============================================================================= #sailfishos-meeting: Sailfish OS, open source, collaboration -- 14th Sep 2023 ============================================================================= Meeting started by rainemak at 07:00:16 UTC. The full logs are available at https://irclogs.sailfishos.org/meetings/sailfishos-meeting/2023/sailfishos-meeting.2023-09-14-07.00.log.html . Meeting summary --------------- * Meeting information and agenda can be found here: (rainemak, 07:00:27) * LINK: https://forum.sailfishos.org/t/community-meeting-on-irc-14th-september-2023/16646 (rainemak, 07:00:32) * Brief introduction (5 min). Please prefix your name/handle with #info (rainemak, 07:00:40) * Otto Mäkelä, community (ExTechOp, 07:00:47) * David Llewellyn-Jones, community (flypig_, 07:01:03) * Raine Mäkeläinen, sailor @ Jolla, chairperson today (rainemak, 07:02:32) * Pami Ketolainen, sailor @ Jolla (Keto, 07:03:11) * JP Walden, sailor @ Jolla (jpwalden, 07:04:55) * Sebix, community (sebix[m], 07:05:42) * SailfishOS Edition of the Pinephone Pro (10 mins -- asked by piggz) (rainemak, 07:06:10) * I have been approached by Pine64, who would like to make available an edition of the PinephonePro pre-loaded with (rainemak, 07:06:20) * my SailfishOS port on the Pine store. I am more than happy for this to happen (with some caveats) and would like to (rainemak, 07:06:20) * know if Jolla would be against this in any particularly strong way. There is precedence for something like this, the (rainemak, 07:06:20) * “Jolla Devices” store sells sony xperias pre-loaded with the free editions of SailfishOS, and I would of-course not be (rainemak, 07:06:20) * shipping any of the components that typically require the SailfishX license. I see this good PR and advertising for (rainemak, 07:06:23) * Jolla and it will only make Sailfish more known. (rainemak, 07:06:26) * (the caveats are that i feel the Pro version of the sailfish os port is not as stable/usable as for the regular (rainemak, 07:06:28) * pinephone just yet, so i would like this to be made clear so as to not disappoint users, and hopefully it would bring (rainemak, 07:06:31) * in more developers) (rainemak, 07:06:34) * with Pine. If you have any conditions, then I would of-course be more than happy to do my best. (for example, DanctNix (rainemak, 07:06:41) * which is based on arch, does not use the name Arch so to be less associated with that project, would you want to not (rainemak, 07:06:44) * use the Sailfish name?) (rainemak, 07:06:47) * First, it’s great to hear that you were approached by Pine64 regarding your Pine Phone Pro port. There shouldn’t be (rainemak, 07:07:04) * major roadblocks in this. However, there are different aspects that we should clarify & check. We shared some details & (rainemak, 07:07:04) * questions to piggz. (rainemak, 07:07:04) * Please do not mix that webshop that sells pre-installed free editions of SailfishOS to this. As far as I know, it is the free version Sailfish OS by Jolla that is produced in Sailfish X program not a community edition. (rainemak, 07:07:14) * piggz community ... seems matrix isnt bridging here :) (piggz, 07:11:54) * regarding keyboard input issues in android apps (6 mins -- asked by thigg) (rainemak, 07:17:59) * LINK: https://forum.sailfishos.org/t/aliendalvik-aas-typing-characters-after-deleting-typed-text-often-inserts-the-deleted-string/14046 (rainemak, 07:18:12) * there are two issues with keyboard input in some android apps. One is that deleted text appears again and the other (rainemak, 07:18:22) * that after placing the cursor added text sometimes is inserted at the previous location. It looks like the issue is (rainemak, 07:18:22) * around for quite a while and I’d like to help to find the cause. Can you describe how the input is actually passed and (rainemak, 07:18:22) * if we have a chance from the community side to investigate into this? (rainemak, 07:18:23) * We don’t know exactly what the issue is, more investigation needed. We have seen the issue, but this hasn’t been a (rainemak, 07:18:57) * priority. There are some stub Android methods in our “RemoteKeyboard” component and those might be relevant. (rainemak, 07:18:57) * Easiest to help via the forum ticket. (rainemak, 07:19:22) * Updating the Linux kernel headers (10 mins -- asked by flypig) (rainemak, 07:23:02) * Currently the Sailfish OS repos contain kernel headers for Linux 3.18. Is there scope for them to be updated in future (rainemak, 07:23:12) * releases? The motivation is that Gecko ESR 91 requires the linux/membarrier.h header (specifically __NR_membarrier) (rainemak, 07:23:13) * which was introduced in kernel headers 4.3. Gecko checks at runtime that the capabilities are available before using (rainemak, 07:23:13) * them. With thanks to @mal for his help with this. (rainemak, 07:23:13) * Related links: (rainemak, 07:23:14) * 1. Upstream change that makes this useful: D88395 Bug 1661016: aarch64: Invalidate icache when compiling on a background thread; r=nbp,lth (rainemak, 07:23:16) * LINK: https://phabricator.services.mozilla.com/D88395 (rainemak, 07:23:19) * 2. Sailfish kernel-headers repo: GitHub - sailfishos/kernel-headers (rainemak, 07:23:21) * LINK: https://github.com/sailfishos/kernel-headers/ (rainemak, 07:23:23) * 3. Sailfish pre-build linux 5.4 headers: Show nemo:devel:hw:native-common / kernel-headers - SailfishOS Open Build Service (rainemak, 07:23:25) * LINK: https://build.sailfishos.org/package/show/nemo:devel:hw:native-common/kernel-headers (rainemak, 07:23:28) * 4. Comparing Linux headers: __NR_membarrier identifier - Linux source code (v4.3) - Bootlin (rainemak, 07:23:30) * LINK: https://elixir.bootlin.com/linux/v4.3/C/ident/__NR_membarrier (rainemak, 07:23:33) * 5. Background: Gecko-dev Diary (rainemak, 07:23:37) * LINK: https://www.flypig.co.uk/gecko?&list_id=855&list=gecko (rainemak, 07:23:39) * Thank you for raising this. Something that we should look at some point. (rainemak, 07:23:43) * Stabilizing TextArea.focusOutBehavior (5 mins -- asked by direc85) (rainemak, 07:29:13) * Jolla Messages keeps the keyboard open after the message has been sent, so sending the next message is easier. (rainemak, 07:29:22) * Whisperfish, a Signal client, would have benefited from the same behaviour from the start, but we couldn’t figure out (rainemak, 07:29:22) * how it was done. Eventually I stumbled upon the solution while looking for something else – see Whisperfish issue 489 (rainemak, 07:29:22) * 1 – it turned out to be the TextArea.focusOutBehavior, which is an undocumented property, inherited from the private (rainemak, 07:29:23) * side. Could Jolla stabilize the property and document it so other applications could easier replicate the same (rainemak, 07:29:26) * behaviour? (rainemak, 07:29:29) * LINK: https://gitlab.com/whisperfish/whisperfish/-/merge_requests/489 (rainemak, 07:29:31) * Example apps sailfishsilica-qt5-demos and sailfishsilica-qt5-demos-private are good examples that could be used to (rainemak, 07:29:37) * check how components can be used. Whether a property is inherited from the private TextBase or not is not a problem. (rainemak, 07:29:38) * TextField documentation already have properties and methods documented that are coming from the TextBase. Your (rainemak, 07:29:41) * suggestion to document focusOutBehavior sounds like a good idea. (rainemak, 07:29:44) * Improving Silica documentation (10 mins -- asked by direc85) (rainemak, 07:31:50) * There are typos, mistakes and simply wrong information in the Silica documentation 1. Example: By documentation, (rainemak, 07:31:59) * PageStack.push() takes var page, PageStack.replace() takes int page and PageStack.nextPage() takes Page page – by (rainemak, 07:31:59) * experiments I can tell they all take Page type variable. I mentioned about this in the #sailfishos channel some (rainemak, 07:31:59) * months ago, but it was not reacted to, even there was now active conversation after that for some time (if my memory (rainemak, 07:31:59) * serves me right). Another example of vast amount of missing data is list of inherited methods and properties. As Qt (rainemak, 07:32:02) * documentation website defaults to Qt 6, manually jumping to version Qt 5.6 needs to be done often, so having even the (rainemak, 07:32:07) * Qt part of the list would be most helpful. Could the current documentation be opened to pull requests so community (rainemak, 07:32:10) * could contribute to it? If not, what is the preferred way for the community to contribute? (rainemak, 07:32:13) * Thank you for your input. There’s clearly an error in the PageStack.replace() – let's fix that. The PageStack.nextPage (rainemak, 07:32:40) * is not similar to PageStack.push nor to PageStack.replace as PageStack.nextPage is a lookup function that returns the (rainemak, 07:32:40) * page that follows fromPage argument (instance of a Page). Whereas both push and replace allow lazy page creation by (rainemak, 07:32:40) * passing Page Component or by passing url (usually used along with Qt.resolvedUrl). In addition, both push and replace (rainemak, 07:32:40) * accept a Page instance or an array of Page instances. This is the reason why these two take var argument. (rainemak, 07:32:43) * (rainemak, 07:33:11) * Hint: from application start up and Page creation point of view many times best approach is to use resolved url as it (rainemak, 07:33:11) * defers both component compilation and object creation at the point when a page is really needed. Thus, avoids (rainemak, 07:33:11) * unnecessary compilation and instance creations. Instantiating Page instances ahead of time is not usually good idea. (rainemak, 07:33:11) * You can accidentally write your QML so Pages are created too early for example by declaring needed Pages somewhere (rainemak, 07:33:12) * without wrapping them to Components. (rainemak, 07:33:15) * (rainemak, 07:33:17) * Documenting Qt properties and methods as part of Silica documentation is tricky. Certainly, we understand and see your (rainemak, 07:33:45) * point that it would be helpful to have. (rainemak, 07:33:45) * (rainemak, 07:33:45) * We haven’t really considered opening only documentation parts of closed repositories as that’s currently part of the (rainemak, 07:33:57) * same repository as the source codes. (rainemak, 07:33:57) * (rainemak, 07:34:14) * Preferred way for community to contribute would be raising API documentation issues in the forum. We can consider (rainemak, 07:34:14) * creating a separate category for API documentation discussion if there would be plenty of discussion. If that is (rainemak, 07:34:14) * preferred approach, please let us know. (rainemak, 07:34:14) * Open PR discussion (10 mins -- asked by jolla / flypig) (rainemak, 07:42:35) * These are all Gecko ESR 91 related. These aren’t stuck, I’m just raising them in case there’s any need to discuss them. (rainemak, 07:43:07) * (rainemak, 07:43:07) * ICU 70.1 (thanks for looking at this already) (rainemak, 07:43:07) * Open PR URL: Update to version 70.1 by llewelld · Pull Request #1 · sailfishos/icu · GitHub 1 (rainemak, 07:43:07) * LINK: https://github.com/sailfishos/icu/pull/1 (rainemak, 07:43:08) * This was merged already. (rainemak, 07:43:18) * GCC backported patch (special thanks to direct85 for this) (rainemak, 07:43:34) * (rainemak, 07:43:34) * (rainemak, 07:43:34) * Open PR URL: [gcc] Prefer bsl/bit/bif for copysignf. (backport GCC-8) by direc85 · Pull Request #4 · sailfishos/gcc · GitHub (rainemak, 07:43:34) * LINK: https://github.com/sailfishos/gcc/pull/4 (rainemak, 07:43:34) * Related thread: Compiling GCC for SDK (rainemak, 07:43:37) * Query: Is there a way to get the aarch64 build on OBS to complete so I can test this; alternatively is there (rainemak, 07:43:39) * a way to do ‘dumb’ builds using the local SDK? (rainemak, 07:43:42) * LINK: https://build.sailfishos.org/package/show/home:flypig:gecko-esr91/gcc (flypig_, 07:44:56) * cbindgen 0.19.0 (thanks for all the useful feedback on this as well) (rainemak, 07:46:27) * Open PR URL: [packaging] Update to version 0.19.0 by llewelld · Pull Request #7 · sailfishos/rust-cbindgen · GitHub 1 (rainemak, 07:46:27) * https://github.com/sailfishos/rust-cbindgen/pull/7 (rainemak, 07:46:27) * This was merged as well. (rainemak, 07:46:27) * Any other open PRs to discuss? (rainemak, 07:51:05) * Untracked bug reports (5 min -- asked by pherjung) (rainemak, 07:51:29) * Untracked bug reports (rainemak, 07:51:38) * LINK: https://forum.sailfishos.org/t/community-meeting-on-irc-14th-september-2023/16646/3 (rainemak, 07:51:38) * The Community Bug Coordination Team have done a superb job once again this fortnight. (rainemak, 07:51:45) * (rainemak, 07:51:45) * As a result of their work, we now have: (rainemak, 07:51:45) * 9 high quality bug reports now recorded internally and tagged as "tracked". (rainemak, 07:51:45) * 4 bug reports marked as "pending" more info. (rainemak, 07:51:45) * 3 bug reports tagged as "fixed". (rainemak, 07:51:46) * 1 marked as duplicates and closed. (rainemak, 07:51:48) * General discussion (5 min) (rainemak, 07:53:38) * Next meeting time and date (2 min) (rainemak, 07:58:32) * Next meeting will be held on Thursday 28th September 2023 at 07:00am UTC: 2023-09-28T0700Z (rainemak, 07:59:33) Meeting ended at 08:00:10 UTC. Action Items ------------ Action Items, by person ----------------------- * **UNASSIGNED** * (none) People Present (lines said) --------------------------- * rainemak (179) * flypig_ (36) * piggz (20) * ExTechOp (5) * abr (3) * sailbot (2) * Alan[m] (1) * Keto (1) * jpwalden (1) * sebix[m] (1) * nephros_ (1) Generated by `MeetBot`_ 0.1.4 .. _`MeetBot`: http://wiki.debian.org/MeetBot