it was good to be reminded about it. 08:19:43 Thanks for the fix 08:19:57 #topic Small/low screen mode for the longphones (15 min -- asked by attah) 08:20:01 #info Continued from the discussion on the forum. 08:20:05 #link https://forum.sailfishos.org/t/small-screen-or-double-screen-for-the-long-models/6310?u=attah 08:20:10 #info Would it be possible to get a cropped display mode on the longphones? 16:9 would be especially relevant and should work out of the box with apps and UI. 08:20:14 #info Use-cases include better one-handed usability, producing screenshots in the correct aspect ratio for Harbour, and adapting to possible future video output (Xperia 10 III). 08:20:26 Thanks again for the questions attah. Here's our next answer. 08:20:34 #info Good idea, although we cannot really give roadmap promises for such new features. 08:20:40 #info Technically it is definitely possible, although it would require changes to closed-source UI packages like the homescreen so e.g. external contribution is not really possible. 08:20:46 #info When working with the feature phone proof-of-concept we have had versions of the homescreen draw the system UI and apps in a smaller area (and e.g. with a virtual phone keypad below). 08:20:51 #info Productisation naturally requires a different level of effort, e.g. to do this in an architecturally maintainable way. 08:21:53 That's the answer. Does that answer your question attah? Feel free to discuss it further. 08:22:11 Sounds good, and yeah -i was afraid it was partially in the closed bits 08:22:19 I also found this since https://github.com/mer-hybris/qt5-qpa-hwcomposer-plugin/pull/93 08:22:46 "Support for external displays #93" 08:22:46 attah, could you summarise what's behind that link for us? 08:22:55 #link https://github.com/mer-hybris/qt5-qpa-hwcomposer-plugin/pull/93 08:23:33 So a PR to the hwcomposer plugin that can take into account and use external displays... i only glanced at it 08:24:05 I assume it deals with mirroring only, and i saw some things with geometry that looked like it could take change sinto account 08:24:33 regarding that PR, it is only for older HWComposer (pre-Android 8) 08:24:53 Oh, i see 08:24:54 some extra adaptation is needed for HWC2 when the first one is merged 08:25:12 Is that planned TheKit[m]? 08:26:19 I ask because it looks like this is all open, so it could be looked at by the community otherwise? 08:26:45 I don't know if it is planned, but could be nice to have on Xperia 10 III 08:27:03 Definitely would be :) 08:27:23 And while you are here... how far away is video output on a device like the Xperia 10 III? (not so much in time as effort) 08:28:49 Mostly the similar improvement, or a long line of things? 08:29:07 hwcomposer2 part itself is not much, but it will be screen mirroring with current MR and different resolution of main display may cause issues 08:30:31 So, it would only be expected to work with 2520 x 1080 displays? 08:31:18 it can be scaled, but then we have either black borders or cropping because of different aspect ratio 08:31:41 That still sounds like a nice feature. 08:34:07 We have a few mins more on this topic, anything more to add? 08:34:34 Indeed, already a cropped output is better than none... and then if the os eventually grows a way to get this low display mode, i'd say it sorts pretty much all use-cases 08:34:56 I think output to 4:3 is mostly academic at this point 08:35:35 Academic because it's uncommon, or because those two cases would cover it? 08:35:47 Because it is uncommon 08:36:36 A letterboxed version would do in a pinch... and most apps wouldn't know what to do with it anyway 08:37:13 So at least from my outside view cost/benefit stops looking good there 08:37:23 I don't recall what the Tablet aspect ratio was. 08:37:41 wider than 4:3 at least 08:37:50 or taller.. 08:38:15 it was 4:3 08:38:17 Just checked... 1536 x 2048 pixels, 4:3 ratio in fact. 08:38:18 16:10 by eye 08:38:21 :O 08:38:29 need to have my eyes checked 08:38:36 :) 08:38:53 attah: there's a lot of black in the top and bottom, that's why it looks taller 08:39:30 True, i did look at just the screen but was fooled all the same 08:40:05 Okay, we used up the extra time, so we should move on to the next topic. 08:40:09 Non-square pixels is common in television, not so much with computers. 08:40:20 Thanks for another nice question attah. 08:40:37 #topic Duplex Telegram bridge for #saiilfishos-meeting (5 mins -- asked by flypig) 08:40:40 Thank you for good discussion 08:40:47 #info During the last meeting we discussed how we could improve the community meetings. 08:40:53 #info A nice suggestion from pherjung was to make full participation possible via Thaodan's Telegram bridge, which is currently read-only. 08:40:59 #link https://irclogs.sailfishos.org/meetings/sailfishos-meeting/2022/sailfishos-meeting.2022-02-03-08.00.log.html#l-24 08:41:02 #link https://t.me/joinchat/AAAAAFcbasJX67Fu-aGxxQ 08:41:07 #info We said we'd look into it, and this will be our chance to give an update. 08:41:13 #link https://irclogs.sailfishos.org/meetings/sailfishos-meeting/2022/sailfishos-meeting.2022-02-03-08.00.log.html#l-74 08:41:25 So, not really a question, but just an update on where things are at. 08:41:30 #info We looked further into what's needed to provide full-duplex support for Thaodan's Telegram bridge. 08:41:35 #info We concluded that meetbot would need updating to interpret the prefixes sent by the bridge. 08:41:41 #info Since the user assigned to the bridge can change, we've not yet decided how to tackle the issue of impersonation. 08:41:47 #info We plan to look at this further in the coming weeks and try to get something sorted by the end of May. 08:41:52 #info Either way, we'll set a deadline to give a further update at the 31st March meeting. 08:41:56 #info The meetbot code is open source python, so if anyone from the community wants to help with the update, let us know. 08:41:59 #link https://github.com/emonty/meetbot 08:42:32 In short: sorry, we didn't get it sorted yet, we hope to, but can't promise anything just yet. 08:42:32 I'm not a python expert at all, but I'll be happy to try and help 08:43:14 I'm not a python expert either, but the code looked quite clean to me. I think Keto looked at it in more detail though. What do you think Keto? 08:46:06 Yeah, so the bot runs Limnoria, and the initial idea is to see if a event filter plugin can be implemented that does some conversion on the messages coming from the bridge 08:47:04 Limnoria... This one? https://limnoria.net/ 08:47:11 yep 08:47:31 #info Meetbot runs Limnoria, the initial idea is to see if a event filter plugin can be implemented that does some conversion on the messages coming from the bridge. 08:47:38 #link https://limnoria.net/ 08:47:52 flypig we are using this meetbot https://github.com/keto/meetbot 08:48:24 Thanks Thaodan for the correction. 08:48:26 #info Jolla's meetbout source is https://github.com/keto/meetbot 08:48:32 #link https://github.com/keto/meetbot 08:48:39 the meetbot plugin does not matter if the message mangling can be done on lower level 08:49:45 anyway, if someone wants to take a look at that, you can ask me for more details 08:49:46 We hit time on this topic, I thought it'd be a quick one, so let's go to general discussion and we can continue. 08:50:03 pherjung, please do follow up with Keto! 08:50:23 #info Ask Keto for more info if you'd like to help. 08:50:29 should I create a topic on the forum? 08:51:02 perhaps other people would help 08:51:09 sure 08:51:48 pherjung, can I give you that as an action? 08:52:13 yes, I do it now 08:52:31 Great, thank you! 08:52:43 #info pherjung will create a topic on the forum about the IRC-Telegram bridge update. 08:52:56 #topic General discussion (20 min) 08:53:32 PSA: chum for version 4.3.0.15 finished building overnight 08:53:39 Thank you, Jolla, for considering the suggestion of one of the last meetings to release an interim bug fix version if needed. This was a welcome surprise for the security issues. 08:54:26 #info PSA from piggz: chum for version 4.3.0.15 finished building overnight 08:54:58 piggz, that's a rebuild of every package in chum? 08:55:08 yes, on 3 architectures 08:55:40 190 packages 08:55:41 Yes, thanks for the update. Question from a dilettante: why was this handled as a OS version bump instead of just making a package update, does this somehow make it more widely available? 08:55:58 fridlmue, ExTechOp, good to hear it was welcomed. 08:56:09 piggz, impressive :) 08:56:49 ExTechOp: Because SailfishOS branches of every version of the OS 08:57:04 It's just how we structure the development 08:57:18 It's really like in the meego/maemo days I think. 08:58:21 flypig: less impressive is that atleast one node has now ran out of disk space ...ping lbt! :D https://build.sailfishos.org/project/monitor/nemo:testing:hw:volla:yggdrasilx:4.3.0.12?arch_aarch64=1&defaults=0&failed=1&repo_sailfishos_4_3_0_12_aarch64=1 08:59:00 Yes, in a sense it was "just" a package update. All package updates come with a version number bump. 08:59:26 out of interest .. aswell as the security update, it was mentioned that it makes upgrade to 4.4 easier .. how/why is that? 08:59:36 piggz: pinged internally. 08:59:41 thx :) 09:00:23 piggz: for some users the repository configuration was wrong. This update fixes it, thus fixing the next upgrade. 09:00:27 Anyone dare to give estimates on 4.4 schedule? 09:01:01 ExTechOp: I was afraied to ask, haha. I would more ask if Jolla would share what are the main blockers ;-) 09:01:30 ExTechOp: Would you rather take "When it's ready" or "soon(tm)" as an answer? :) 09:02:36 fridlmue, the best place to look to see what's being worked on is the repo PRs. I'm not sure you'll get anything more concrete than that I'm afraid! 09:02:58 I already have a new phone I'd like to install :-D 09:03:28 ExTechOp, Xperia 10 III? 09:03:35 flypig: not that I expected anything else... ;-) 09:03:43 Yep. 09:03:50 fridlmue, of course :) 09:04:03 Topic is created : https://forum.sailfishos.org/t/update-meetbot-to-provide-full-duplex-support-between-telegram-and-irc/10331 09:04:12 ExTechOp, technically you already can. 09:04:37 pherjung, many thanks! 09:04:58 My English isn't so good, so if somebody can review it to avoid misunderstandings, could be nice :) 09:05:17 #info pherjung created a topic for discussion of the Telegram-IRC bridge update 09:05:20 #link https://forum.sailfishos.org/t/update-meetbot-to-provide-full-duplex-support-between-telegram-and-irc/10331 09:05:58 pherjung, looks excellent to me. Thank you. 09:06:05 you're welcome! 09:06:09 flypig: haven't we seen some 4.4 previews already running on the 10th year celebration in berlin? It looked relatively mature already. So my hope is a very well tested an solid release! 09:06:44 Yes, I think you're right fridlmue. 09:07:17 About the 4.4 update, why Jolla couldn't be a little more transparent? Perhaps Community can help with resolving som blocking issues :) 09:08:18 one feature missing from the gitlab -> github transition is the lack of the "activity" view which gave a handy overview of what was being worked on 09:08:29 was great for spying 09:09:11 like this https://github.com/sailfishos/sailfish-browser/pulse ? 09:09:26 no, project wide, across all repos 09:09:51 oh, much easier actually! 09:10:37 pherjung, we do try to be transparent, and we're always getting useful input from the community, it just maybe isn't always flagged as being release-related. 09:11:59 I'm sure there was a discussion about github visibility on the forum too, but I forget where exactly. 09:12:32 This is from there: https://github.com/pulls?q=is%3Aopen+is%3Apr+user%3Asailfishos 09:13:15 attah: looks like a lot of ofono activity :D 09:13:27 How would you describe that link for the minutes attah? 09:13:29 Pretty good, but closed PRs are not there of course... so maybe the query could be improved 09:13:56 List of open PRs to all sailfishos (organization) projects 09:14:10 Perfect; thanks! 09:14:23 #info List of open PRs to all sailfishos (organization) projects 09:14:25 #link https://github.com/pulls?q=is%3Aopen+is%3Apr+user%3Asailfishos 09:14:37 i guess this is the list of stuff that recently got merged https://github.com/pulls?q=is%3Aclosed+is%3Apr+user%3Asailfishos+ 09:14:38 So titles of features, and repos touched can be seen easily 09:14:52 piggz: yes, good job 09:14:54 Now the same for mer-hybris and we are on track. :-) 09:15:14 #info List of PRs that recently got merged 09:15:16 #link https://github.com/pulls?q=is%3Aclosed+is%3Apr+user%3Asailfishos+ 09:15:33 We're bumping up against time I'm afraid. Any final comments? 09:15:57 Oh, on the forum there is now a tracked tag. Some bug report don't have this tag, does it mean that Jolla didn't see it? Should the community ping Jolla's dev? Or just wait? 09:16:46 Pinging doesn't hurt I think. 09:17:05 As long as you don't go crazy. 09:17:26 Alright, we have to wrap up I'm afraid. Thanks for he great discussion today. 09:17:30 #topic Next meeting time and date (5 min) 09:17:34 Proposing Thursday 3rd March at 08:00am UTC 09:17:46 Any objections to the 3rd? 09:18:53 Okay, no objections raised :) 09:19:00 #info Next meeting will be held on Thursday 3rd March 2022 at 08:00am UTC: 2022-03-03T0800Z 09:19:10 Thanks again all! 09:19:11 #endmeeting