08:00:00 #startmeeting Sailfish OS, open source, collaboration -- 25th February 2021 08:00:00 Meeting started Thu Feb 25 08:00:00 2021 UTC. The chair is sledges. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:00:00 Useful Commands: #action #agreed #help #info #idea #link #topic. 08:00:08 #info Meeting information and agenda can be found here: 08:00:11 #link https://forum.sailfishos.org/t/community-meeting-on-irc-25th-feb-2021/4782 08:00:21 I am the meeting's chairperson today, and will be doing my best to keep time and order. Please behave and re5tecpa. 08:01:18 #topic Brief introduction (5 min). Please prefix your name/handle with #info 08:01:42 #info Simonas Leleiva -- privateer for Jolla 08:01:47 #info Ville Nummela - Sailor@Jolla 08:01:52 #info fridlmue - community 08:01:52 #info Andrew Branson - Jolla 08:02:03 #info Zyuc2G community member 08:02:21 #info Ruben De Smet - The Whisperfish guy@community 08:02:45 #info Joona Petrell - Sailor@Jolla 08:03:46 #info Björn Bidar - Sailor @ Jolla 08:03:46 #info cartron, community member 08:04:28 #info Lukáš Karas - community member, developer 08:04:59 #info David Llewellyn-Jones - Sailor@Jolla 08:05:44 Jolla 6:5 community 08:06:31 yay Jolla wins this time :D 08:06:31 #topic Has Jolla/SailfishOS resolved the licensing issues with GPLv3? (5min -- asked by Zyuc2G) 08:06:40 #info Recently SailfishOS appears to have added GPLv3 licensed applications including BASH 5, and grep 3.6. Has Jolla/SailfishOS resolved the licensing issues with the GPLv3? 08:06:51 #info If a Sailfish OS user decides so, they can install GPLv3 apps (such as gnu-bash) that Jolla packages merely for convenience (and for use on non-device infra, including SDK). 08:07:06 #info However such packages still cannot be distributed within the flashable images or pre-flashed devices. 08:07:43 interesting, thanks 08:10:12 can you package, just for convenience, newer qt and silica with it. so users, of they desire it, would be able to enable separate repo and install from there? :) 08:10:25 one can get many more business customers without GPLv3, that's what's mainly keeping Jolla's lights on 08:10:37 rinigus:) 08:11:04 ... assuming that gpl3 is keeping qt from us as well 08:11:23 rinigus: it'd be quite difficult to keep the system compatible with both 5.6 and the latest 08:12:07 I think fear is a major issue when it comes to Gplv3 08:12:28 abranson: that's correct. that hypothetical user would have to drop qt 5.6 then 08:12:35 gnu-bash et al. is easy because they don't require separate repos and are actively used in infra that builds images and sdk 08:12:47 I'm just thinking aloud here, but one option is to have GPLv3 in the standard community distribution, and I assume charging your paying customers for a Qt proprietary license 08:13:09 Thaodan: oh I don't know. it's a valid desire to prevent your users from messing with things in developer mode, especially for work devices. 08:13:24 Jolla can still offer all the benefits of opensource, while avoiding GPLv3 for its customers 08:14:00 it's other way round: we cannot get customers if we present a rootfs with GPLv3 08:14:10 abranson: There is true however that is about devices the user does not own (the employer is the owner) 08:15:04 Thaodan: that sounds feasible to us, but we are not lawyers 08:15:17 So the user of the device would not be able to change something 08:15:51 imho the fsf or fsfe should help clear up the situation 08:16:02 I believe that's already been investigated in the past 08:16:40 problem with lawyers is that they always take cautions approach. if they don't know, they prefer to avoid it. unless you find more reputable lawyer who is sure that it is OK. 08:16:44 they have lawyers and invented the tivo and non TiVo version of gpl3 08:17:56 good thought; but we're twice overtime now, let's exhaust Qt aspects in the topic after next 08:17:59 #topic Plan of VoLTE (5min -- asked by bionade24) 08:18:06 #info Lots of providers in countries all over the world are deactivating their 3G networks this year. Some providers in the USA even disable both 2G and 3G, breaking the call function on current SFOS versions. 08:18:14 #info What is the plan for VoLTE on Jolla, since you have to support it anyway once Rostelecom (which is a shareholder of Jolla) takes down der 2G and 3G network. 08:18:25 #info We have VoLTE on the roadmap, however have not committed to defined timescales at the moment. 08:19:04 i believe there could be some potential synergies with other mobile GNU/Linux 08:19:25 in other words don't expect anything soon :) 08:19:26 Is there some development effort by Jolla done already or a repo where we could have a look and probably help? 08:21:18 ApBBB: it is hard to work on something that is not public 08:21:26 it would start with investigation phase 08:21:34 we are studying VoLTE, but nothing to share now sorry 08:21:45 e.i. has almost no information available 08:22:29 We are open to work with others, if you have worked on that contact us 08:22:29 Thaodan public from jollas side or other? 08:23:21 just to get update on volte - sony aosp does support it, right? 08:23:24 VoLTE requires more from the OS side vs. 2G/3G calls handled a lot on behalf of the OS. and there doesn't seem to be much OSS available for VoLTE 08:24:05 ApBBB: from any side, even aosp is not that public about it e.i. documentation 08:24:21 So Jolla could found together with others e. g. a GSOC fore some students to set the corner stones?! 08:25:15 rinigus: aosp supports it but it has code that is written with non public information i think 08:26:40 assuming that jolla will write code for the VoLTE is this going to be able to be used by the ports or we will end up with a situation liek the fingerprint implementation 08:28:21 Thaodan: thanks! ApBBB good question. assuming that they wanted to work with other mobile linux os'es, maybe not. but good to get confirmed 08:28:49 What about creating some repo to collect information and docs and sources on VoLTE. Perhaps the community can help or other projects join? 08:29:55 fridl, if the community could help bring info together, I'm sure it would be a good contribution. 08:30:28 we need to move along, this is a big topic, so please raise further plans during next meeting 08:30:32 #topic Qt on SFOS moving forward (15min -- asked by KeeperoftheKeys) 08:30:40 #link https://forum.sailfishos.org/t/what-version-of-qt-are-we-on/4964/2 08:30:45 but sounds like we need some repo to collect such info. 08:30:51 #info As mentioned in the topic we seem to be somewhere in limbo between Qt 5.2 and Qt 5.6 neither of which have been supported for the past year and a half, newer versions of Qt bring lots of features developers are jumping to use beyond the very important aspect of securiy patches. 08:31:54 #info Situation is similar to VoLTE, we have Qt upgrade on the roadmap, however no schedule to share at this time. We want the big Qt upgrade as much as you do. 08:33:25 would moving the apps to flatpack solve the issue. 08:33:44 oh fuck i forgot about the wayland situation :/ 08:33:47 from the discussion above, the only similarity is a wish to have it. Qt is mainly licensing issue 08:33:54 flatpak requires newer Wayland, which requires newer QtWayland in lipstick.. so chicken and an egg :) 08:34:07 ApBBB: and wouldn't that still require silica rebase? 08:34:29 sledges probably 08:34:39 flatpak would be awesome to have though 08:34:42 assuming that we will use silica 08:35:00 #info The comment about some Qt modules lagging the Qt 5.6 is good, 3rd party developer pain-points which we can already alleviate. We recently upgraded Qt Positioning to 5.4, and now are updating Qt Wayland to 5.6. Some modules are already GPLv3 in 5.6 level. 08:35:00 id favor rpm os tree then instead of flatpak 08:36:04 about licenses.. newer Qt (5.9 and 5.12 for example) are LGPLv3 .. these discussions often mention GPLv3? Is LGPLv3 problematic for jolla? 08:36:39 and the current Qt 5.6 is LGPLv2 08:36:45 LGPLv3 isn't, but QtWayland is going for GPL only in 5.15 08:36:56 the QT situation need to be resolved at some point soon. it keeps the OS behind. i was expecting news with the release of SFOS 4 but .... 08:37:10 sledges: ok 08:37:10 ApBBB: +1 08:37:25 well, at least Qt 5.12 would be an improvement already.. 08:37:38 yes lGPLv3 is problematic, by saying GPL3 sometimes both lgpl3 and gpl3 is meant, sorry for the confusion 08:37:44 one step at a time:) i.e. 5.6 as mentioned ^^ 08:37:50 the situation on qt isn't getting any better even on the desktop 08:38:33 jpetrell: what exactly is problematic? the fact that lgplv3 requires users to be able to rebuild and upgrade qt libs? eg. can't lockdown the device? 08:38:52 sledges: that stop is not good enough. it used to 5.9, now 5.6. not the direction it is expected to move 08:39:06 they slow down releasing just slow enough to avoid the qt goes bsd clause 08:39:48 Thaodan bsd??? 08:40:04 pasik: yeah many industry customers see that as an issue 08:40:09 Thaodan: yes, but we are way behind desktop right now 08:40:40 jpetrell: ok, thanks 08:41:26 seems that the business model works for Qt well, e.g. many companies are willing to pay for commercial version to avoid opening their embedded stack 08:41:26 Ok, and something like deleting the encrypted user data Partition if Developermode is activated and Phone is in MDM-Mode is not an option for them? (Yeah, some warnings before that for sure) 08:41:30 ApBBB: read the contract that Qt has setup with Kde e.V. = over 12 months no foss release, kde gets a bsd licensed version of Qt 08:41:48 ApBBB: Qt/KDE have an agreement that Qt needs to remain FOSS enough for KDE, or it will trigger a relicensing of Qt to BSD -- or something to that effect: https://kde.org/community/whatiskde/kdefreeqtfoundation/ 08:42:00 rinigus: given the predicament, anything with version increments (not decrements) is hoped to be welcomed, we understand how you feel, we feel the same.. 08:42:14 but like said Qt is on our roadmap, you shouldn't read it "never going to happen" even if we cannot share schedule now 08:42:17 Zyuc2G: exactly that is what I.meant 08:42:38 ha. i had no idea about that contract 08:43:20 it is a self suicide imho for Qt when they don't follow the rules 08:44:38 * sledges 1min left 08:45:29 with the whole situation i still thing that choosing Qt as a toolkit for SFOS was a bad idea :/ 08:45:36 think 08:45:40 anyway 08:45:58 ApBBB, you would have stuck with GTK? 08:45:59 not at a time. and maybe it is ok now as well. 08:46:26 well they followed MeeGo as opposed to Maemo (Nokia N9 was the reason I quit the job and joined Jolla:) 08:46:49 let's move it on:) 08:46:51 #topic How to change playback speed of media files without changing the pitch (10min -- asked by thigg) 08:47:09 #info This is a blocker of multiple apps (gpodder, talefish, podqast) for some while now and nobody found a usable solution (asides from using QT5.12 features). Are there plans to do something about that? Does anyone have a workaround in mind? As far as I understand this, it could be done by modifying the pulseaudio pipeline. 08:47:57 #info We don't have any plans to implement this at the moment, but should be possible to do by using Gstreamer directly(?) 08:48:02 #link https://developer.gnome.org/gstreamer/stable/gstreamer-GstEvent.html#gst-event-new-seek 08:49:14 thigg is MIA, KeeperoftheKeys was mentioned as substitute:) 08:49:30 (the current dev of gPodder ;) 08:49:53 sorry had a bunch of meetings at work 08:50:59 no worries 08:51:31 for QML-only apps perhaps a suitable plugin could be developed? 08:53:47 what do you mean? 08:54:18 there isn't a way for modifying a pipeline in QML-only apps 08:55:29 abranson could shed more light on :) 08:55:46 as far as I understand it's a feature in one of the newer Qt releases (don't remember by heart if its 5.6 or 5.12) 08:56:36 We use gstreamer for media pipelines. Qt Multimedia wraps that up for Qt/QML apps. 08:56:38 we're currently discussing the workaround 08:57:12 It's already possible to do seeking in gstreamer, which I guess is what Qt Multimedia does in later versions 08:59:12 So it might be possible to build a plugin that wraps a seekable audio pipeline, and exposes it to QML 08:59:22 in your apps 08:59:30 I think if there is any workaround it is not harbor approved 08:59:52 unless you non qml code to the app 09:00:15 Could someone submit a patch to Qtmultimedia? https://git.sailfishos.org/mirror/qtmultimedia/blob/5.6/src/plugins/gstreamer/mediaplayer/qgstreamerplayersession.cpp#L376 09:00:27 (I mean, to patch our version) 09:01:18 if it's already done in a later version, isn't it a bit dodgy to backport the functionality 09:02:32 If it's a "cleanroom" implementation then technically should be okay (IANAL etc.). Practically though, it may be hard to demonstrate. 09:03:00 abranson: it is but if the backport was done independent than it should be fine 09:03:43 #info (as a workaround ^) 09:03:57 flypig: that code you linked looks like it already does it? 09:04:11 5.6 ? 09:04:44 branch said 5.6 09:07:31 It looks like this is the commit we're currently using: https://git.sailfishos.org/mirror/qtmultimedia/blob/1c0cfdf4/src/plugins/gstreamer/mediaplayer/qgstreamerplayersession.cpp#L317 09:07:51 so should it already work? 09:08:58 I'll try again, I haven't tried it for a while now. 09:10:05 #info with the currently used Qt Multimedia such functionality should already work 09:10:08 #link https://git.sailfishos.org/mer-core/qtmultimedia/blob/mer-5.6/src/plugins/gstreamer/mediaplayer/qgstreamerplayersession.cpp#L377 09:10:22 *could ;) 09:10:28 *might 09:10:33 Grabbing slightly back to the previous question - knowing exactly what versions we are on would be helpful to know to try something again, at the moment Qt is in limbo as far as I'm concerned 09:10:48 #info void QGstreamerPlayerSession::setPlaybackRate(qreal rate) 09:11:07 shoulda woulda coulda:)) 09:11:09 5.6.2 09:11:19 +git21 09:11:41 https://git.sailfishos.org/mer-core/qtmultimedia/-/tags 09:12:18 wait we have been able to set playbackrate for ages 09:12:30 the issue is that it is without pitch correction 09:12:50 the newer qt version adds pitchcorrection 09:12:58 #undo 09:12:58 Removing item from minutes: 09:13:01 ok, that makes more sense 09:13:02 #undo 09:13:02 Removing item from minutes: 09:13:05 #undo 09:13:05 Removing item from minutes: 09:13:14 :P 09:13:57 without pitch correction people start sounding like chipmunks 09:14:06 Did I get the wrong branch/commit? 09:14:33 thigg mentioned Qt 5.12 to have pitch correction 09:14:54 setPlaybackRate looks the same in 5.15 09:16:09 let's clarify in the next meeting 09:16:10 #topic General Discussion (15+ min) 09:16:38 jolla people. what is up with the recorder. why its not in the store yet 8) 09:17:11 we don't know where to put it :D 09:17:31 on top. its brand new and shiny 09:18:24 potential troll question - why are we being sent here when we can tag you guys in forum topics and also have wider community involvement there? 09:18:59 ApBBB: it's a system app, but we don't have a proper category to reflect everything about it 09:19:03 most devs don't spend time in the forum and here you know you will get an answer 09:19:11 +1 09:19:19 I don't mind IRC from time to time, but I also don't understand why if there is a serious topic on the forum it can't be discussed by Jolla people on the Forum 09:19:36 we have limited time and only to review burning questions 09:19:37 after all the forum replaces the mail list >:-) 09:20:29 working like elon musk - timeboxing ;) 09:21:37 @jpetrell a question. have you put any thought or looked into speeding up the UI a bit. (transitions and stuff like the patch in patchmanager) 09:24:11 also on the subject of patches. is there anyting that stops the extra search engines patch from being merged in the browser? 09:24:55 ApBBB: As a sailor who is not involved in UI development, but sitting in the same room with jpetrell and other UI guys, I can tell you: He and the entire UI team spend a *lot* of time trying to make the UI as smooth as possible. 09:27:09 ViGe i am mostly refering to the transition speeds (ie from home to notification) which feels slow. 09:27:20 ApBBB: which search engine patch is that? 09:27:53 there's something along those lines happening in the browser repo right now if you browse the recent commits 09:28:09 other than that there are a few places that i see issues. ie the media player, 09:28:23 (or check the next l10n round strings when that's out;) 09:28:52 ok. i'll wait for the translation strings 09:30:27 ViGe or to explain  what i mean by "slow". it doesn't follow the gesture speed nicely. like its "top speed" is limited or something 09:30:57 flypig: By any chance, did you have an opportunity to look into the WebView-captcha thingy? 09:31:06 it doesn't feel snappy in other words 09:31:08 (not that I have a lot of time today to look together, just informative) 09:31:59 rubdos[ma], thanks for flagging it up. I had a quick look, but nothing substantial to add yet I'm afraid. 09:32:28 Ok, thanks for the feedback :-) 09:33:05 ApBBB: It doesn't feel slow to me. Perhaps my fingers are slower than yours... :) 09:33:06 Haha, no problem :) I'll try to take a proper look at it and then we should discuss futher. 09:33:14 ApBBB: i think SFOS has a certain flow rate that all animations move at uniformly (Top menu, App grid, transition between pages) 09:33:47 flypig: Nice Update on the Community News :-) Great, that you care about that! 09:34:17 #link https://forum.sailfishos.org/t/sailfish-community-news-25th-february/5179/1 09:34:18 Yeh, the community newsletters are really nice :) 09:34:49 Thanks fridl, rubdos[ma]. It's worth saying that I get to post it, but it's a group effort. 09:34:51 Indeed, thank you flypig for the community newsletters on the forum ! 09:34:58 encourage all the devs and peeps you know to contribute to the digest via comments:) 09:35:13 Yes, that really makes things easier when we get the news given to us! 09:36:10 and on this note, it's time to end! 09:36:13 #topic Next meeting time and date (5 min) 09:36:15 Proposing Thursday 11th March at 8am UTC 09:37:01 11th March LGTM. 09:37:07 (Lithuanian Independence Day haha!) 09:37:26 +1 09:37:28 #info Next meeting will be held on Thursday 11th March 2021 at 8:00am UTC: 2021-03-11T08Z 09:37:38 thanks y'all! 09:37:39 #endmeeting