09:00:09 #startmeeting Sailfish OS, open source, collaboration – 19th of March 2020 09:00:09 Meeting started Thu Mar 19 09:00:09 2020 UTC. The chair is Jaymzz. Information about MeetBot at http://wiki.debian.org/MeetBot. 09:00:09 Useful Commands: #action #agreed #help #info #idea #link #topic. 09:00:27 Welcome to the new meeting room! 09:00:43 #info Meeting information and agenda can be found here: https://lists.sailfishos.org/pipermail/devel/2020-March/009083.html 09:01:18 I am the meeting’s chairperson today, and will be doing my best to keep time and order. Please behave, respect the timings and be gentle. This one is gonna be a long one, so let's respect the timing as we should. 09:01:43 #topic Brief introduction (5 min). Please prefix your name/handle with # info 09:01:55 #info James Noori - sailor @ Jolla 09:02:10 #info Ville Nummela - sailor @ Jolla 09:02:26 #info Nico - developer/user, community 09:02:36 #info BirdZhang - developer/user, community 09:02:44 #info Pami Ketolainen - sailor @ Jolla 09:02:56 #info Leszek, dev/user, community 09:03:11 Vesa-Matti Hartikainen - running projects @ Jolla 09:03:28 #info Daniel Eriksson user prospecive app developer 09:03:37 #info Leif-Jöran Olsson, community developer 09:03:38 #info Andrew Branson - sailor 09:04:05 Nice, we have a lot of sailors attending today :) It's fitting with the amount of topics too! 09:04:48 Probably a lot of people at home, I guess? 09:04:51 #info Vincent Knecht - community dev/porter 09:05:54 Nico[m]: True. Many sailors are currently working from home. 09:06:27 #info Timur Kristóf - sailor :) 09:06:32 First topic coming up in a few seconds 09:06:53 #topic Official support for PinePhone and Xperia 10 || (5 min – asked by cquence) 09:07:05 #info Is there any information you can share with regards to these 2 devices and getting official Sailfish OS images for them? 09:07:12 And here comes the answer 09:07:25 #info We are currently looking the next possibilities for Sailfish X devices. Requirements for a new device in general is availability for the community and enough openness by the vendor to allow providing an image to the device, similar to what we have done with Sony devices in the past. The mentioned devices are both potential candidates, but currently we can not confirm any plans. 09:07:37 #info If you have alternative in mind in addition to the mentioned two devices feedback welcome. 09:08:15 #info pasik, community 09:08:31 I don't see cquence online. Does anyone have anything to add? Otherwise we shall move on in a minute :) 09:08:47 Well, my favourites are the 2 mentioned devices :D 09:08:48 Xperia 1 flagship device with SailfishX would be cool 09:09:00 for sure 09:09:31 also could mean some effort will be put into the camera app which is very weak 09:10:00 #info David Llewellyn-Jones - late sailor @ Jolla 09:10:02 yeah, I don't think is is possible to ingore that anymore 09:10:18 Jaymzz: So it's officially that Pine64 sent you some devices and Jolla evaluates to support them? 09:10:24 some news on upcomming devices from OMP/Russian providers? 09:10:27 official support for Tama devices ? (XZ2 etc.) 09:10:48 vknecht if you have bugs report them on github 09:11:05 bionade24: Where did you get that info from? :) 09:11:12 Jaymzz: https://forum.pine64.org/showthread.php?tid=9086&pid=62012#pid62012 09:11:36 pm-guest We can't comment on customer device plans. 09:11:51 bionade24: can't really comment on that yet actually :) But stay tuned! 09:12:01 I guess they need a device with aosp10 to ensure that all works in upcoming devices as well 09:12:01 OK thank you) 09:12:17 ApBBB, I only have Loire and idol3 :-) 09:12:34 or moving to the linux based ones, but has to be available 09:12:46 Jaymzz: Would be nice to have a device without libhybris, and at least many people in the german user forum have the same opinion, too. 09:12:53 ach ja and Volla phone was financed. Maybe SailfishX for it would be also cool 09:13:00 bionade24: Noted 09:13:11 #info Damien Caliste, community, sorry to be late 09:13:20 dcaliste_: Welcome 09:13:27 We need to move on to the next topic guys 09:13:38 we have to keep the timing as it is packed today 09:13:42 moving on 09:13:55 #topic Qt update (10 min – asked by rinigus) 09:14:06 #info please give an update regarding planned Qt update. which versions are considered? has the work started? is it in active phase now? 09:14:27 Answer coming up: 09:14:29 #info A topic that has been on the table for a long time already, and for many reasons it has not been going as fast forward as we would have hoped. We have been working still improving the development offering and tooling by introducing new toolchain in the next release. 09:15:20 Does next release mean literal next release or some release in the future? :D 09:15:20 So it's not a GPLv2 ->v3 problem? 09:15:37 Sefriol: Next release means next release :D 09:15:51 in a few days probably 09:15:51 3.4? 09:15:56 Important bits and pieces to this. Qt 5.9 will be EOL. Qt 5.12 is the only viable option 09:16:00 Jaymzz: was that a full answer? in this case, which version of Qt is considered? 09:16:11 bionade24 The license still one of the issues we see. 09:16:22 Sefriol: And just to be 100% clear: new toolchain does not mean new Qt version. 09:16:25 Sefriol, to be clear, in next release, it's for GCC. 09:16:47 well, that has been very long overdue 09:16:52 very 09:16:56 rinigus We did a lot of work for 5.9, but as time has passed on from that we need to re-evaluate if it still makes sense to finalize as first step. 09:17:10 last year it was said QT5 license is not a problem anymore? 09:17:59 It should be priority right now. We devs won't sit here and wait for another year on an outdated Qt5 version for sure 09:18:09 do you have anyone working on it? like during last 2 months? 09:18:09 * schmittlauch[m] is planning to bring up Qt and GPLv3 licensing issues in the next meeting (finally) 09:18:09 pasik Well, for Jolla and for example Sailfish X it would not be an issue. With licensee's now it gets more complex. 09:18:30 veskuh_web: right, ok 09:19:07 veskuh_web: I could see the works on 5.9, but it seems to have stopped a while ago 09:19:30 schmittlauch[m]: will that include the recent (qt company) backporting/LTS policy changes? 09:19:34 veskuh_web: it sounds like a dillemma. You can't make both happy and need to decide on who you gonna dissapoint 09:19:45 *disappoint 09:20:13 3 minutes left 09:20:17 I think many devs in general are tired of using old software stack and some do not want to make an effort to start since some of the issues need to be fixed via hacks or just upgrading the stack :p 09:20:18 I understand hat 5.9 worrks will be finsihed with next release, correct? 09:20:42 Keitto: I don't think that's the case? 09:20:51 Keitto: no not at all 09:20:52 I don't think there is a way around it though - it needs to be upgraded soon (or later is long not an option) 09:20:59 gcc and libc will be updated not Qt 09:21:08 Qt 5.6 rebuild for newer gcc only 09:21:35 all alternatives are worse and likely impossible (switching to other toolkit or Jolla effectively maintaining a disjoint fork forever, including features people want) 09:21:40 Jaymzz: this question has not been really replied to. its not clear whether anyone is working on it actively 09:22:05 2 minutes :) just a reminder that we have to respect the timing strictly today. Otherwise it'll clash with Finnish lunch time and people will leave to eat ;) 09:22:05 unfortunately we currently won't yet upgrade from Qt 5.6, not because we don't want to, but due to the same reasons that were discussed almost every time this came up in the past. 09:22:32 No one wants to work on it if the license thing is not cleared up. Otherwise it is like the Qt 5.9 work and it will never be released and a waste of time 09:23:00 this has to be solved though. 09:23:11 sfos is held back due to this 09:23:24 Venemo: so, it's license, right? 09:23:29 like I said. Devs will vanish 09:23:49 and if people need latest wayland stuff smoething older than 5.12 wont cut it 09:24:57 On top of that we can only collaborate in limited areas with Ubtouch and Plasma Mobile on devices like the Pinephone. On apps almost not at all 09:25:09 ApBBB: taking into account that we have qtwayland 5.4, anything higher may make life better 09:25:15 We need to move on. If any discussion is left, please have it during the general discussion at the end of the meeting. ' 09:25:26 Yes, we are aware that this is challenging situation. With the updated toolchain we are pushing also other ways to help developers and hopefully we could also progress the public API. 09:25:45 rinigus if you need the latest protocol you need 5.10 -12 + 09:25:50 xdg 09:25:57 #info With the updated toolchain we are pushing also other ways to help developers and hopefully we could also progress the public API. 09:26:08 that wasn;t working on desktop defore 5.10 09:26:20 Moving on to the next topic, time is over on this 09:26:28 #topic Developer Forum (5 min – asked by Sefriol) 09:26:38 #info Current progress? Will this also include a roadmap promised years ago? Future of TJC? :) 09:26:52 #info Progress has been slow. The plan is to get it up and running during Q2/2020. The plan is to merge TJC to the new forum later. 09:26:52 ApBBB: thanks 09:27:16 Sefriol: Anything to add? or shall we move on? :) 09:27:38 More or less, I think ease of communication is the key to move Sailfish forward :) 09:28:09 Indeed. Hopefully having all means of communication in one place will help massively with that. 09:28:11 I'm fine to move on, unless somebody has anything to add 09:28:22 moving on in 30 sec :) 09:28:29 Jaymzz: Will the forum be more of bulletin board/ discourse style or more like Askbot? 09:28:48 schmittlauch[m]: Discource style :) 09:29:02 And you didn't say anythign about the roadmap ;) 09:29:05 yay 09:29:26 Sefriol: Can't comment on that now unfortunately :( 09:29:42 Sure, let's move on. Talk in general if needed 09:29:45 Point 1 on th roadmap: Don't talk about the roadmap :P 09:29:49 :D 09:29:53 Alright moving on 09:29:59 #topic gcc c++ support and qt upgrade progress (15 min – asked by Sefriol) 09:30:02 Discourse style, will it be to discuss technical matter also ? Or only user experience oriented ? 09:30:10 #info Personally I have been holding on doing more with Sailfish because of the aging software stack that was supposed to be upgraded last year. What's the current status and how big of a priority is this? 09:30:35 Oups, sorry, forget my question, let's keep the current topic. 09:30:36 dcaliste_: It's *developer* forum 09:30:46 dcaliste_: it is called a "Developer Forum" so the name would be silly if technical matters weren't allowed 09:30:52 dcaliste_: Hopefully everything. The point will be to have different forums there for different topics :) 09:31:05 Answer to this topic: 09:31:12 Ok, sorry, forgot the name. :/ 09:31:14 #info Next release contains gcc 8.3, you can see this work already in the repositories https://git.sailfishos.org/mer-core/gcc/ surely this is not the latest and greatest, but at the time we started the work 8.3 was the latest supported toolchain in the same Linaro series we have used in the past. Now hosted by ARM though. We have noted that gcc 9.2 is also there available now, and there is discussions to do yet another updat 09:31:14 e on toolchain and do these in a bit faster phase than before. Also the updates should be easier now that we have relatively new toolchain already. 09:31:26 #link https://git.sailfishos.org/mer-core/gcc/ 09:31:49 cquence: We discussed your topic first. Please read the logs :) 09:32:03 Thank you, sorry I am super late 09:32:11 Do we need to rebuild applications? What ABI will be used by default? C++11 ABI? 09:32:26 cquence: No worries! 09:33:27 Nico[m]: You shouldn't have to rebuild applications. 09:33:59 I guess lipstick and others are compiled against newer gcc to gain better security and a bit better speed? 09:34:17 ViGe: Sounds great, when can I get my hands on it? :D 09:34:25 leszek: sure, everything in the OS has been rebuilt 09:34:52 Nico[m]: As soon as it's ready to be published to early access users. 09:35:27 I was hoping for a date, but alright :3 09:35:39 I think I can wait for the next release 09:35:58 Nico[m] translatios were done a few weeks ago so it wont take long 09:35:59 Nico[m]: We can't really give dates, as there is always a possibility that a new blocker bug is found next minute 09:36:09 probably next week I assume 09:36:20 if no bug 09:36:30 Do you feel that future upgrades are still going to take as much time as 8.3? 09:36:42 Yeah, I understand that, I'm usually in the same position ;-) 09:36:46 Like are we going to see 9.2 in this year or the next? :D 09:36:49 Next week sounds awesome! 09:37:19 Sefriol We are hoping that keeping toolchain uptodate would now be easier and we could do it more frequently 09:37:32 Sefriol: Well, they said they are discussing faster updates in the future 09:37:41 getting from 4.8 to 4.9 was the hard bit ;) 09:39:17 a bit off topic but what community apps couldn't be used because of the old gcc?? 09:39:23 About dates, I'm sure you have target release dates? Why not share those? Ofc, it's up to community to know what target means :p 09:39:31 That's why the Roadmap would be nice 09:39:55 Like many of us are devs who understand how this world works 09:40:11 Maybe they're agile ;p 09:40:42 Oh, many firms also say that they are "agile" 09:40:46 agille projects use roadmaps too 09:40:48 The problem is with non-devs which will start complaining if you're late by more than 5 minutes 09:40:49 ApBBB: libs used by pure maps and osm scout server. had to compile with my own copy of newer gcc 09:40:55 Sefriol, I'm afraid they've been bitten already by some taking planning dates as promises... It makes them over cautious now. 09:40:57 5 min left on this one 09:41:03 A bit offtopic too: Will the meson package be updated too? I want a feature, where the version was 1 to low before, iirc 09:41:11 Sefriol but many arent devs and there is going to be a shitload of "but you said this week" posts 09:41:37 ApBBB exactly 09:41:47 Yeah, I know. But now there are a lot of complains of not having any dates :p 09:41:49 rinigus its the app i wait the most to hit the store once the other issue with systemd is resolved 09:42:30 next topic pls 09:42:55 Nico[m], Sefriol this has nothing to do with 'agile' 09:43:04 Nico[m], meson is 0.50.1 in sailfishos gitlab. 09:43:09 also, agile is kind of just a buzzword these days 09:43:27 Venemo As I said :D 09:43:48 But yeah, I'm fine moving to next topic 09:43:58 dcaliste_: Hm, I think I need 51 or so, anything I can do to update it in the future? 09:44:05 1 min 09:44:50 Nico[m], Propose a MR in sailfishos updating the version. It's using submodule, so update in the submodule directory and change the version in spec file. 09:45:01 Then, of course, check that nothing breaks... 09:45:16 dcaliste_: Thanks, I will do that :3 09:45:21 https://git.sailfishos.org/mer-core/meson 09:45:30 I'll give this 1-2 more minutes as the previous topic was cut short 09:45:49 Nah, I'm ready to move on 09:45:57 ok 09:45:59 #topic Alien Davik support for community devices (15 min – asked by Sefriol) 09:46:02 Here is the commit done by mal some months ago to move to 0.50.1 : https://git.sailfishos.org/mer-core/meson/commit/d10ac625cc090a65ff3a1adf36323b238849f594 09:46:10 #info Ongoing topic as well. Would Jolla be interested in doing a blog post which would clarify the current situation (What AD and Android support is, what does it consist, how it is licenced) and what could be done by the community (either by some users with NDA or in general, for device specific ports or Android version specific). "Tech and Strategy behind AlienDalvik" or something similar would be really interesting. 09:46:23 And the answer: 09:46:28 #info We are looking for possibilities for providing Android App Support for wider audience. The new Android App Support is more flexible on this sense and allows us to provide it with smaller effort for non official ports as well. As long as the device adaptation full fills certain requirements. This requires still some effort and we are not yet there. 09:47:24 dcaliste_: Thank you for all the links, that helps a lot! 09:47:46 the current AD is snapdragon centric afaik 09:48:11 Maybe simply unite and use Anbox? 09:48:38 bionade24: that is not "simply" done 09:48:48 ad is proprietary 09:48:58 anbox missing the sfos integration 09:49:08 Is AD still licensed from a third party? Could it maybe be opened up a bit? 09:49:16 if I had 20 cents each time I heard "simply" since I'm working in the software field, I would not need to work anymore because i'd be a millionaire 09:49:18 leszek: Of course, but we don't have to develop similar things all the time. 09:49:18 AD is a nice feature ... but if SFOS is supporting the functions direct like (VoIP, a.o.) then AD would be not so important because we get more native apps 09:49:38 * ApBBB wishes android disappears and more native stuff 09:50:02 pm-guest: That requires a lot of effort though! 09:50:10 Nico: Unfortunately Jolla doesn't seem to be eager to open up more of their "strategic asset" code. 09:50:27 Maybe they can't 09:50:41 For my part, I only need 2 android apps, one which I will replace soon myself (Matrix client) and the other one is the browser... 09:50:41 Android Support is so advanced topic that I think it should be covered in a blog post. 09:50:49 It has so many aspects to cover.. 09:50:51 From what I can see the integration part into the OS regarding wifi, bt and other stuff has been not much done 09:50:57 * schmittlauch[m] points at the effects of the PSD2 regulations @ApBBB 09:51:16 Venemo: Let's start putting €1 aside every time we hear "simply" ;) 09:51:24 Nico[m] But when will we start ... before or after the right time? 09:51:31 Nico: for browser you will be able to use angelfish starting from sfos 3.3. (assuming you are on official device) 09:52:36 rinigus: great work. You also get a lot of praise from the plasma mobile community 09:52:49 any tips about some minimal requirements ? overlayfs ? some defconfig settings ? 09:52:50 rinigus: That sounds great, although I actually like the default browser, but it is just incompatible with some websites. But if anglefish works well, I will finally uninstall Android support! :D 09:53:57 Nico[m]: works for ff-gecko 52 upgrade have begun afaik. Though not sure if early april joke because ff 52 is EOL since 2018 09:54:07 thanks. I tried to move angelfish more towards sfos browser design. cannot fully, as we have to be consistent with plasma, but in part it should be there 09:54:17 pm-guest: I'm sure people are already working on that stuff, but there are only so many hands. And there are also a lot of other issues, that need work, to bring such advanced features up. 09:54:43 leszek: Oh, that sounds great and a bit scary :D 09:54:48 leszek It's step for building newer version. not intemding to release that version 09:54:59 I see. Good 09:55:07 5 min left. 09:55:37 let's also include "just" and "only" besides "simply" 09:55:40 vknecht: min requirements are kernel >= 3.5, as far as we learned so far. see sfos flatpak project page, main repo 09:56:00 Venemo: We're gonna be filthy rich :D 09:56:04 Nico[m] ... that's the job to solve problems to build a better SFOS. Waiting is a solution to loose market shares. 09:56:29 market share never was the target 09:56:56 Jaymzz Do you feel a blog post about the Android Support is good idea or are we too far from AD being available for other devices that it's waste of time to talk about it? 09:57:02 rinigus, asked about community Android App support ;-) but infos about flatpack are welcome too 09:57:41 :) 09:57:48 leszek: That' not a fun project anymore ... people are working and living for the project. So the market is the level to make income. 09:58:03 For many who are interested in Sailfish, lack of AD support is pretty much the roadblock that drives many away 09:58:05 pm-guest: Jolla needs to make enough money to feed its developers and can't just employ an infinite amount of them to work on features. There simply is a limit to how much a small team like Jolla can work on. (I hope that use of simply was appropriate :D) 09:58:05 Sefriol: I think we should have it concretely in place first before a blog post is due :) 09:58:13 Since Xperias are not the most popular of devices 09:58:43 vknecht: getting compatibility with container features would be in common there 09:58:44 xperias are really nice devices 09:59:00 2 more minutes left on this one 09:59:21 Jolla should really start talking with Pines T.Lim to get some millions for more devs :) 09:59:24 abranson, sounds good 09:59:48 1 min 10:00:32 So, time's up. Moving on in a few seconds 10:00:53 #topic Add aptx, aac, ldac bt support (15 min – asked by Leszek) 10:00:55 just a comment regarding AD: I don't think I would have actively worked on many projects if I would have AD installed on my device 10:01:05 #info Bluetooth headphones get more popular by the date. SailfishOS should support current technologies like AptX, AAC and LDAC. There is a possibility to do this with this pulseaudio modules: https://github.com/EHfive/pulseaudio-modules-bt/ 10:01:05 The only thing I see that would need discussion besides the technical solution is the licensing. I am not sure if its possible to support AptX and LDAC in this way without acquiring a license. Though at least AAC and better SBC could be supported this way. 10:01:24 #link https://github.com/EHfive/pulseaudio-modules-bt/ 10:01:48 Answer coming up 10:01:51 #info The implementation is already a fork from mainline PulseAudio, which shows that upstream is not willing to take the implementation, probably mainly due to licensing issues. The same applies with us, we cannot ship these without license, which we do not have. I haven't looked at the latest developments with the fork but probably just taking SBC improvements would not be that simple. As our own (PulseAudio fork) bluetooth 10:01:51 modules are already diverging quite a bit from upstream, it wouldn't be so easy to for example just add the extended bt modules to the side as they most likely wouldn't work. So it would mean the two forks (ours and extended-bt) would need to be merged to get working implementation on Sailfish OS. That doesn't mean it isn't possible of course, both sources are out there. 10:02:29 a note about "technical solution" .. while the EHfive provided pulseaudio bt modules are really nice to have, they're also non-upstreamable, and the author doesn't seem to be willing to work with upstream to get them merged. 10:03:01 upstream bluez 5.52+ actually already has the required bits included, to allow supporting multiple a2dp audio codecs, and switching between them 10:03:10 do we need them to merge? I mean its a module for a reason 10:03:28 AAC should not be a licensing issue afaik. Even AOSP has support for it 10:03:36 but the pulseaudio bits are *still* being discussed by the developers, and not yet merged. developer called Pali is trying to get the pulseaudio bt codecs support merged. 10:03:48 At least we could try compile them for SFOS and see what happens 10:03:54 so yeah, that's the "upstream situation" 10:04:50 leszek: think AAC still needs licensing for encoding 10:05:14 TJCs related to this: https://together.jolla.com/question/11008/bluetooth-apt-x/ , https://together.jolla.com/question/201820/ldac-support-for-better-audio/ , https://together.jolla.com/question/216872/update-sailfishos-bluez-to-version-552-for-multiple-a2dp-codecs-support/ 10:05:22 and that is not already in sfos? I mean aac encoing? 10:05:32 #link https://together.jolla.com/question/11008/bluetooth-apt-x/ 10:05:35 leszek: yeah trying those modules on sailfish would be good! 10:05:41 #link https://together.jolla.com/question/201820/ldac-support-for-better-audio/ 10:05:45 leszek: yeah, but not in ffmpeg. it's complicated. 10:05:51 oh 10:05:53 #link https://together.jolla.com/question/216872/update-sailfishos-bluez-to-version-552-for-multiple-a2dp-codecs-support/ 10:06:37 AOSP btw includes LDAC codec out-of-the-box these days 10:06:51 which is the high quality preferred codec 10:06:53 I mean it could be also a community thing for now. Not shipped officially but just being compiled on sailfish-build and then land on openrepos but yeah it needs ffmpeg with aac 10:06:59 is that the one for sony headphones? 10:07:09 yes 10:07:13 abranson: well, not just sony 10:07:24 abranson: originally developed by sony :) 10:07:26 sony invented it 10:07:49 that doesn't need licensing iirc 10:07:50 most of the pro/hifi bluetooth has started supportin LDAC 10:08:11 libldac is included in fedora for example 10:08:13 so should be fine 10:08:33 I guess it could be in openrepos or something 10:08:40 yes ldac is fine 10:09:05 the pa modules could also be compiled without aac I think. Only ldac and improved sbc 10:09:42 would be cool if someone at jolla could test compiling it on mer-build and test it 10:09:57 there are opensource apt-x / aptx HD codec libraries these days aswell 10:10:00 even in ffmpeg 10:10:03 5 min left on this 10:10:08 it's merging the EHfive changes into the sfos fork that's the tricky party 10:10:25 abranson: why? 10:10:42 I compiled it against normal debian pa without patching the pa 10:10:47 and it works 10:11:05 let's hope the support for multiple bluetooth a2dp audio codecs gets finally merged to upstream pulseaudio in the next months.. 10:11:14 leszek: in the answer above - you can't just build the EHfive version. there are changes in the sfos fork that need to be merged in. 10:11:15 I think Jusa would know more, but I don't see him here now 10:11:34 * schmittlauch[m] suppresses the urge to say mean things about the MPEG LA 10:11:34 pasik: source code openness and required patent licenses are 2 different things 10:11:41 schmittlauch[m]: of course 10:11:50 pasik: yeah that's a big prerequisite 10:12:16 anyway. It would be bad if in one year nothing changes on that front and bringing this up here was just a waste of time 10:14:22 So moving on then 10:14:34 #topic general discussion (15 min) 10:16:23 Anuthing to bring up? I see a lot of people dropping out and coming back in the room 10:16:28 anything* 10:16:33 qt and gcc aside many parts of sfos start to show their age or be plainly bad. the camera interface for example. and performance doesn't seem to be good. i mean i use a phone that can run fortnite on it and SFOS feels kind of sluggish 10:16:34 just as note. QtWebkit 5.6 that Jolla ships is full of security and usability bugs, broken css3, javascript and others. I decided for the next webcat upgrade to warn users about that. As Jolla refuses to upgrade to Qt 5.212 and I see no coperation on this matter this is the first step I will take 10:16:47 Next step will be blocking all old QtWebkit users 10:16:55 Was Jolla suppose to participate to MWC? 10:17:15 QtWebkit as a whole is AFAIK obsolete, in favour of Qt WebEngine - correct me if I'm wrong 10:17:23 Jaymzz: got a connection error( that was the reason. 10:17:44 Pekkeli: Yes but with no booth. A few sailors were supposed to be there to have meetings etc 10:17:57 so, we intend to move away from QtWebkit, but it's kind of too early to boast about any details on this. 10:18:03 Venemo: QtWebkit 5.212 exists though that is not full of security nightmares and bugs 10:18:25 Ok, so there was not suppose to be any press event by Jolla? 10:18:30 Venemo: I brought this up 2 years ago also 10:19:25 QtWebEngine would need a api as the upstream Qt one is abysmal 10:19:46 any news on lipstick getting an overhaul or something?? 10:19:51 leszek: so, about 5.12 - see my previous message above. I would personally love to use 5.12, about QtWebkit - I think we don't really use it in any critical places, do we? still, like I said we would like to move away from QtWebkit completely, eventually. 10:19:57 niko[m] and others - May Jolla has less devs for features but if a feature like VoIP-receiving calls is working and in the next release not is not very comfortable. 10:20:04 ApBBB: what kind of overhaul would you be interested in? 10:20:25 Venemo: accounts and email uses qtwebkit 10:21:01 upstream firefox (75+) has recently gotten vaapi/ffmpeg hardware accelerated video decoding/playback on wayland, which is nice. doesn't directly help sfos, I know. 10:21:02 rinigus needs it to support xdg shell latest. and performance. interacting with sfos feels slow. 10:21:06 Little update about Xperia X aosp9 based port: 10:21:07 https://build.merproject.org/project/show/home:vknecht:hw:sony:loire 10:21:07 http://merproject.org/logs/%23sailfishos-porters/%23sailfishos-porters.2020-03-18.log.html#t2020-03-18T19:13:52 10:21:29 main problem is booting right every time 10:21:37 pasik: that's good for the pine! 10:22:41 jaymzz, where can I find log for this channel ? 10:22:50 Venemo: problem is that QtWebkit 5.6 is still used and users of my webbrowser complain a lot about broken websites and crashing app. 10:22:59 A webbrowser is a critical app 10:23:06 It's not listed in http://merproject.org/logs/ 10:23:27 dcaliste_: I'll send out links after the meetng is over :) 10:23:41 Ah yes, sure, thanks. 10:23:47 leszek: our web browser doesn't use it 10:23:48 Anytime! 10:23:53 dcaliste_: https://irclogs.sailfishos.org/meetings/sailfishos-meeting/2020/ 10:24:02 Venemo: I know. But mine does. 10:24:05 Thanks mal. 10:24:13 leszek: what can I say? 10:24:32 I don't know. It should've happen 2 years ago. The upgrade to 5.212 10:24:35 5 min left 10:24:41 leszek: 5.212??? 10:24:43 also accounts and email suffer from it aswell 10:24:52 Venemo: yes 5.212 10:25:04 you mean 5.12 10:25:09 no 5.212 10:25:17 what? 10:25:42 is QtWebkit versioned differently than the rest of it? 10:25:43 Venemo (IRC): https://github.com/qtwebkit/qtwebkit/releases 10:25:48 okay 10:25:49 Venemo: https://github.com/llelectronics/lls-qtwebkit 10:25:50 its qtwebkit 10:26:38 well, like I said, we would like to move away from QtWebkit, so I would suggest you to do the same for your app, too. there isn't really a good reason to invest in it anymore since it is obsolete 10:27:26 5.212 is in active development 10:27:36 and updated just a couple of days ago 10:27:40 Venemo: but there is no harm in updating qtwebkit to the latest version on sfos 10:27:47 3 min 10:27:55 any news on the fingerprint for community devices?? 10:28:02 now that i remembered 10:28:06 its surely better than the current one, if I understand leszek correctly 10:28:14 miles better 10:28:26 css3 works, js is not shit and way less security issues 10:28:30 leszek: maybe you should just make MR at sfos repos with webkit update 10:28:31 also IIRC QtWebeingine has pretty crazy depndencies and build requirements - there are issues with keeping it building even on Fedore quite often 10:28:31 ? 10:28:37 how do you justify putting effort into something that we don't actually use? 10:28:47 rinigus: did it 2 years ago 10:28:54 no one was interested 10:29:05 try to ping it there 10:29:13 Venemo: you use it. Like I said in email and accounts it is used 10:29:23 Venemo: if we use it in email then its there 10:29:33 leszek: ok so you're saying that you do have an MR for upgrading to the version you desire, but we haven't responded to that yet? 10:29:57 email app also needs love 10:29:59 Venemo: exactly. No one responded. 10:30:01 using such an ancient QtWebkit for parsing arbitrary html mail does not seem very secure... 10:30:16 Venemo: it was brought up in multiple meetings before aswell 10:30:23 not to mention the abysmal performance for more complex html emails 10:30:28 it's not true that no-one was interested. you might not have got a response, but the amount of work required to assess that was too high for anyone to properly assess it. 10:30:40 it's not an MR. it's a replacement branch 10:30:42 I'll give this a couple more minutes. But try to wrap it up as we are out of time (and I'm sure most of the Finnish sailors are already having lunch) 10:30:45 Venemo: though the MR is outdated nowadays surely. I did not bother after 2 years of hearing no feedback 10:30:47 https://git.sailfishos.org/mer-core/qtwebkit5/merge_requests/12 10:31:57 abranson: what else can I assume when no one contacts me or shows interest? 10:32:29 maybe we should take a risk and update qtwebkit? or should we commit and test on our devices and report back? 10:32:36 as leszek has it on openrepos, I think 10:32:43 abranson: can we get that accross to the relevant people? 10:32:44 exactly 10:33:40 Venemo: I asked around two years ago, but it's really hard to assess, and high risk as it's used in accounts and email. 10:33:43 Let's wrap up :) 10:33:56 abranson: how about after we stop using it in those places? 10:34:00 note that the 5.212 version is marked alpha on github too 10:34:11 abranson: its an higher risk to iuse the outdated qtwebkit version in mail and accounts 10:34:22 Venemo: if/when that happens, we'll have a better alternative to webkit ;) 10:34:37 moving on in 30 sec 10:35:00 leszek: security risk, not a risk of people not being able to read their email or sign up for accounts at all. 10:35:14 #topic next meeting time and date (5 min) 10:35:24 #info Next meeting will be held on April 2nd 2020 at 09:00 UTC 10:35:27 abranson: you had plenty of time testing that the last 2 years with my qtwebkit version 10:35:41 I did mention how hard it was to assess as a new branch at the time 10:35:54 #info if the current situation continues with the Covid-19, we may have to postpone the next meeting. We'll keep you updated. 10:35:56 what do you mean, plenty of time? 10:36:18 I think he means that we had 2 years to test it by now 10:36:24 hmm 10:36:43 Jaymzz: it's an online meeting, why is it affected by the covid craze? 10:36:45 abranson: the code is there since 2 years. Updated by me all the time. Testing that against accounts and email is done by thousands of people already that did not had a problem using my qtwebkit. Jolla could've done the same 10:37:14 Venemo: Sailors may need to concentrate their time and effort on other projects. This comes from the management team :) 10:38:11 Alright ending the meeting 10:38:15 Stay safe out there 10:38:16 I see 10:38:17 okay 10:38:21 Till next time :) 10:38:21 Venemo: yeah - I'm also WFH due to that right now and yxou can't even leave your house in Czech Republic right now without a mask or respirator, not to mention closed borders schools and shops 10:38:28 leszek: it's about balance. it's a minor part of sfos, but not one we can afford to break. like Jaymzz said - we have to be careful with what we spend time on 10:38:39 M4rtinK: you can tune into an online meeting from home can't you 10:39:05 #endmeeting