r0kk3rz | piggz_: oh good, rinigus got you sorted | 01:37 |
---|---|---|
piggz_ | r0kk3rz: well, sorted myself ;) | 07:57 |
rinigus | piggz: exactly! | 07:58 |
T42 | <adampigg> rinigus: except its still not working...im sure ill get to the bottom of it. loop devices exist but it stils fails in the kickstart stage, adding debugging to it | 09:26 |
r0kk3rz | what device? | 09:40 |
T42 | <adampigg> r0kk3rz: https://gitlab.com/sailfishos-porters-ci/t5-ci/-/jobs/687604373#L3164 | 09:59 |
T42 | <adampigg> so, in the run-mic-lvm script i have added the mknod commands, and early in the job i print them out and they are there | 10:00 |
r0kk3rz | seems to have built something | 10:05 |
r0kk3rz | 18mb seems a little small though ;) | 10:06 |
T42 | <adampigg> Yeah, it misses,userdata | 10:11 |
r0kk3rz | on mine the command i use is mic create loop | 10:12 |
r0kk3rz | you're using mic create fs | 10:12 |
r0kk3rz | aside from that, it seems identical to what i did for f5321 | 10:12 |
r0kk3rz | same with tama | 10:13 |
T42 | <adampigg> Ah, ok, will.change to loop later | 10:46 |
T42 | <adampigg> Ta | 10:46 |
rinigus | adampigg: good luck with conversion to lvm | 10:59 |
T42 | <elros34> @kabouik if you are interested I have pushed some cursor test branch to qxcompositor. It works with touchscreen but I didn't test it yet with mouse. | 14:28 |
T42 | <edp_17> @elros34 : I am also interested. How can I test it? | 15:00 |
T42 | <elros34> just build and install | 15:01 |
T42 | <edp_17> Okay, thanks. 👍 | 15:03 |
T42 | <elros34> not sure if you get it right, it's not system wide cursors, just for qxcompositor. If that will work then it can be also used as patch for lipstick for system wide cursor. | 15:05 |
T42 | <kabouik> Oh thanks for pinging me @elros34, interesting. I will look at it later, I have to check if my chroot installation still works correctly after I installed LXC some months ago, LXC builds qxdisplay too, perhaps there's a conflict | 15:50 |
T42 | <kabouik> Related: by starting the X session with Lightdm in LXC, we managed to get a cursor working showing up too. However, its position is updated only when an event is called (right click, left click, touching the screen), as if it was still considered a touch screen. Also, thist led to losing multitouch support and touch support in Firefox | 15:52 |
T42 | with xinput2 | 15:52 |
T42 | <elros34> ok so maybe I need to handle also QMouseEvent::HoverMove. Anyway all events will be printed to journal | 15:54 |
kalube | Can anyone help me set up udev rules to create /dev/block/bootdevice/by-name/boot for example on a device with slots, right now boot_a and boot_b are created which means the boot slot has to be hard coded into fstab/mount units and making it slot agnostic would make the install experience a lot more friendly | 16:09 |
kalube | It's possible to extract the active boot slot from GPT headers or more easily the kernel cmdline contains something like "androidboot.slot_suffix=_a" | 16:10 |
T42 | <Max %lastname%> Can someone try to port 3.3 release for Athene? | 16:38 |
kalube | Do it dude :D | 16:39 |
T42 | <adampigg> r0kk3rz: gah, it failed again! | 17:26 |
rinigus | @adampigg: do you need to have --pack-to in your mic command? I don't have it and it seems like loop mounting is after running pack scripts | 17:39 |
T42 | <adampigg> Ah, good spot, i only managed to diff the kickstart before having to go out again | 17:41 |
mal | @adampigg are you following this command https://github.com/mer-hybris/droid-hal-device/blob/master/helpers/build_packages.sh#L408 | 18:17 |
T42 | <ankaos> sfos version update? | 19:14 |
T42 | <adampigg> rinigus: mal: thats fixed it | 19:55 |
T42 | <adampigg> removed —pack-to | 19:55 |
rinigus | @adampigg: great! | 20:37 |
T42 | <adampigg> rinigus: mal: i assume i know what sailfishos and hybris-boot images are, but what are fimage and vendor? | 20:46 |
mal | @adampigg I'll check those tomorrow | 20:54 |
kalube | mal: can you set up nemo:testing:hw:oneplus:enchilada for me on OBS? Finally getting OTAs ready :D | 21:04 |
mal | kalube: https://build.sailfishos.org/project/show/nemo:testing:hw:oneplus:enchilada | 21:12 |
kalube | Brilliant thank you :D | 21:14 |
kalube | Mister_Magister: do you have the handy dandy script to clone everything from devel to testing? | 21:24 |
Mister_Magister | nah i don't but here is what you can do. osc co nemo:devel:device; osc co nemo:testing:device, copy all dirs from devel dir to testing dir and remove .osc dir in every of them and then while being in testing do osc ar; osc ci and it will create all the projects (you might have to do last step multiple times_ | 21:25 |
Mister_Magister | thats how i do it | 21:25 |
T42 | <elros34> example of such a script is in faq | 21:28 |
piggz | mal: according to the scripts, the vendor.img gets flashed to the vendor_a/b partitions, but is very small ....same on tama | 21:33 |
piggz | and fimage gets flashed to system_b (only) | 21:34 |
Mister_Magister | piggz: you doin lvm thingy? | 21:34 |
piggz | aye,just wondering what the content of everything is before i kill my pro1! | 21:35 |
piggz | sailfish.img goes on userdata | 21:35 |
Mister_Magister | :P | 21:35 |
Mister_Magister | good luck | 21:35 |
piggz | it wont be tonight tho | 21:36 |
piggz | feel free to download the image and write it to your pro1! | 21:36 |
Mister_Magister | xd | 22:12 |
kalube | r0kk3rz: can you add people to CI? have testing building on OBS so getting CI going too would be nice https://build.sailfishos.org/project/show/nemo:testing:hw:oneplus:enchilada | 22:14 |
r0kk3rz | kalube: request access through gitlab | 23:00 |
kalube | r0kk3rz: done! | 23:02 |
r0kk3rz | kalube: you're in, welcome | 23:40 |
r0kk3rz | do you understand how it works? | 23:40 |
kalube | Good to be here :D | 23:41 |
kalube | deathmist has been helping me out, I've already got a ks file that uses OBS so it's just a case of cloning oneplus 5 repo and changing "cheeseburger" to "enchilda" | 23:42 |
kalube | That sound about right? | 23:43 |
r0kk3rz | pretty much yeah | 23:43 |
r0kk3rz | for most devices its quite simple | 23:43 |
kalube | Cool | 23:43 |
kalube | I have a few ideas on how to get fajita builds to without having to duplicate OBS and CI | 23:43 |
r0kk3rz | its not really duplicate | 23:44 |
r0kk3rz | the ci only packs the image | 23:44 |
kalube | ie just adding a second metapackage (and device specific packages) for the 6T and running mic build twice | 23:44 |
r0kk3rz | oh, multiple devices with 99% the same stuff | 23:44 |
r0kk3rz | yeah | 23:44 |
kalube | yeah that's what I mean sorry | 23:44 |
r0kk3rz | yeah thats always annoying | 23:45 |
kalube | like with cheeseburger and dumpling (5 and 5T) the devices are almost identical just with a separate kernel and device packages | 23:45 |
r0kk3rz | i just have duplicate everything with scorpion and scorpion_windy | 23:45 |
kalube | I /think/ if on OBS I add all the device specific packages for both, create a common metapackage and then specific ones for each device and then modify CI so mic build gets run twice by adding a token to https://gitlab.com/sailfishos-porters-ci/cheeseburger-ci/-/blob/master/Jolla-@RELEASE@-cheeseburger-@ARCH@.ks#L23 then we should be good... | 23:47 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!