**** BEGIN LOGGING AT Wed Apr 22 02:59:57 2020 Apr 22 07:28:03 Slothman21: Cosmo Communicator more likely v.s. Palm Apr 22 07:28:09 Palm didn't release the full kernel so far Apr 22 07:28:22 Important bits are missing to get the kernel building Apr 22 07:28:32 Morning! Apr 22 07:29:01 Morning! Apr 22 07:30:11 Cosmo Communicator looks nice; I'd say why not, depending on available time & source code & maybe a volonteer :p Apr 22 07:33:24 But the Gemini PDA, in the same fashion, is a more secure choice, as we know there are already some other ports available for this target Apr 22 07:35:43 though frankly I have my hands more than full these days with pinephone and the other targets, so I can help but I won't do the port myself Apr 22 07:40:43 true. I listed the cosmo mostly because of the ramp up in specs and similarity of build (and price). They're also now working on a more standard form factor called the astro as well. I'd imagine there's enough to keep you guys busy for a long time with all the other devices as well, especially the pinephone which if I'm understanding right, is like a blank slate. Fairly sure it came out with zero software so it's been built from scratch by the Apr 22 07:40:43 community. Apr 22 07:47:32 Herrie: sorry, I didn't realize the kernel wasn't available. I knew as part of the root, you had to package something like the phone's image and send it to a rooted android device to get it rooted and installed. I just assumed that that image would be all that was needed. I'm not very experienced in these things. Apr 22 07:57:57 Slothman21: Well I got a kernel from them, but was unable to build due to missing pieces. Not sure how they got TWRP working Apr 22 07:58:02 Do you have a link to the TWRP? Apr 22 07:59:09 Sorry, I think I might not have written something clear enough. They've listed the pvg100 on their "to do" list. whether or not it gets done, I don't know. Apr 22 08:00:46 https://twrpbuilder.github.io/downloads/twrp/ I believe it's number 44 (at the last time I checked) on their "builds in queue" Apr 22 08:00:48 Slothman21: Anyway Android targets are a bit harder v.s. Pinephone for example Apr 22 08:01:15 Yea, I'd imagine Apr 22 08:02:55 Problem is that there's not really a standard solution for "newer" Android versions. For Android 5 and 7 based devices we have the "Halium" approach which is pretty generic and easy to port to new target devices. However for Android 8/9/10 based devices this isn't there yet Apr 22 08:03:27 Halium basically is a very minimal Android build that alternative OS-es such as LuneOS, UB Ports, SFOS use to access the various hardware bits Apr 22 08:03:57 Which is required because all drivers are proprietary and provided in binary blob format mostly, so cannot run using a standard linux kernel. Apr 22 08:04:36 Targets like Pinephone, Librem are a lot more open therefore don't need so much of an intermediary layer Apr 22 08:21:17 As I understand it, the Gemini and the Cosmo have dual boot capabilities with the code being open source. I don't think it's necessary to build over android Apr 22 08:27:10 Seems they are a bit more open, but since they're using MediaTek chipsets, they're a bit of a pain. But in general when SFOS, PmOS etc run on it, LuneOS shouldn't normally that hard to get working Apr 22 08:38:15 It should be ok I guess. And the work we did for pinephone (u-boot, initrd, sensors...) can likely be reused there Apr 22 09:09:52 That's pretty cool. I think something like webos would really shine on such a device. Especially the keyboard and overall feel of the design. Apr 22 09:14:03 I imagine you guys would need physical devices to test with. Or do they have virtual emulation for this kind of stuff? Apr 22 09:54:27 Yes, without something to test with, it's almost impossible to get it right, there is always some particular stuff for each device Apr 22 10:38:36 Yocto 3.1 Dunfell was released yesterday, anyone willing to test unstable builds in runtime? :) Apr 22 10:40:30 Herrie: can you easily retest https://bugreports.qt.io/browse/QTBUG-81790 with 5.15 beta4 in your windows? it has chromium 80 so in theory it should work now, but in LuneOS it doesn't seem to work still Apr 22 12:54:55 latest image with 5.15 beta4 http://build.webos-ports.org/luneos-unstable-dunfell/images/qemux86-64/luneos-dev-emulator-qemux86-64-unstable-0-121.tar.gz Apr 22 13:00:44 Tofe: I've applied your libhybris change in meta-smartphone and tissot build is almost finished locally, now http://jenkins.nas-admin.org/job/LuneOS/view/unstable/job/luneos-unstable_tissot/84/console should be successful as well, thanks Apr 22 13:01:06 the change from https://github.com/webOS-ports/meta-webos-ports/pull/421/commits/d27a08a7c50abc8eb3ba1b983d3ce8d509cc0d5f Apr 22 13:03:00 JaMa: do you want me to remove that from zeus ? it's true that it's not really needed there... Apr 22 13:03:46 as you wish Apr 22 13:04:05 but we need to prevent applying it in both places Apr 22 13:04:41 the emulator image with 5.15 still has multiple issues (not sure if it's all because 5.15 or dunfell as well) Apr 22 13:04:58 browser doesn't start (which is reproducible in last testing build with 5.14 as well) Apr 22 13:06:26 the CardShell.qml now loads Luneos.Components OK, but there is still some size issue (the black bar at the bottom is completely missing) Apr 22 13:06:35 and there is still the refresh issue https://bugreports.qt.io/browse/QTBUG-81790 Apr 22 13:06:53 will rebuild 5.15 with zeus now to see how it works there Apr 22 13:18:47 JaMa: ok, well, it's a bit of mess :) Apr 22 13:19:15 I'll remove the patch from my zeus PR, it has no use there Apr 22 13:20:59 done Apr 22 13:21:20 ... and let's merge that PR, it's been there for a while Apr 22 13:22:03 ... done Apr 22 13:34:52 Tofe: what qt do you currently have on pinephone? 5.14.2 as current testing build? Apr 22 13:50:47 JaMa: yes Apr 22 14:06:27 JaMa: I can try to install 5.15 beta4 on my machine and see what it does Apr 22 14:06:37 But will probably be tomorrow Apr 22 14:13:41 thanks, no rush, the ticket probably won't get resolved before final 5.15 anyway, just wanted to confirm that it's still reproducible on systems they are more likely using as well **** ENDING LOGGING AT Thu Apr 23 02:59:58 2020