17:00:09 #startmeeting Sailfish OS, open source, collaboration -- October 29, 2026 17:00:09 Meeting started Thu Oct 29 17:00:09 2026 UTC. The chair is rainemak. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:09 Useful Commands: #action #agreed #help #info #idea #link #topic. 17:00:10 #info Meeting information and agenda can be found here: 17:00:10 #link https://forum.sailfishos.org/t/35051 17:00:10 I am the meeting's chairperson today, and will be doing my best to keep time and order. Please respect the timings and bee-hive. 17:00:11 #topic Brief introduction (5 min). Please prefix your name/handle with #info 17:00:13 #info Raine Mäkeläinen, Jolla 17:00:22 #info Otto Mäkelä, community 17:00:28 #info Ville Nummela, community 17:00:37 #info Balázs Szulovszky, community 17:00:50 #info yusssufff, community 17:02:17 #info CLMA31, community 17:02:46 #info poetaster community 17:03:14 #info mal, Jolla 17:03:22 #info zhasha, community 17:05:18 #topic Tracking of non-resolved Community Meeting topics 17:05:18 #action Community task board 17:05:18 #link https://forum.sailfishos.org/t/community-meeting-on-23rd-april-2026/29030/3 17:05:18 #action Jolla to update regarding recommended charging cables and chargers. 17:05:18 #link https://forum.sailfishos.org/t/community-meeting-on-7th-may-2026/29253/2 17:05:19 #action The Privacy Switch in relation to Ambiances 17:05:21 #link https://forum.sailfishos.org/t/community-meeting-on-july-16-2026/30378/7 17:05:23 #info No actions taken yet regarding community task board nor charger / cables. 17:06:32 let's go 17:06:38 #topic Let’s talk about PIM capabilities. (15 mins -- usbee) 17:06:38 #info It is a bit complex topic but I feel like they are heavily 17:06:38 #info connected and it is worth to discuss it as a whole package. 17:06:39 #info Supporting Oauth doesn’t work by most providers. As a result it is 17:06:40 #info not possible to sync IMAP, CalDAV and CardDAV inside a singular 17:06:42 #info account. So setting up Fastmail for example needs to be done 3 17:06:44 #info times as 3 separated entries. Also syncing notes and todos is not 17:06:46 #info possible at all. Even so there is no ToDo app as part of Sailfish 17:06:48 #info OS. There is a notes app but it only works as an offline local 17:06:50 #info app. There should be an option besides generic IMAP and generic 17:06:52 #info CalDAV/CardDAV accounts for generic Oauth accounts too. I do 17:06:54 #info believe that this set of features is essential on a phone to be 17:06:56 #info able to use it as a PIM device so having a consideration and a 17:06:58 #info timeline for it would already be a great thing. 17:07:00 #info Thank you, let's have an open discussion. 17:07:34 Originally this topic was huge, thanks for cleaning it up 17:07:42 I think I was able to summarize the issue as clear as I was able to do. 17:07:58 To me this is a critical weak point in SFOS. 17:08:10 I think many of us have defaulted to using nextcloud as the central point for pim 17:08:30 noted, however this meeting shall not become item prioritization meeting 17:08:47 I picked Fastmail (over Protonmail) specifically to be able to have proper WebDAV support. The problem is not on their side. 17:08:50 poetaster, I have seen that happening as well 17:09:22 It's difficult for me to make a contribution since I run both my own nextcloud AND mail services. 17:09:27 @rainemak Okay, I understand. 17:10:00 I guess I'm the counter data point here. I've only ever used generic IMAP since getting my first Jolla back in early 2014 17:10:54 Separating CalDAV and CardDAV would already be an improvement because the current method is very alien even for the average Linux user who is socialized on Thunderbird or Evolution. It's also broken with several providers like Fastmail as I observed. 17:11:38 We should also consider grouping certain core functionality like this as a company. Maybe it could be a nextcloud instance provided by Jolla who knows. 17:12:01 but it's not only about PIM. 17:12:12 That could work, though I don't want to change my e-mail address once again, it's a long and painful road. :D 17:12:37 how would generic OAuth account know how and where it is supposed to connect for mail, callendar, contacts, etc.? 17:13:10 rainemak, that is the value add that google and apple use to tie you to their platforms :) 17:13:19 usbee, understand that but simpler to help / ease changing from big tech 17:13:30 Well, there should be a database for that. Thunderbird solves that without having to select provider manually, it even auto detects if the provider uses Oauth or not. There is no extra user input needed there. 17:14:10 poetaster, true that and there's also value when using different service provider for different purpose 17:14:14 Yup, I see that as an added value as well, just like how Thunderbird went with Thundermail too. 17:14:35 It's just very unlikely I will ever change my address unless my provider dies or something. 17:15:19 usbee, that is why I do my own services :) But, the question is, is this about a SINGLE point to coordinate the services or easier 'on-boarding' 17:15:34 I also run my own email server, caldav server, carddav server and nextcloud servers. That makes three accounts in my phone. I've never really seen that as a huge issue. 17:16:24 With Google it is possible to have all with a single login. But do we want to convince people to keep using Google? I doubt that. For now that's the most convenient provider on SFOS. 17:16:24 ViGe, exactly. But DKIM signing is not everyone's cup of tea. 17:16:58 usbee, you mean mail, calendar, notes, contacts ...? 17:16:59 Maybe a bit of a chore for people who don't want to run their own services, but then again, I wouldn't say SFOS is for the non-technical people anyway. 17:17:15 @poetaster And yup! 17:17:18 @poetaster yup! 17:17:26 (Generating DKIM is not that involved with modern software packages and all that) 17:17:32 Notes is missing though, but that's a different story. 17:18:12 Would be cool to get notes in Nextcloud to visible in native Notes app 17:18:20 I guess .well-known could be used to query provided/supported services? 17:18:33 There are some other little things too, for example if I create/delete/modify a contact, it won't trigger a sync, I have to wait for the next scheduled sync for it to be uploaded which is kinda weird. 17:18:57 Keto, always good! ok, the 'short term' solution is a nextcloud / imap service. Who'll host :) 17:18:59 @CLMA31 Or any WebDAV notes in that regards. 17:19:54 That's true 17:20:13 time is almost up 17:20:24 I think we got the point though 17:20:30 Yeah, okay. 17:20:37 Thank you usbee 17:20:42 There are plenty of room for improvements, yes. 17:20:47 Maybe a sailmates project. collective nextclound/imap 17:20:54 #topic Is it time to tell us about Jolla’s plans to update QT? (10 mins -- PeegeeTips) 17:20:54 #info The topic itself is not new; we all know that QT5.6 is 17:20:54 #info completely obsolete and that there are ‘reasons’ why Jolla 17:20:54 #info hasn’t redeveloped components like silica using a current 17:20:54 #info version of QT6. To a degree this didn’t matter as long as 17:20:55 #info Jolla (and SailfishOS!) was on life support still eking out 17:20:57 #info an occasional minor update on the aging Xperia 10 hardware, 17:20:59 #info but this is not where we are now... Right now, we’re all in the 17:21:01 #info happy position of looking back at a successful launch of new 17:21:03 #info Jolla hardware, so I am being forced to look forward to 17:21:05 #info consider what a healthy SailfishOS ecosystem look like? I 17:21:07 #info don’t believe that looks like asking prospective app 17:21:11 #info developers in 2027 to get excited about using an obsolete GUI 17:21:13 #info toolkit from eleven years ago. One that bears little relation 17:21:15 #info to tooling used right now to create their desktop and desktop 17:21:17 #info apps. So, what’s the plan? Update QT6 / Adopt GTK / Something 17:21:19 #info else... 17:21:21 #info We update when we have something to update, sorry. Last time we 17:21:23 #info discussed about PWAs. I'd proposed that we would use this topic to 17:21:25 #info discuss about potential toolkits. What other toolkits would you 17:21:27 #info like to see? 17:22:12 Flutter (ducks'n'runs) 17:22:35 I'd like to see Kirigami app being functional on SFOS, so apps made for Plasma Mobile. For now plenty dependencies are missing for it. 17:23:01 usbee, I'd like to avoid kde library dependancy hell at all costs :) 17:23:21 I don't have technical knowhow one bit (and don't know any toolkits), but in Jolla's eyes: is it easier to do the whole Qt upgrade or get rid of Qt all together? 17:23:37 Since KDE is also Qt based it would be nice to see some cooperation. I would even like to see a SailfishOS version shipped with Plasma Mobile. On desktop it's normal for a distro to have multiple DE choices. 17:24:06 You're looking for Mer 17:24:53 @poetaster Well, whether you like it or not there are plenty of native mobile Linux apps that are not functional whether they are GTK or KDE apps, because of lacking libraries. It is significantly easier to run the Android version of those apps if they exists than the native one. 17:25:01 usbee, I contributed some to rinigus and piggz efforts with 5.15 and the whole kde ecosystem. While there are 'benifits' I personally won't go there again since I don't want any of the kde apps. 17:26:03 @zhasha I'm looking for a mobile distro with Android AppSupport or equal and DE options. There aren't any since Waydroid is pretty inferior to AAS in my eyes. 17:26:36 The support overhead for something like that would be enormous 17:26:47 Ok, we're drifting. Jolla, QT6 plans outside of piggz & rinigus efforts? 17:27:43 I apologize before hand my technical lag of understanding and stupid question... But would it be possible to natively ship something what piggz and rinigus have pulled of so app development could direct more towards Qt6? 17:27:43 and then we have gtk which could be an option 17:27:55 switching toolkit would be a lot bigger effort than qt update 17:28:07 Thanks mal! 17:28:30 more like adding another toolkit as an option 17:28:31 yeah, let's stick to QT please. 17:28:56 I'm thinking more like application developer point of view 17:29:00 rainemak, did you see the experimental flutter hello world? 17:29:35 https://github.com/mrcyjanek/flutter-sailfishos 17:29:45 thanks, will check as well 17:30:51 Would it be possible for Jolla to ship the Qt6 packages alongside the current once so apps could direct to that? Similarly like currently is achievable through piggz and rinigus effort 17:32:01 CLMA31[m], I'd rather not start discussing technical possibilities here as that ends up being endless 17:32:23 Thanks for the answer 17:32:36 CLMA31[m]: well you also need to take into account disk space usage 17:33:08 But to me that would seem like something that possible could be done without throwing huge amounts of money to Qt? 17:33:37 Sure! But with Jolla phone and 256gb I would imagine it to be feasible 17:34:09 It also could be optional, like AAS or MicroG and other stuff. 17:34:15 Thank you all for your input... let's move on 17:34:28 well we have the separate root partition which causes some more thinking 17:34:51 #topic Softening a specific Harbour restriction (10 mins -- yusssufff) 17:34:51 #info Add a “call” class to classify streams. Until now only the 17:34:51 #info game and player call are available: Specifically, currently 17:34:51 #info libaudioresource only offers player and game. A call type in 17:34:51 #info libaudioresource should help any store app doing voice or 17:34:53 #info video calls. 17:34:54 #info 17:34:56 #info Devs are working on getting video calls to work with the 17:34:58 #info native telegram’s client Fernschreiber. There is a will to 17:35:00 #info have the calls and video calls done in a Harbour-compatible 17:35:02 #info way. They identified that to be published in Harbour, the 17:35:04 #info “call” class should be available in libaudioresource. You can 17:35:06 #info dig in here: Add Telegram voice & video calls (tgcalls + 17:35:10 #info tg_owt) by JimKnopfIoT · Pull Request #583 · Wunderfitz 17:35:12 #info /harbour-fernschreiber · GitHub 17:35:14 #link https://github.com/Wunderfitz/harbour-fernschreiber/pull/583#issuecomment-6254871945 17:35:16 #info This is a rather broad topic. Raising a "call" class is not 17:35:18 #info necessarily good way to approach this problem statement whilst 17:35:20 #info "call" policy could be one. We should instead talk about what's 17:35:22 #info missing from the Harbour so that it'd be more attractive for 17:35:24 #info developer, users and at the same time it would be a trusted source 17:35:26 #info for app installation. 17:35:28 #info 17:35:30 #info Please also note that this kind of changes tend to take time 17:35:32 #info whilst being important item. 17:36:38 Allright. And why call "policy" would be better approach than "class"? 17:37:01 yusssufff, you're getting it wrong 17:37:11 policy == class 17:37:14 same same 17:37:22 oh okay. vocabulary :) 17:37:44 "call" won't make a rain 17:38:58 having that 2nd binary delivered by a harbour app is a good starting point 17:39:01 It sounds good anyhow. To answer what would be "missing" in Harbour, well a way to do a video call with an app sounds quite a reasonable expectation? I don't know if Telegram makes it easier or harder, but no devs managed to package video (and yet nor audio) calls there yet 17:39:02 Again, apologies for my non existing technical knowledge, but my understanding is that if one day Whisperfish lands in Harbour it won't have voice recording as that brakes some Harbour rule. I don't know how to classify this, but audio recording, VoIP calls, video calls, fingerprint API etc. would be nice additions 17:39:22 rainemak, is the reason for the restrictive policy to protect incoming phone calls? 17:40:00 for example, we need figure out a way how apps can be granted ipc calls -> how to tell it to end users 17:40:04 I'd been wondering about this since I built a hack screen recorder with audio support but won't release it since I don't know what harm it might do to calling :) Probably none :) 17:41:00 CLMA31[m], not quite correct 17:41:15 rainemak: From end user POV, could it just be another permission which is listed at first launch of app? 17:42:04 CLMA31[m], not quite that simple but half way correct 17:42:58 unfortunately, I don't have good answer now whether we could just go and add "call" audio routing policy 17:42:59 @ranemak, so you suggest Fernschreiber devs for now to publish a 2nd binary as an approach to calls? 17:43:05 as allowed 17:43:07 Yeah, I'm very much aware that I probably understand 5% of technical stuff talked regarding Sailfish :D 17:44:23 yusssufff, nothing to do with the "call" class -> just an example that opens up new possibilities 17:45:02 maybe some voip call policy? to separate it from the actual calls, not sure if that makes sense? 17:45:26 What little I understand that would make sense 17:45:27 mal, I think it's almost like that under-the-hood 17:45:52 As a end user I would like to know whether app makes actual calls or voip calls 17:46:00 let's say Signal calls from AppSupport requires it 17:46:46 it == something like mal elaborated above 17:47:03 I don't quite get why it's so complicated. It's three lines of c++ all related to the policy/class which modify routing behaviour. 17:47:19 #action Jolla to check whether "call" or similar could be allowed in Harbour 17:48:25 points are taken 17:48:27 moving on 17:48:29 #topic Getting the community work on the Xperia 10 V’s cameras merged in @Jolla ‘s build (5mins – yusssufff) 17:48:29 #info Devs are working on getting Xperia 10 V Camera’s to work with 17:48:29 #info the blobs version 4b. Most of the work actually doesnt imply 17:48:29 #info messing with the binaries themselves (exept eventually for 17:48:29 #info telephoto aka 0.6, but telephoto could be left on the side, as 17:48:31 #info quite a non-essencial lense). Could @mal maybe have a look at 17:48:33 #info incorporating that work into the official build? Even fast, as 17:48:35 #info a quick hotfix for the 5.2 release? Having cameras now working 17:48:37 #info on the 10 V would makes a huge instant change in usability, 17:48:41 #info therefore the faster the better... Here is the link of the 17:48:43 #info community work: [GUIDE] How to enable all cameras on Xperia 10V 17:48:45 #link https://forum.sailfishos.org/t/guide-how-to-enable-all-cameras-on-xperia-10-v/35267 17:48:47 #link https://forum.sailfishos.org/t/xperia-10-v-zambezi-enabling-all-cameras-hardware-video-on-sailfish-os/35811 17:48:50 #info At some point we could look it this. Meanwhile, you can certainly 17:48:52 #info contribute back. We are very much busy with Sailfish OS 5.3, Jolla 17:48:54 #info Phone and Commodore Callback. 17:49:35 Allright. so i guess Github PRs are the way to go? 17:49:56 mal, do you have better ideas? 17:51:31 I have only so far had a quick look at those changes, I do wonder about that part where the instruction copy some odm camera files, I didn't quite follow what that does 17:51:32 MMh the thing is the two approaches taken in both threads are somehow quite different... I dont know exactly which one would fit the Jolla's rights ... We arent very familliar with what is "allowed" to play around with Sony's side of the things... 17:53:37 parth007 's way seems to need to download sony's firmware from the phone itself for exemple. 17:54:35 possibly balta's approach would be easier for Jolla to implement, from what i could judge...But hard to say from outside of Jolla... 17:55:33 we could actually continue this same topic in the next topic 17:55:49 as Parth007's topic is quite similar 17:55:53 if I understood correctly balta's version for main camera only needs kernel patch, that is doable 17:55:54 Sure. 17:56:11 mal : yeah i think so too 17:56:26 even just main camera would be an improvement 17:56:35 indeed 17:56:42 #topic Xperia 10 V (5.1.0.11) – community fixes for SIM loss after boot, slow boot, mic and fingerprint (10mins – Parth007) 17:56:43 #info I’ve been debugging the Xperia 10 V on 5.1.0.11 and have 17:56:43 #info working fixes, all in one package: 17:56:43 #link https://forum.sailfishos.org/t/guide-xperia-10-v-faster-boot-mic-sim-recovery-fingerprint/35340 17:56:43 #info 17:56:45 most people only use that anyway 17:56:45 #info 1. “No SIM card” after some boots: ofono crashes (SIGSEGV) ~60 17:56:45 mal ! yesss!!! and a bit one! 17:56:47 #info s after boot during SIM setup, right after “session_read_info 17:56:49 #info not implemented”. The restarted ofono sends setRadioPower 0 17:56:51 #info while the SIM is already active, so the modem powers the card 17:56:53 #info down (uim POWER_DOWN, card error 0x17). Restarting only the RIL 17:56:55 #info (vendor.qcrilNrd / qcrilNrd2) brings the SIM back; I use an 17:56:57 #info automatic watcher for that now. ofono also segfaults when 17:56:59 #info simply restarted. Is this known, and can it be fixed in ofono / 17:57:01 #info ofono-binder? 17:57:02 *big 17:57:03 #info 17:57:05 #info 2. Slow boot: bluebinder.service is Type=notify, so boot waits 17:57:07 #info for Bluetooth, and the BT HAL often fails chip init at boot. 17:57:11 #info Making bluebinder non-blocking cut a full reboot from 4:02 to 17:57:13 #info 2:23. Could this go into the 10 V adaptation? 17:57:15 #info 17:57:17 #info 3. Mic: primary_audio_policy_configuration.xml only allows 24 17:57:19 #info -bit for “primary input”. A 16-bit recording makes pulseaudio 17:57:21 #info -modules-droid loop until RLIMIT_RTTIME kills PulseAudio, 17:57:23 #info giving quiet or broken recordings and no sound until reboot. 17:57:25 #info Adding a 16-bit profile (plus +12 dB gain) fixes it. Can this 17:57:27 #info be shipped? 17:57:29 #info 17:57:31 #info 4. Fingerprint: the open-source HAL fails (“SetDataPath failed 17:57:33 #info rc=1”) with the current egisap TA. Sony’s stock HAL plus a 17:57:35 #info small shim to the et51x driver works (enrol, unlock, survives 17:57:37 #info reboot). Is official fingerprint support planned for the 10 V? 17:57:41 #info Thank you for your questions. Quite a technical set of questions. 17:57:43 #info Let's hope that we have somebody from Jolla for this. There looks 17:57:45 #info to be good pointers. 17:58:27 audio policy change is doable, just needs testing, bluebinder issue is annoying, not sure why it randomly fails 17:58:57 this is 10v specific? 17:59:14 I think so, probably some bt firmware issue or something 17:59:15 mal : if you or anyone had time to review those, and yeah the kernel patch to unlock the main camera following balta's work, that'd be really so cool. Those set of improovements would make it almost day and night for patient owners of 10 V... 18:00:37 modem issue needs more checking, could be some timing issue or wrong order of binder calls 18:01:41 btw, was there a chance to implement mal's work on 3.5mm jack unlock for the 10 IV/V? i couldn't hear yet improovements on that side with the 5.2, so that was maybe kept for 5.3? 18:01:57 fingerprint is a bit of a problem, that driver in open source is missing something and needs reverse-engineering it figure out what 18:03:22 I'll have a look at the audio policy configuration along with that audio recording issue once then worst busy times are over 18:03:55 tmo, any small set of improovement for daily usability soon-ish (main camera, sound and jack, boot issues) would be hugely welcome. Leaving aside for now harder nuts like fingerprint or telecamera would really not make much people sour i think 18:05:46 yusssufff: does bluetooth recover from that eventually? just to make sure the service that blocks boot won't keep restarting forever? 18:06:05 ah that different person complaining about that but still 18:06:33 mal: i think parth007's work got published quite recently, so not much "testers" yet 18:07:05 but I will do some testing on different devices to see how bt behaves with different service types 18:07:46 sound wonderful. Are the guides, such as balta's one, on the forum detailled enough for Jolla to step in, or shall the devs start PRs on github actually? 18:08:00 Let's see what could be done for Xperia 10 IV and 10 V. yusssufff do we have this topic covered? 18:08:19 yep all seems shared. 18:08:26 thank you 18:08:28 the bluebinder service change is simple to test but just needs many reboots to test on different devices 18:08:29 let's move on 18:08:39 #topic Open Pull Requests (PRs) to discussion (5 mins -- asked by Jolla) 18:08:56 https://github.com/sailfishos/sdk-harbour-rpmvalidator/pull/189 18:11:23 jeez. I get to distribute ffmpeg :) 18:11:44 just when I'm about to rebuild everything with libav :) 18:13:45 poetaster: what? 18:13:59 mal, ViGe 's rpmvalidator PR 18:14:16 ViGe, thanks for reminding 18:14:22 moving to general discussion 18:14:25 It's not mine, it's karry's 18:14:35 sure, but you linked 18:14:42 #topic General discussion (10 mins) 18:14:42 Thank you everybody who participated to Jolla Phone Roadshow Stockholm in person or via peertube 18:14:42 #link https://forum.sailfishos.org/t/roadshow-community-meeting-stockholm-oct-22/34246 18:14:58 Do the Jolla Phone has DP Alt mode support? 18:15:17 nope 18:15:24 Is that hw or sw limitation? 18:15:30 hw 18:15:33 errr, what is DP Alt mode? 18:15:34 I see okay. 18:15:44 Display Port Alt mode 18:15:44 DisplayPort over USB C practicallíy 18:15:48 yeap 18:16:11 but also sw would require changes 18:16:26 I mean multiple monitor support 18:16:30