2016-06-05 00:00:30	Nokius	:( the docks have log shipping time cheked ebay.fr
2016-06-05 00:18:34	@kido	yes it's hard to get one
2016-06-05 08:38:25	locusf	good morning
2016-06-05 09:28:59	@kido	hey locusf
2016-06-05 09:29:31	@kido	do we hack your zenwatch today ? :)
2016-06-05 09:30:13	@kido	you gotta wear asteroid for the sailfish developer day
2016-06-05 09:31:54	locusf	sure
2016-06-05 09:32:32	locusf	lets do this
2016-06-05 09:32:43	@kido	cool,  turn on some good progressive metal and let's go :)
2016-06-05 09:32:44	locusf	now that I'm a bit more recovered from the flu :p
2016-06-05 09:33:01	@kido	oh you had flu ? that's a pain
2016-06-05 09:33:16	@kido	how cold is helsinki nowadays ?
2016-06-05 09:33:27	locusf	the thing is that it wasn't even cold
2016-06-05 09:33:41	locusf	nearly heat-wave hot the day before and weeks before
2016-06-05 09:33:44	locusf	really warm
2016-06-05 09:34:01	locusf	well, at least when compared to the other part of the climate :D
2016-06-05 09:34:27	@kido	sounds like a strange place to live in anyway :)
2016-06-05 09:34:46	locusf	I hear that a lot ...
2016-06-05 09:35:00	@kido	hehe
2016-06-05 09:36:39	@kido	so where were we? rootfs booting with splashscreen shown but no way to gain an adb shell to debug lipstick's behavior?
2016-06-05 09:37:33	locusf	yeah
2016-06-05 09:37:37	locusf	pretty much
2016-06-05 09:37:45	locusf	I'd like to see if systemd has even booted
2016-06-05 09:37:54	@kido	if you have bootscreen then systemd worked
2016-06-05 09:38:19	locusf	yes
2016-06-05 09:38:23	@kido	I just think your defconfig is messed up by the "simplifying" script of OE
2016-06-05 09:38:26	locusf	journalctl might be very interesting to see
2016-06-05 09:38:47	@kido	I talked about that with kernel devs and they said that it's a current behavior with the usb subsystem in old kernels (like our 3.10 ones)
2016-06-05 09:38:56	@kido	s/current/common/
2016-06-05 09:39:37	@kido	according to me the android usb otg module is just not compiled
2016-06-05 09:41:00	@kido	locusf: try to add this: KCONFIG_MODE = " " to your linux recipe
2016-06-05 09:41:14	locusf	ok
2016-06-05 09:41:20	@kido	and then recompile the kernel and fastboot your new boot.img
2016-06-05 09:44:06	locusf	gah
2016-06-05 09:44:17	locusf	how do I properly clean the kernel directory ..
2016-06-05 09:44:59	@kido	bitbake -c clean linux-athias
2016-06-05 09:46:42	@kido	by the way have you continued playing with OE for nemo?
2016-06-05 09:47:21	locusf	it didn't recompile anything after clean
2016-06-05 09:47:26	locusf	yeah just tried CHIP
2016-06-05 09:47:35	locusf	but it fails in flashing the fel image
2016-06-05 09:47:48	@kido	oh cool, let me know if it works on CHIP, I got one for my summer's internship
2016-06-05 09:48:31	@kido	I'll develop an open source v4l driver for the VPU acceleration, it would avoid the pain of allwinner's proprietary blob and the current open source libvdpau implementation
2016-06-05 09:48:50	@kido	I'll also try to get that upstream if I can
2016-06-05 09:48:51	locusf	woah
2016-06-05 09:48:57	locusf	thats awesome
2016-06-05 09:49:25	@kido	yes that's really cool, and my boss will be the kernel's maintainer of allwinner code in upstream
2016-06-05 09:50:04	@kido	so it would be a pleasure to try nemo on this
2016-06-05 09:50:22	locusf	yup
2016-06-05 09:50:32	locusf	it was cleanall
2016-06-05 09:50:34	@kido	"it didn't recompile anything after clean" <- then it's time to be more agressive and use cleanall instead of clean in my previous command
2016-06-05 09:50:40	@kido	well you figured it out :)
2016-06-05 09:50:41	locusf	heheh
2016-06-05 09:54:08	locusf	hmm
2016-06-05 09:54:19	locusf	there is a modules.tgz created but rootfs is not updated at all
2016-06-05 09:54:31	@kido	rootfs has no reason to be updated
2016-06-05 09:54:50	@kido	just check your zImage-dtb-athias.fastboot
2016-06-05 09:55:13	locusf	ok
2016-06-05 09:55:21	@kido	or there might have been another place where you could get your boot.img ? I remember you had to pack it differently than dory's kernel
2016-06-05 09:55:35	@kido	so maybe you have to run abootimg or whatever by hand... maybe you remember what you had to do ?
2016-06-05 09:56:15	locusf	throws up partitions
2016-06-05 09:56:31	@kido	yes that's the problem we had, you have to pack it with another tool
2016-06-05 09:56:37	@kido	let me read logs
2016-06-05 09:58:16	@kido	locusf: http://pastebin.com/BCG4fRUY
2016-06-05 09:59:11	@kido	(but this time you gotta use the asteroid's ramdisk)
2016-06-05 10:00:17	@kido	we probably got the bootimg.cfg by extracting a working boot file
2016-06-05 10:06:40	locusf	my console history doesn't go far enough :/
2016-06-05 10:08:54	locusf	it was roots history
2016-06-05 10:09:33	@kido	did you keep working boot.img files? (from android)
2016-06-05 10:10:09	@kido	call build/tmp-glibc/sysroots/x86_64-linux/usr/bin/abootimg -x the_android_boot_img_file
2016-06-05 10:10:21	@kido	it will bring you the bootimg.cfg file you want
2016-06-05 10:10:23	locusf	yes
2016-06-05 10:11:56	@kido	where zImage is in your build/tmp-glibc/work/athias-oe-linux-gnueabi/linux-athias/3.10+lollipop-r0/git/arch/arm/boot/zImage-dtb
2016-06-05 10:12:45	@kido	and initrd.img is in your build/tmp-glibc/deploy/images/athias/initramfs-android-image-athias.cpio.gz
2016-06-05 10:12:52	locusf	ok
2016-06-05 10:15:51	locusf	that work directory is new
2016-06-05 10:15:59	locusf	boot logo !!!!!
2016-06-05 10:16:16	locusf	to throwup ...
2016-06-05 10:16:18	locusf	fuck ..
2016-06-05 10:18:12	@kido	mmhhh
2016-06-05 10:19:17	@kido	can you watch your computer's dmesg -w when booting your connected watch ?
2016-06-05 10:19:34	@kido	Just to see if adb appears even for a second or so
2016-06-05 10:22:49	locusf	hmm
2016-06-05 10:22:59	locusf	I wonder if the root partitions is right after all
2016-06-05 10:23:39	@kido	well it might be worth disabling some services to check what's going on
2016-06-05 10:24:01	@kido	the best way to do this would be from the rootfs
2016-06-05 10:24:17	@kido	sorry I meant the ramfs
2016-06-05 10:24:58	locusf	there is no rootfs.ext inside userdata
2016-06-05 10:25:14	locusf	hmm oh well
2016-06-05 10:25:30	@kido	in your /sdcard ?
2016-06-05 10:25:42	@kido	you should have a linux directory
2016-06-05 10:25:49	locusf	I dunno which one is /sdcard
2016-06-05 10:25:55	@kido	with a rootfs file
2016-06-05 10:26:41	@kido	it depends on your device but from android with adb you should have it in /sdcard/linux dir
2016-06-05 10:28:01	locusf	yeah
2016-06-05 10:28:15	locusf	it is /storage/emulated/legacy
2016-06-05 10:28:55	@kido	indeed
2016-06-05 10:30:50	locusf	does it somehow loop-mount the image?
2016-06-05 10:31:03	locusf	rootfs
2016-06-05 10:31:56	@kido	yes it does
2016-06-05 10:33:28	@kido	https://github.com/locusf/meta-athias-hybris/blob/master/recipes-core/initrdscripts/initramfs-boot-android/init.sh#L85
2016-06-05 10:35:11	@kido	if you want to try to get a shell from ramdisk in order to debug the rootfs (and make sure adb works) use that https://github.com/locusf/meta-athias-hybris/blob/master/recipes-core/initrdscripts/initramfs-boot-android/init.sh#L51
2016-06-05 10:38:51	locusf	ok
2016-06-05 10:40:21	locusf	I don't know which one is the sdcard device file :/
2016-06-05 10:42:44	@kido	well it shouldn't matter, the script should figure it out by itself
2016-06-05 10:43:09	@kido	if you see the splashscreen then it found the rootfs and systemd does something
2016-06-05 10:43:43	locusf	ok
2016-06-05 10:44:05	@kido	if you want to debug further modify that script or add enable_adb to your kernel's cmdline
2016-06-05 10:44:15	@kido	so that it runs adbd
2016-06-05 10:44:31	@kido	and let's see if you can get a shell from this
2016-06-05 10:45:24	locusf	yeah
2016-06-05 10:47:37	locusf	fingers crossed
2016-06-05 10:53:58	locusf	boot logo but then blackness and nothing in dmesg
2016-06-05 10:54:40	@kido	did you force adbd running from ramdisk this time ?
2016-06-05 10:55:05	locusf	I did try that but it resulted in blurting partitions
2016-06-05 10:55:20	@kido	damn
2016-06-05 10:57:38	locusf	hmm
2016-06-05 10:57:51	locusf	how does it create /usr/bin/android-gadget-setup ?
2016-06-05 10:58:11	@kido	it should come from android-tools I guess
2016-06-05 10:59:08	@kido	no it doesn't
2016-06-05 11:00:38	@kido	mhhh why did I miss that script before
2016-06-05 11:00:54	@kido	I mean, I added the same thing to init.sh but it wasn't needed
2016-06-05 11:01:59	@kido	try to comment the android-gadget-setup line
2016-06-05 11:02:11	@kido	it will try to rewrite what we do just before
2016-06-05 11:03:12	locusf	I did comment it
2016-06-05 11:03:17	locusf	it boots apparently ok now
2016-06-05 11:03:24	locusf	but just that nothing happens after boot logo
2016-06-05 11:03:34	locusf	it won't even blurt the partitions anymore
2016-06-05 11:03:41	locusf	and doesn't reboot after 120s
2016-06-05 11:04:41	@kido	ok so next time we meet remember to slap me for those shitty boot script I did a really bad job there
2016-06-05 11:05:06	@kido	anyway, we are getting a little bit further
2016-06-05 11:05:11	locusf	meh
2016-06-05 11:05:13	locusf	yeah
2016-06-05 11:05:17	locusf	don't slap yourself
2016-06-05 11:05:21	locusf	its a learning process :)
2016-06-05 11:05:27	locusf	now
2016-06-05 11:05:39	locusf	I'll pull the rootfs image
2016-06-05 11:05:44	locusf	and see if it has journals
2016-06-05 11:05:51	@kido	I don't think it saves journals
2016-06-05 11:05:58	locusf	damn
2016-06-05 11:06:06	locusf	I'll still see :)
2016-06-05 11:06:31	@kido	https://github.com/AsteroidOS/asteroid/wiki/Boot-Process#troubleshooting-the-boot-process
2016-06-05 11:06:42	@kido	"Logs are normally stored in a tmpfs filesystem which is lost whenever the watch is rebooted. To disable the tmpfs mount, first mount the sdcard card containing the rootfs, and then the rootfs:"
2016-06-05 11:07:02	@kido	thanks to our apple :)
2016-06-05 11:07:16	locusf	ah
2016-06-05 11:07:18	locusf	brilliant
2016-06-05 11:09:40	locusf	bad geometry: block count 425124 exceeds size of device (120192 blocks)
2016-06-05 11:09:45	locusf	when mounting ..
2016-06-05 11:10:08	@kido	where did you get that?
2016-06-05 11:10:44	locusf	my desktop
2016-06-05 11:10:50	locusf	when trying to mount the image
2016-06-05 11:11:56	@kido	try to run e2fsck on it
2016-06-05 11:14:03	locusf	its very corrupted
2016-06-05 11:14:22	@kido	ok
2016-06-05 11:14:31	@kido	try to push a new rootfs to your watch
2016-06-05 11:14:36	@kido	(fresh out of OE)
2016-06-05 11:14:47	@kido	boot it, pull it, mount it
2016-06-05 11:14:51	locusf	yes
2016-06-05 11:31:40	locusf	the grind ...
2016-06-05 11:31:45	locusf	so much grind :)
2016-06-05 11:36:30	@kido	 ?
2016-06-05 11:38:40	locusf	cannot execve('/system/bin/logd'): Permission denied
2016-06-05 11:39:43	@kido	oh nice
2016-06-05 11:40:04	@kido	where did you get that log ?
2016-06-05 11:40:24	locusf	https://gist.github.com/locusf/c9aa87938990911b448a367f4871fb9c
2016-06-05 11:40:24	@kido	chmod +x /system/bin/*
2016-06-05 11:40:26	locusf	jorunal
2016-06-05 11:40:55	@kido	good job :)
2016-06-05 11:45:29	@kido	[USB] asus_chg_detect_work: g_usb_boot is 0 , add more 2 sec
2016-06-05 11:45:46	@kido	this might also explain the usb gadget not working
2016-06-05 11:46:04	locusf	also some firmware crap
2016-06-05 11:46:37	@kido	yep for  video
2016-06-05 11:46:57	@kido	we might be missing a directory from android containing some shitty blobs
2016-06-05 11:49:59	locusf	yeah
2016-06-05 11:50:44	@kido	tell me how it goes with the chmod +x
2016-06-05 11:53:46	locusf	nothing happened
2016-06-05 11:54:40	@kido	ok and nothing of interest in journal ?
2016-06-05 11:54:42	locusf	except partitions bluirt
2016-06-05 11:55:44	locusf	which is strange
2016-06-05 12:00:12	@kido	https://android.googlesource.com/kernel/msm/+/android-msm-anthias-3.10-marshmallow-mr1-wear-release/drivers/media/platform/msm/vidc/msm_vidc_common.c#2242
2016-06-05 12:00:54	@kido	well this seems to be only about video encoding/decoding so it shouldn't be a problem
2016-06-05 12:12:16	locusf	no help from chmod
2016-06-05 12:15:36	locusf	https://www.youtube.com/watch?v=PEcWqIdXEAc
2016-06-05 12:16:02	locusf	warning for anyone with epilepsy
2016-06-05 12:16:21	ledixus	Melanosuchus: borgy isnt here
2016-06-05 12:19:10	@kido	locusf: good stuff
2016-06-05 12:19:36	@kido	at least finland is a good country for metal :)
2016-06-05 12:26:10	Melanosuchus	ah forgot to add this to the auto join
2016-06-05 12:26:50	ledixus	:D
2016-06-05 12:31:38	ledixus	EvilAntCyborg: can i use asteroidos on an apple watch?
2016-06-05 12:31:39	EvilAntCyborg	Ask again later
2016-06-05 12:32:37	ledixus	EvilAntCyborg: can i use asteroidos on an apple watch?
2016-06-05 12:32:37	EvilAntCyborg	Signs point to yes
2016-06-05 12:34:16	Melanosuchus	EvilAntCyborg: can I use asteroid when on steroids?
2016-06-05 12:34:16	EvilAntCyborg	Maybe not
2016-06-05 12:34:33	Melanosuchus	EvilAntCyborg: can I use asteroid when on a steroid?
2016-06-05 12:34:33	EvilAntCyborg	My sources say no
2016-06-05 12:56:44	@kido	"kernel: lipstick (731): undefined instruction: pc=471e6108" wtf
2016-06-05 12:59:39	locusf	yep
2016-06-05 13:00:57	@kido	I'm cloning the msm code to find what can go wrong with usb otg
2016-06-05 13:01:12	@kido	I hope that fixing usb otg will bring us a shell and make our lives easier
2016-06-05 13:19:58	@kido	https://android.googlesource.com/kernel/msm/+/7f5041bb64f734366ed6ba8ded86b91ba19931a9%5E!/drivers/usb/phy/phy-msm-usb.c mhhh I don't think it should be a problem for gadget
2016-06-05 13:31:36	locusf	ok
2016-06-05 13:55:58	filtoid	hi
2016-06-05 13:59:58	@kido	hi filtoid :)
2016-06-05 14:00:11	@kido	interested in asteroid ?
2016-06-05 14:00:31	filtoid	I am - I?m currently looking at getting a smart watch but am insterested in open source
2016-06-05 14:00:54	filtoid	basically I?m just wondering if this project is being actively developed
2016-06-05 14:01:04	@kido	it is
2016-06-05 14:01:08	filtoid	am also keen to help out if I can
2016-06-05 14:01:10	filtoid	cool :)
2016-06-05 14:01:30	@kido	I was busy with a couple of things recently but good stuff is coming soon
2016-06-05 14:01:58	@kido	help is always welcome so it would be a pleasure :)
2016-06-05 14:02:03	filtoid	Cool
2016-06-05 14:02:10	@kido	do you think you could port asteroid to a new watch?
2016-06-05 14:02:28	filtoid	I don?t know that much about hardware - is it difficult?
2016-06-05 14:02:38	@kido	it's not about hardware
2016-06-05 14:02:39	filtoid	I?ve not done a lot of systems programming unfortunately
2016-06-05 14:02:45	@kido	most of us are software guys
2016-06-05 14:02:51	filtoid	what sort of things does it take
2016-06-05 14:02:56	@kido	and it doesn't really involve programming, just messing around with a linux boot process
2016-06-05 14:03:08	filtoid	well I could give it a go
2016-06-05 14:03:19	filtoid	are there any watches that you are keen to port to?
2016-06-05 14:03:50	@kido	so you can find a list here: https://github.com/AsteroidOS/asteroid/wiki/Porting-Status
2016-06-05 14:03:59	filtoid	thanks
2016-06-05 14:04:14	@kido	you can also find some info on what it takes here : https://github.com/AsteroidOS/asteroid/wiki/Porting-Guide
2016-06-05 14:04:43	filtoid	brilliant thanks
2016-06-05 14:05:12	@kido	but if you think you won't be able to do a new port (this can be painful, ask locusf we are on his zenwatch since this morning :) ) you might prefer buying a g watch or g watch urbane
2016-06-05 14:05:47	filtoid	I was looking at the G watch - there don?t seem to be many about though (or at least not in the UK)
2016-06-05 14:06:03	@kido	you can also find more info on what OpenEmbedded is here https://github.com/AsteroidOS/asteroid/wiki/OpenEmbedded (porting guide talks about it and it will be your best friend while porting)
2016-06-05 14:06:41	@kido	the g watch is the one I've been working on for a long time, it's cheap and well supported
2016-06-05 14:08:36	filtoid	I was looking at the G as it?s the one mentioned on the site
2016-06-05 14:08:53	filtoid	but I?d be more than happy to try and help porting to a different device
2016-06-05 14:13:34	@kido	so... g watch is easy to get working and showcase the capabilities but it is starting to get old. g watch urbane is nice and works ok. g watch r is on my desk but my dock, it should work easily but my dock is not working so I can't play with it. moto 360 needs a custom cable which is quite painful to get working. sony smartwatch 3 is tricky and I'm on it (I might get sony to release an aosp code for this).
2016-06-05 14:13:34	@kido	gear live is cheap and should be easy to make work but I will probably work on it during the summer with a friend. zenwatch is currently bothering locusf. there has been some efforts on the zenwatch 2 but I don't have any news on that
2016-06-05 14:16:58	@kido	once you get your watch working there are a couple of things where you can help, depending on what you want to do :) for example you can help on the QML side by coding new watchfaces, components or apps or just improving the current UX
2016-06-05 14:17:17	filtoid	that sounds like something I could help with
2016-06-05 14:17:56	filtoid	Is the sony smartwatch 3 looking promising?
2016-06-05 14:18:19	filtoid	I?ve found one for a reasonable price - so if you think that might be worth looking into I could get one of those
2016-06-05 14:18:55	@kido	mhh... the current state of sw3 is that I have asteroid booting, framebuffer more or less working, asteroid-launcher runs, its backtrace is completely fine but the screen is completely black and I don't know why
2016-06-05 14:19:14	filtoid	ah
2016-06-05 14:20:31	@kido	I've been working with sony recently and if I get "ion" working inside their unified AOSP kernel we should get AOSP for sw3 and hopefuly a source code release for the broadcom's "hwcomposer" that would allow us to debug that further and hopefuly make it work
2016-06-05 14:21:21	filtoid	that sounds cool - I?m not sure there?s anything I could do to help with that unfortunately.
2016-06-05 14:21:26	@kido	sw3 would actually be a really cool platform to work on but this black screen is remaining and it is bothering me
2016-06-05 14:21:51	@kido	well yes... it's a little bit too soon
2016-06-05 14:22:07	filtoid	Looking on amazon - there seem to be a whole bunch of ridiculously cheap smart watches
2016-06-05 14:22:25	filtoid	I?m guessing they are not worth it
2016-06-05 14:22:56	@kido	right, most of the really cheap watches that you can find on amazon (< 50$) are bad chineses watches with mediatek soc and running a real time OS "nucleus"
2016-06-05 14:23:05	@kido	they can't run linux and we can't get asteroid to work on them
2016-06-05 14:23:29	filtoid	ah i see
2016-06-05 14:23:41	filtoid	I assumed there were running an old android version
2016-06-05 14:23:57	@kido	well there are also some cheap android smartwatches but it might be harder to get the kernel source code than with android wear watches
2016-06-05 14:23:59	filtoid	but I hadn?t considered the existence of a cheap os
2016-06-05 14:24:20	filtoid	that?s cool - I?d probably rather start with something I stand a chance of getting running
2016-06-05 14:26:05	@kido	well, the least I can say is that you have the choice :)
2016-06-05 14:26:16	filtoid	hehe :) Yes that?s true
2016-06-05 14:26:26	filtoid	That list you sent me is very helpful
2016-06-05 14:27:19	@kido	it might not be exaustive though. I don't stay updated on the new watches that are released, you might find interesting new models but probably more expensive too
2016-06-05 14:27:35	@kido	it also depends on the money you want to spend on it
2016-06-05 14:28:00	filtoid	I?m not looking to spend too much - particularly if it won?t work
2016-06-05 14:28:25	filtoid	I?m looking around the ?100-200 (~$200-400) mark I think
2016-06-05 14:28:32	@kido	sure I understand :)
2016-06-05 14:28:35	filtoid	Although preferably on the lower end
2016-06-05 14:34:29	@kido	locusf: by the way, do you remember what conclusions we came to about "/sys/devices/platform/msm_hsusb/gadget/wakeup" <- I think that init.asus.usb.sh might be a clue for us
2016-06-05 14:35:01	filtoid	I might go with an LG G then look at getting one of the ones on a list at a later date - Many thanks for all your help Kido
2016-06-05 14:35:13	@kido	filtoid: good! :)
2016-06-05 14:35:15	@kido	by the way
2016-06-05 14:35:42	@kido	if you want to try to help porting without spending anything you might give a try at porting asteroid to an emulator
2016-06-05 14:36:08	@kido	there will probably be some tricky parts but it can be useful for other developers
2016-06-05 14:36:27	filtoid	ah ok, that might be a good way to get started without having to wait for the post
2016-06-05 14:36:40	@kido	and one more thing, if you like QML you can already start playing with it from your computer
2016-06-05 14:37:11	@kido	since QML is just interface description, if you have ideas of QML components for "qml-asteroid" you can prototype that from your computer and test on a watch later
2016-06-05 14:37:36	filtoid	that?s a good idea - I shall look at that too
2016-06-05 14:38:01	@kido	same is true for watchfaces https://github.com/AsteroidOS/asteroid/wiki/Watchfaces-creation (kind of bad guide for now since we don't really have watchfaces at the moment)
2016-06-05 14:38:45	@kido	anyway, just keep in mind to have fun and it will be ok! :)
2016-06-05 14:39:49	filtoid	Will do - thank you :)
2016-06-05 15:23:44	@kido	locusf: there is one more strange thing I don't understand in your logs
2016-06-05 15:24:04	@kido	l766: systemd[1]: Stopped Android Debug Bridge. <- but it is never started
2016-06-05 15:27:35	locusf	yes
2016-06-05 15:28:10	@kido	it might come from usb-moded but I don't get why it wants to stop adbd instead of starting it
2016-06-05 15:29:25	@kido	are you sure this line is executed: https://github.com/locusf/meta-athias-hybris/blob/master/recipes-core/initrdscripts/initramfs-boot-android/init.sh#L54 ?
2016-06-05 15:29:45	@kido	bring it out of the if condition, I really want to stop that boot and get something
2016-06-05 15:30:16	@kido	for now it's quite normal if you don't get any adb shell, no adbd are started *ever*
2016-06-05 15:32:27	@kido	for the g_usb_boot warning, maybe those two commits can help https://android.googlesource.com/kernel/msm/+/28a3009d0f1a59272572fda0eb970aef4e64ab5a%5E!/ https://android.googlesource.com/kernel/msm/+/2a09bf7457039bbe16f95632e15d2bdf107e350a%5E!/
2016-06-05 15:37:21	locusf	it doesn't go anywhere now
2016-06-05 15:37:29	locusf	when uncommenting just adb
2016-06-05 15:37:32	locusf	adbd
2016-06-05 15:37:46	@kido	yep it is supposed to stop your boot process from the ramdisk
2016-06-05 15:37:59	@kido	what does your desktop's dmesg says?
2016-06-05 15:38:06	locusf	nothing
2016-06-05 15:38:10	locusf	no new usb devices
2016-06-05 15:38:25	@kido	screw this
2016-06-05 15:40:02	locusf	yeah
2016-06-05 15:40:13	locusf	gonna go watch tennis now, thanks again for the assistance :)
2016-06-05 15:40:52	@kido	ok, i'm a bit disapointed though
2016-06-05 17:29:03	filtoid	Hi again Kido, I?m going to try and load Asteroid on VirtualBox, does that sound like something that?s possible or is that going to fail badly?
2016-06-05 17:29:42	@kido	welcome back
2016-06-05 17:29:59	filtoid	And is there somewhere with prebuilt images or would I need to compile my own? Just like to save a couple of hours if I can
2016-06-05 17:30:07	filtoid	(as mentioned on the wiki)
2016-06-05 17:30:59	@kido	it won't fail badly (I think that sailfish uses virtualbox in its SDK doesn't it ?)
2016-06-05 17:31:06	filtoid	yes
2016-06-05 17:31:37	filtoid	I think so
2016-06-05 17:32:00	@kido	however you should know that there will be a couple of things to adjust for example on watches asteroid-launcher runs on top of the qt5-qpa-hwcomposer "QPA" (qt platform abstraction) which uses libhybris and android drivers
2016-06-05 17:32:19	@kido	in vbox you will need to use another QPA, I think there is a drm/kms one
2016-06-05 17:32:45	@kido	you will need to recompile everything though
2016-06-05 17:32:59	@kido	we've never used asteroid on x86
2016-06-05 17:33:05	filtoid	ah ok - this may be trickier than I had anticipated
2016-06-05 17:33:19	filtoid	oh right - x64 only?
2016-06-05 17:33:25	@kido	yes it will be trickier
2016-06-05 17:33:27	@kido	no, arm only
2016-06-05 17:33:35	filtoid	oh i see
2016-06-05 17:33:45	filtoid	of course
2016-06-05 17:33:48	@kido	well it's just a matter of specifying the architecture you want to openembedded
2016-06-05 17:34:05	@kido	I know there are already BSP for QEMU inside OE
2016-06-05 17:35:06	@kido	but there must already have been some guys building an OE distro for vbox
2016-06-05 17:38:52	@kido	you might just want to start by playing with a default openembedded build for qemu, this should be pretty straightforward and will bring you a minimal linux that you can play with. once you understand the config you should configure your asteroid build for it and try to compile everything. once every components are built and you get a shell on this you will be able to tweak the asteroid-launcher's config to
2016-06-05 17:38:52	@kido	run on top of drm/kms
2016-06-05 17:40:06	@kido	as I said this is not the easiest task, I think that mcirsta wanted to help on this too but I haven't seen him for a couple of days. if you meet him maybe you could work together I don't know
2016-06-05 17:48:35	filtoid	sorry dropped off for a second there
2016-06-05 17:48:55	filtoid	This looks like a good starting point though, thanks for the pointers
2016-06-05 17:49:43	filtoid	I shall keep an eye out for mcirsta too
2016-06-05 17:50:20	@kido	don't hesitate to ask if you have any question
2016-06-05 17:50:31	filtoid	thanks :) will do