 @UniversalSuperBox, MIR!  MER!  CAPS!
 @wayneoutthere, Did you know MIR (Caps) actually stands for main inclusion request https://wiki.ubuntu.com/MainInclusionProcess :-)
[00:25] <ubports_bot> Project daily-hammerhead build #346: SUCCESS in 11 min: http://ci.ubports.com/job/daily-hammerhead/346/
[00:27] <ubports_bot> Project ota-push build #1140: SUCCESS in 1 min 42 sec: http://ci.ubports.com/job/ota-push/1140/
[00:28] <tgBot> LennyPenny was added by: LennyPenny
 @UniversalSuperBox, YES! I'm 130% with you!
 Welcome @LennyPenny :)
 @ahayzen, I did not.  So may people teac h me so many things here.  Remember with me that a) I don't know anything and b) I don't like everything and c) I talk alot
 @wayneoutthere, terrible attributes to have
 wow.
 If/when ubports picks up Mir 1.0 it won't be Mir or Wayland, it will be Mir and Wayland.
 @LennyPenny, We're all a work in progress....
 @LennyPenny, you must a) know everything b) like everything and c) remain silent? ;)
 meh.
 mirland!
 Did someone say Mirland?
 (Photo, 1280x720) https://irc.ubports.com/sQrQjmS8/file_1093.jpg
 eglplasma on Mir!
 Woooooooooo~
 On a device that has never run Mir before!@
 nice
 @UniversalSuperBox, i kept reading 'eggplant' but now I get it! Nice work, gents
 Take a break from your day and listen to the discussion on Librem5 vs Ubuntu Edge. Will Microsoft kill free software? Give me a break! Find out more on the latest audiocast:  … https://ubports.com/blog/ubports-blog-1/post/ubuntu-touch-audiocast-012-prism-break-84
 That N5 shutdown/reboot issue is very annoying. Phone freezes always. If battery is low and phone tries to shutdown itself, it freezes and display backlight stays on untill battery is dead. If I force reboot, then I have to reboot second time because any app wont open..
 Now there is one core in 100% all time! Damn.
 (Photo, 960x1280) https://irc.ubports.com/LzjRfHkk/file_1095.jpg
 Reboots and shutdowns wont help because it freezes every time. 😢
 Top shows that "wl_event_handle" takes nearly 100% all time.
 (Photo, 960x1280) https://irc.ubports.com/Bxbyrr97/file_1097.jpg
 And off course my wifi died also.. It wont find anything anymore..
[07:01] <ubports_bot> Project click-import build #20: SUCCESS in 5.7 sec: http://ci.ubports.com/job/click-import/20/
[07:01] <ubports_bot> Project vivid-rootfs-armhf build #155: SUCCESS in 1.4 sec: http://ci.ubports.com/job/vivid-rootfs-armhf/155/
[07:01] <ubports_bot> Project ota-push build #1141: SUCCESS in 0.54 sec: http://ci.ubports.com/job/ota-push/1141/
 Hey people! I have the new UBports installed on my BQ E5HD and BQ M10FHD and I can't get LoquilM to verify the code on either device that was sent to my other phone. Any ideas? Code is 3 numbers followed by a minus (-) and 3 more numbers. Nothing happens when I enter xxx-xxx format, but it seems to try to validate when I enter xxxxxx format, but doesn't continue after trying to validate.
 Is there a changelog for release candidate images? Recently installed rc8 (Hammerhead).
 @Jo_Led, It didn't me neither. I made webapp with webapp-creator. Address was something like web.whatsapp.com  … It works fine. No notifications and ui is a bit strange but messages comes and goes..😄
 +work
 (Photo, 449x75) https://irc.ubports.com/X4FqqHAE/file_1099.jpg When you came back and you have 5 thousands messages to read :))
 😄
 @Jo_Led, I think you should join their Telegram group man and ask there for help
 @Jo_Led, You can use this link: https://telegram.me/joinchat/BlpqfQk1L0l0PJTnM02yxQ to join that group
 @ebetonro, Thank you for that!
 @Jo_Led, You're welcome!
 @samitormanen, This sounds great. Could you please share more info about it? Thanks
 I was thinking in installing Firefox to try there the web interface of what's app
 It works with ubports web browser too. Webapp-creator is available in openstore
 😐
 Thank you sir. I was now aware about that
[08:19] <tgBot> <Milan Korecky> @wayneoutthere & @joeinhere very nice audiocast again, thanks a lot, I enjoy and look forward to listen next one like a small child. You have the skills to pick up the important topics for deeper thought!!! 👍
 @samitormanen, I remember you have to scan the code in order to use it. I will have to explore more about the right procedure
[08:23] <ubports_bot> Project daily-fp2 build #364: SUCCESS in 13 min: http://ci.ubports.com/job/daily-fp2/364/
 Yes, i scanned the code with my tablet
[08:24] <ubports_bot> Project ota-push build #1142: SUCCESS in 1 min 5 sec: http://ci.ubports.com/job/ota-push/1142/
[08:31] <tgBot> <Stefan Kalb> @samitormanen, I had a similar issue with my bq e4.5 wasn't able to see any wifi network. I gave up digging and reflashed with wipe and just backuped photos and sms db. After that it worked again.
[09:19] <xorpad> has no one ever made an arm64 version of ubuntu touch?
[09:20] <xorpad> the build tree that is used for making it only has armv7a toolchains in it
[09:37] <tgBot> <Milan Korecky> @Stefan Kalb, Ben
[09:43] <tgBot> <Milan Korecky> Did it worked the rm command Alexander sent you to restore your WiFi connection?
 Off topic. The UBports podcast is a great way to spend your time and work
 At*
 @alan_griffiths, nice! so, something like mir server talking with mir and waylands clients? mir just need to understand the wayland protocol? xml stuff? me noob X-)
 @popescu_sorin, I would prefer more people actually drawing API and stack documents on a piece of paper, its much easier to comprehend graphically. than in text
 a visual representation help, indeed
 Who has been working on bullhead support other than myself?
 @Xorpad, Erm since we took over from Canonical I dont remember any names
 Before neither ^^
[11:14] <xorpad> well, if anyone has done any work on it, that person should contact me via telegram or irc
[11:17] <xorpad> I got the kernel to compile, now i need a good recovery, then i'll be building and arm64 rootfs i guess?
[11:17] <xorpad> the kernel works, not just compiles:D I can boot lineage with it
 cool ^^
 actually its @UniversalSuperBox and @mariogrip what @bhushanshah just told me (I like to mention people haha)
[11:23] <bshah> (and I worked on halium port for it  :P)
 Cool, do you have anything other than the kernel working so I don't waste time doing work that's already done?
[11:25] <xorpad> the kernel took me like 10 hours to get to compile, It's a mix of Halium's kernel, oreo msm kernel, franco kernel, and some things from mainline linux
[11:26] <xorpad> pretty much I just took the best of all i knew were good
[11:30] <bshah> if you are aiming to run ubports, halium kernel is your best bet
 Yep, but I'm aiming for more than just getting it to run, I don't think doing the bare minimum is enough by my standards
[11:32] <bshah> then I don't know, good luck
 Heh... It's more about time and patience and research than luck.... Although luck might prevent some bricked devices during the process haha
 UBports will favor any Halium-compatible build for shipment, as we really need to build a common platform. So, best bet is to start with that resource, and from there you can patch/modify the kernel as you like. But it should be buildable with our toolchain
 We cnanot make special cases for each device
 Which is why I made sure I asked them what needs to be done to comply with their standards
 They didn't have many requirements apparently, so I just made sure to support every feature they do. And it compiled with the build tree I got while following the wiki
[11:40] <bshah> the whole point of halium is you don't touch Halium kernel and android parts
[11:40] <bshah> at all..
[11:40] <bshah> *just rootfs*
 Okay then
 *shrugs*
 Well @bhushanshah you know how the vendors modded their kernels, probably it will make sense or even be necessary to still have device-specific stuff
 Guess I just wasted 18 hours
 and I agree that battery saving can be done there, which would be great for the N5 for example
[11:41] <bshah> @Flohack, about kernel patching we had this discussion, idea was, you can "enable" some config options if you need them
[11:42] <bshah> e.g. ubuntu needs apparmor, they can enable it
[11:42] <bshah> but default kernel keeps it disabled
 Huh what is the "default" now? Vanilla vendor kernel ?
 I hope not
[11:43] <bshah> Flohack, defconfig provided by halium I mean
 Yeah we ar enot talking about config options, thats clear
 We are talking about modding the kernel
 Of course the config options must be standardized
[11:44] <bshah> what kind of "mods" you are thinking about?
 @xorpad, @bhushanshah Apparently its possible to mix certain kernel parts together =) - To me, if everything works, and we got goodies like better performance, less battery drain etc, then we should investigate this road of course.
 Sorry to have to go now, I need to enter our server room for some time
 well, what you are talking about is backporting things...
 that is fine sure..
 But it should not be ubports specific then
 The problem is I can only get docs on the man SoCs so I can't do the same things to devices based on non-msm kernels
 There's this problem with Android having like 20 kernels that are all different
 @xorpad, arm64 patches are incoming.. there is still a problem in the build setup of the kernel .. the other parts are building..
[11:50] <bshah> xorpad: honestly resource you are putting in this can be better spent with getting devices working with "mainline" kernel...
 @xorpad, whats the point of that? then you could have used the stock vendor kernel..
 The point is that it's better than the vendor kernel. It has lots of governers for CPU and GPU and userspace hooks to customize them
 Many things
 Whatever though it's not good enough that's fine
 People seemed excited when I talked about doing this stuff a week ago
 @Milan Korecky, Yeah I will reinstall it :) I dont have much on it. So no backup needed
 If anyone told me what I'm being told now when I got here and said I could and would do it, I would have saved me a lot of hours
 I think getting Halium and THEN ubports to work on the device is hard enough work, but getting a cool specced kernel is better
 i mean - the point of this all wasn't to have a shared kernel between all ubports or halium devices right?
 kernel code is always device specific?
 @Xorpad your stuff is still cool....... BUT, question is where we want to prioritize work?
 if i speak for myself, i can't even get my kernel to boot halium 😔
 Well, I'm a kernel and driver guy... That's what I do best
 I mean, I'll just...post this on XDA ot something
 do that ... or make sure it works in the build infrastructure .. and works with the rootfs .. so that it can be integrated when everything works..
 Well now I'm confused... You guys come to a consensus on stuff I guess before I keep on this track
 It has every feature the halium kernel does, just lots more and more optimized
 If it works with Halium systems and better than the stock kernel, it could be included, right @bhushanshah ?
 @UniversalSuperBox depends, for instance I am "against" including anything not-in mainline torvalds tree..
 (for instance parts taken from android kernel)
 Okay so you guys should have a little meeting I think and decide what's okay for me to do and what's not... If it's possible to code it, than I probably can do it
 Otherwise I don't see a point in working on this stuff if you don't want it
 I've spent over 100 hours r doing things related to this project and it seems like most of it has been wasted time
 So, it is backports from Android kernels rather than mainline?
 It is mainline, Android kernels, custom kernels, and I started with the halium kernel and kept everything it has
 aha
 For the record though I said I was going to do this and people were nothing but excited... It doesn't exactly make me too thrilled about doing more work that won't be welcome
 @bhushanshah, But the basis of any device kernel is some android stuff
 Are the patches (Minus voltage, custom governers) somewhat portable between devices?  If so, I'm sure there's room for those patches living somewhere accessible.
[12:11] <xorpad> @UniversalSuperBox most of them, I'm really not 100% sure how much variance there is in the chips we will have to deal with
 Can you push your work somewhere? so we can actually see that?
 Because currently it feels like we are just commenting on things from peanut gallery :P
[12:14] <xorpad> Okay, I'll upload it to gitlab
[12:14] <xorpad> but my upload speed is garbage
[12:15] <xorpad> like 6mbit I'm suppposed to get, but I get no where close
[12:15] <bshah> hint: fork base repo (let's say halium kernel repo first) and then git push branch on that remote
[12:15] <bshah> will be quicker
[12:30] <xorpad> Now you're making me show my work before it's anywhere near ready which I don't like, but it's uploading now
[12:30] <xorpad> 70kb/s lulz
[12:31] <xorpad> So what is there i can work on that you guys will actually want?
[12:32] <xorpad> because like, i'm a little frustrated right now
[12:33] <xorpad> 3% uploaded
 I'm not responding yet because I see other people typing...
[12:37] <xorpad> okay
 @xorpad, You need to reset your expectations. This is a diverse group of people from casual users to core developers. Each has their own priorities. You work isn't unwanted, but care is needed in deciding what to support with limited resources. What you see now is that stuff that works across all phones is easier to justify supporting than device specific optimisations.
[12:39] <xorpad> my expectations were based on people being excited about me saying what I was going to do, then me doing it
 And when we see the code rather than hypothetical, we can decide.
[12:41] <xorpad> I can't figure out how to push it to github after making it on gitlab
[12:42] <xorpad> it failed after resolving deltas
[12:43] <xorpad> How do you move a project that's already set up with git to another server?
 you add another remove
 some fraction of 800+ were excited, some fraction (who need to support it going forward) want to evaluate the code before accepting it. That's normal.
 and push..
 or pull  ..
 If the code is all on your gitlab server already, though, you can have Github pull from Gitlab.
 By using the 'import repository'
[12:45] <xorpad> well, it's not on gitlab because I cancelled the push at 6% when you said to fork halium and upload based on that
 Aha
[12:45] <xorpad> and now I can't change the remote repoisitory
 You said you started from Halium's repo?
[12:45] <xorpad> I did
[12:46] <xorpad> But I downloaded it as a zip because git pulls the whole history
 Ah, there's the problem.
[12:46] <xorpad> okay
[12:46] <xorpad> well
 Because now you don't have the history, Git can't figure out what pices of code have or haven't changed.
[12:46] <xorpad> I see
 If you know what branch you're trying to clone from, you can do `git clone -b mybranch --single-branch [repo URL]` to reduce the time it takes to get the history.
[12:48] <xorpad> oh
[12:48] <xorpad> See, I don't know that
[12:48] <xorpad> I usually code on my own
 Git is a wonderful and terrifying beast
 But mostly terrifying.
[12:48] <xorpad> apparently it's terrifying me because I can't get it to forget the gitlab thing
[12:48] <xorpad> even if I delete all the config files
 I'd clone into a new folder to avoid overwriting your old changes
[12:49] <xorpad> erm... what would overwrite my changes?
 Then merge your changes into the new folder and commit as you go
 Removing a remote like you're trying to wouldn't
[12:50] <xorpad> okay good
 I'd stop messing with it, though.
[12:50] <xorpad> okay...
[12:50] <xorpad> I'm gonna make a tarball and put it somewhere safe before i keep trying
[12:53] <xorpad> oh... I see the problem
[12:53] <xorpad> well one of them
[12:53] <xorpad> this isn't the right kernel base
[12:54] <xorpad> I thought i started with halium but in the .git/config file, I can see that this is from franco kernel
[12:54] <xorpad> ugh
 Whelp
[12:54] <xorpad> I guess I start over or work backwards or something
[12:54] <xorpad> I don't know
 Hmm.
[12:55] <ubports_bot> Project daily-mako build #139: SUCCESS in 12 min: http://ci.ubports.com/job/daily-mako/139/
[12:55] <ubports_bot> Project ota-push build #1143: SUCCESS in 0.59 sec: http://ci.ubports.com/job/ota-push/1143/
 Starting from Halium would be the best option since that's where it'd ultimately be merged.
[12:55] <xorpad> Okay
[12:55] <xorpad> I'm gonna make a tool to analyze changes between kernels
 There's always `diff`, too, but it'd be suboptimal with that much code.
[12:56] <xorpad> yeah
[12:56] <xorpad> Now I'm frustrated cause not only is this not what you want, i screwed it up
 Bah, take a break then. No point in getting frustrated over it.
[12:57] <xorpad> but I've only been coding for 22 hours
[12:57] <xorpad> lulz
[12:57]  * xorpad doesn't sleep much
[12:57] <xorpad> or have a life
 I've heard 'never code drunk/tired/high'. But coding frustrated is dangerous too
[12:58] <xorpad> You never heard of the balmer peak?
[12:58] <xorpad> re coding drunk
[12:58] <xorpad> https://xkcd.com/323/
 Remember Windows ME?
 Bah, you ninja'd me.
[12:58] <xorpad> hehe
 Anyway, I'll be right back. Reconvene later?
 @xorpad, Lool
[12:59] <xorpad> sure
[12:59] <xorpad> I'm gonna do something else, possibly something more productive
[13:00] <xorpad> I'll work on the recovery
[13:00] <xorpad> or something
[13:01] <xorpad> I don't know
[13:01] <xorpad> I'll... eat food... I haven't done that in a long time
 review the things that make this hard. Pick one simple issue and fix that on its own.
[13:02] <xorpad> Okay...
[13:03] <xorpad> one of the things that makes this hard is that I'm not used to working with other people
[13:04] <xorpad> My last OS may have taken 6 years just to become basic functionality, but I did it all myself and never had to deal with other people caring
 I get that, my password manager app only became cool after about half a year but it's awesome to have your own stuff with no dependency on others
 A good rule for working with people: http://retrospectives.com/pages/retroPrimeDirective.html
[13:06] <xorpad> is that telling me I should go back to working alone?
[13:08] <xorpad> Okay, I found the instructions for helping port the bullhead
[13:08] <xorpad> I'll just... follow those
[13:09] <xorpad> I don't really know
[13:09] <xorpad> i'm feeling like this is the wrong project for me
 Don't be discouraged! It sounds to me that you have useful ideas and could make a difference, but you need to understand what others need to achieve their objectives.
[13:13] <xorpad> well, there's nothing stopping you from using things I make
[13:16] <xorpad> I don't even know which of these kernels is the halium one
[13:16] <xorpad> but i know I don't want to just compile the halium kernel and move on
 https://github.com/halium?utf8=%E2%9C%93&q=kernel&type=&language=
 That's all of the Halium kernels
[13:18] <xorpad> most of them are msm devices
 Yep
 Nothing mtk yet because mtk is really hard
 To but it simply
[13:19] <xorpad> anything is hard when you don't have the hardware documentation
 Indeed.
[13:19] <xorpad> You can't write a driver or kernel module without the documentation unless you just try random things and figure out what they do after you run the code
 That's commonly what we do for things like the camera.
[13:19] <xorpad> which, I've only ever heard of one person doing on one project
[13:20] <xorpad> really?
 Well, it's not from scratch
 It's just looking at how it segfaulted this time and work around that.
 Damned Bionic drivers
[13:20] <xorpad> I knew a guy who made an intel gpu driver by making a kernel that logged everything and throwing random instructions at it
[13:21] <xorpad> I think he said he bricked over 500 netbooks before the driver was stable and complete
[13:21] <xorpad> crazy guy
 he must have had a crazy good deal on netbooks
[13:22] <xorpad> or he was just crazy and didn't care
 imagine how much beer he could have bought for that
[13:23] <xorpad> well, it looks like the halium kernel is literally just the cm 14.1 kernel, and franco kernel is completely comaptible with cm 14.1
[13:23] <xorpad> so...
[13:23] <xorpad> maybe I can work with what I started?
 The problem is getting Git to be able to play all of the changes one over the other
[13:24] <xorpad> Yeah, I was thinking to make a python script, that indexed every function of every file, and then prepared a comparison report
 Mismatched commits in either your or our history (either out of order or missing / added) means Git can't resolve.
[13:25] <xorpad> and for comparing functions i could just use diff
[13:25] <xorpad> but for finding things and knowing what to comapre with what, the python script would create a database of the stuff
[13:25] <xorpad> if would be useful for you to analyze what things i've done
[13:25] <xorpad> and it would be useful for me to find things I want to backport or merge
 I'm not a kernel developer, so I'll take your word on what is or is not useful to a kernel developer.
 If you find any CVEs that apply in our kernel that haven't been patched, those are always really nice to backport/merge. ;)
[13:27] <xorpad> it's more what's useful for the situation we have that what's useful as a kernel developer
 But one important thing is that we'll need a good commit history if you're adding a lot of features
[13:28] <xorpad> well, I made a folder, inside my ubports folder, called new-start
[13:28] <xorpad> so now I'm started over
[13:28] <xorpad> and I will upload as soon as I put together the build tree
 There will be some things that you've worked on that don't apply to Halium, or would be dangerous to merge in.
[13:28] <xorpad> well, perhaps that is possible
[13:29] <xorpad> I can't say for sure
[13:29] <xorpad> there's a reason I don't just upload stuff like this as I'm working
 I'd recommend forking the Halium kernel you're starting from, cloning your own fork, and starting there.
[13:29] <xorpad> but i'll make a private repo on gitlab, and keep track of everything, I'll make commits obsessively
[13:29] <xorpad> yep, I cloned it already
[13:30] <xorpad> now I'm cloning the build tree the ubports website recommends to use for porting to bullhead
 Oh jeez, are you looking at the old porting information on the wiki?
[13:30] <xorpad> then I'm gonna put it all together, and create a repo on gitlab
[13:30] <xorpad> I guess it might be old
 The one that ends with your kernel out-of-tree?
[13:30] <xorpad> I don't know
 This page? https://wiki.ubports.com/wiki/Nexus-5X-Developer-Information?hl=bullhead
[13:31] <xorpad> yes that
[13:31] <xorpad> should I ignore it?
 You should.
[13:31] <xorpad> okay
[13:31] <xorpad> should add a thing that says when the page was last updated on all the wiki pages
 Way down at the bottom, "Updated by -blank-"
[13:32] <xorpad> ahh, 10 months ago
 I was really proud of myself. Got all the way to the kernel not panicking every time it booted. :P
[13:33] <xorpad> :D
 But anyway, that's old information now. This kernel: https://github.com/Halium/android_kernel_lge_bullhead/tree/halium-7.1
 And the halium-7.1 branch
[13:33] <xorpad> yeah I'm pulling that
 Is where you want to start from
[13:33] <xorpad> oops
[13:33] <xorpad> wrong branch
[13:33] <xorpad> I got the default branch, which is cm-14.1
[13:34] <xorpad> it's mostly behind the lineage kernel in terms of commits
 git checkout halium-7.1
[13:34] <xorpad> not sure it really makes a difference between the 2, since 14.1 is 7.1
 halium-7.1 is ahead and it has patches for Halium booting
[13:36] <xorpad> okay
[13:36] <xorpad> does it have kexec hardboot support?
[13:36] <xorpad> wait you prob don't know
 attempts to defer to bshah
 Any idea when we flash a phone, on which partition the files are pushed ?
 thanks
 @xorpad, If the upstream kernel doesn't have kexec, the downstream one doesn't either.
[13:38] <xorpad> it supports kexec
[13:38] <xorpad> we need that
 @dal
 @Kev
 Well now it's not funny any more
 each time i push enter ... instead of clicking on it...
 You can also use Tab
 Ah thanks
 Any idea, for this guy : https://forums.ubports.com/topic/650/nexus-4-flash-hangs
 Ugh, udf is such a pain.
[13:47] <xorpad> is there a 7.1 halium build tree I should use?
 Could you have him try the UBports installer? It's at least a little more robust
 And doesn't get stuck when bootstrap is enabled
[13:47] <xorpad> magic-device-tool works great
 @xorpad, You want the full tree for Halium?
 Er
 For Bullhead?
[13:48] <xorpad> I guess... I don't know, I need a build tree, unless you don't mind me building in cm-14.1?
 @UniversalSuperBox, I've already submitted this option to the person., let see
[13:48] <xorpad> the build tree I used last time wasn't proper for compiling 64 bit arm code
 @xorpad, Well, I assume you know repo and everything, so...
[13:48] <xorpad> so I had to pull in tons of stuff from cm-14.1 anyways
 (Document) https://irc.ubports.com/0DN0PR0P/file_1100.xml
[13:49] <xorpad> yeah but I'm asking what build tree I should use
 wasn't the tree based on CM 14.1 ?
[13:49] <xorpad> no, it was based on lolipop aosp
 You'll want to `repo init -b halium-7.1 https://github.com/Halium/android/tree/halium-7.1`
[13:49] <xorpad> the one I had, but it was from the nexus 4 guide
 That is not a repo command.
[13:50] <xorpad> i'm confused
[13:50] <xorpad> you give me a command that is repo, and tell me it's not a repo command... Am i just misreading?
 I put the wrong link in there.
[13:50] <xorpad> oh
 repo init -b halium-7.1 -u https://github.com/Halium/android/tree/halium-7.1
 Untested, but that should be it.
 WHAT
 It didn't paste the right link.
 https://github.com/Halium/android.git
 Mumbles something about copy-paste
[13:52] <xorpad> lulz
[13:52] <xorpad> just let me download it, I know how to clone a git repo
[13:52] <xorpad> that's about all i know about igt
[13:52] <xorpad> git
 Oh, that's not the full tree. It's just the manifest that tells the repo tool all of the projects in the full tree
[13:53] <xorpad> okay, that, i don't know how to use
 How were you getting Android trees before?
[13:54] <xorpad> from the android repository
[13:54] <xorpad> but I don't know how to use an xml file to grab using repo
 repo init -u https://github.com/Halium/android -b halium-7.1
 There we go
[13:55] <xorpad> okay
 Starting at http://docs.halium.org/en/latest/porting/index.html#step-2-create-a-new-directory-to-download-the-halium-tree
[13:55] <xorpad> okay sure:D
 And for the local manifest, use the bullhead.xml that I linked earlier. Unless @bhushanshah has one that isn't missing what i probably am.
 I might have changed things in order to build for potter.
[13:56] <xorpad> I don't know, but I need to drink more redbull
[13:56] <xorpad> while this is syncing
[14:00] <xorpad> ugh, I need something to do, this syncing is not happening fast
[14:00] <xorpad> I thought it was a minimal build tree
 It is...
[14:01] <xorpad> it's only at 9%
 Android is big.
 Even tiny android
[14:01] <xorpad> ahh
[14:02] <xorpad> AOSP including prebuilts is like 12GB, but to sync it with repo you need like 60gb for mashmallow
[14:02] <xorpad> because of the .repo folder and such
[14:02] <xorpad> I used to have a full mirror of AOSP locally, but I can't find my external drive
 My Halium folder, with pieces for both Potter and Bullhead, is 29 gigs
[14:02] <xorpad> well I still have it.... somewhere
[14:03] <xorpad> I wish I could find that drive
[14:03] <xorpad> It has 220gb of AOSP lol
[14:06] <tgBot> Nioury was added by: epierre
[14:06] <xorpad> 54/197
[14:19] <xorpad> okay, it's at 92%
[14:19] <xorpad> soon I can start
[14:20] <xorpad> or rather start over
[14:20] <xorpad> Maybe I'll just port it and do the fancy stuff later
[14:20] <xorpad> work on the rootfs or something like that
 I have the steps to start testing Mir. Trying to eek more out of Marius. :P
 ^^
 I think this is our bottleneck, and where all hands are needed, getting 16.04 to boot from Halium with Mir and unity8 not broken
 But first you check if Mir is working, then if qtMir is working
 @UniversalSuperBox, speaking about mir, will start the build of qtmir with 7.1 headers
 Oh, that's why it didn't work
 mumbles something about segfaults
[14:33] <xorpad> segfaults lulz
[14:34] <xorpad> Ryzen much?
 Everything segfaults.
 Everything.
[14:34] <xorpad> My ryzen used to segfault like crazy before I RMA'd it
[14:34] <xorpad> my threadripper i still need to send in
[14:35] <xorpad> But I don't want to go without it while i wait for a replacement
 threadrippers weren't affected by the segfault problem...
 Mine is
 Anything made before July 1 is, so very few TRs
 Mostly the ryzens because they made them for like  month before fixing the fabrication
 Huh. … I read that threadrippers were never affected by it,my bad.
 Bad luck I guess.
 I don't know where you read that but my 1950 def segfaults
 Yeah... The worst part was they knew about the fabrication issue since early June, yet still released the chips hoping they could patch it away with microcode
 Not buying amd again I think
 3 amd chips in row with critical bugs
 Only one bad Intel chip... If
 Intel is expensive though
 What other AMD chips had issues like this?
 Not like this, but my fx-8120 had an issue with scheduling the shared fpus, leading to complete instability of the system when running a large variety of software, especially games
 It took 6 months for them to make the microcode patch, but another 6 for the patch to make it into my motherboard firmware, so the box I bought for Windows gaming took a year before I could game on it
 After the first 6 months I could use Linux because it patches microcode at boot time
 Fair enough. … But, eh, pinning AMD on this seems a bit disingenous. Intel has issues too, there was the Skylake corner case with their SMT, there was the Atom C2000 issue where some clock pin would fail preventing a system from booting entirely.  … I think that last one meant a recall of a ton of embedded systems. … Both manufacturers screw up.
[14:53] <xorpad> Well AMD didn't do a recall
 This didn't cause a bunch of systems to be unbootable.
[14:53] <xorpad> they are accepting RMA's, so unless you do cpu bound video encoding, or compile stuff on linux, you won't know there's an issue and won't RMA it
 iirc it was also worked around in a recent AGESA version too.
[14:53] <xorpad> which means if you decide after warranty period to do something and have a problem, you're screwed
[14:54] <xorpad> AMD not doing a full recall is upsetting to me
[14:54] <xorpad> it's like, they are taking the easy way out and some people will later down the road be screwed when they change the way they use their system
[14:56] <xorpad> I've tried every workaround possible, and my motherboard firmware has the latest AGESA version on my ryzen 1700x, which was good for OC'ing my ram up to 4000mhz but useless for the segfaults
 Ooh, they fixed the FMA bug. My bad.
 Yeah... There is one thing that works... Install Gentoo, then compile gcc 6.3 or 6.4, then use gcc to recompile every package in the system. The way those 2 gcc versions compile everything stops the segfaults for some mystery reason
 Alright. Fair enough on those points. … I had to go and double check what I read a couple months ago. … I have an R5 1600 manufactured before July and ran the kill_ryzen script for a couple hours and it never did anything unusual. I guess I just lucked out.
[15:03] <xorpad> Only certain models were affected, not sure which ones execpt the ones i own
[15:04] <xorpad> and i bought the ryzen 1700x so i could develop linux kernel modules for it
[15:04] <xorpad> like, I bought it as a challenge lol
 What a strange situation.
[15:06] <xorpad> Okay, I got the halium 7.1 build tree, it's done downloading, it's syncing the working treee then it needs to check out the files
[15:07] <xorpad> it's not checking out the files like repo usuallyu does
[15:09] <xorpad> oh, it checked them out into the folder that I was inside a folder within, so it checked out the files to ../
[15:09] <xorpad> lol
[15:31] <xorpad> okay, so first thing is... it doesn't build lol... at least not the standard way aosp/cm/lineage do
[15:31] <xorpad> but I'm past that now, on to the next issue
[15:49] <xorpad> the whole build system is messed up, makefiles should do things in order if you don't specify something, but this is backwards
 Any build instructions you need are in the Halium docs
[15:49] <xorpad> thanks universalsuperbox
[16:31] <xorpad> the reuqired files for bullhead seem to be not getting grabbed
[16:36] <xorpad> I can probably pull the files from AOSP driver binaries
[16:36] <xorpad> since it's the hardare/lge folder I can't get
 @Flohack, Does this hel? https://community.ubuntu.com/t/mir-architecture/467
 @alan_griffiths, Yes somehow 😆
 great visualisation! :D
 Is there a way to inject css styles into ogra webapps?
 I really wish webapps could look more native
 I can try doing that sometime
 once I get the SDK to work on 17.04..
 You can do that with a `UserScript` on a custom `WebContext`
 @DanChapman, Oh, is there a write up of that anywhere?
 Marius should really use some of that patreon dosh for a programming competition
 @padraic7a, There's nothing official on using them but take a look at the "How can i add userscripts" section here https://dpniel.gitlab.io/blog/ubuntu-webview/ and also take a look at dekko it injects a custom style for converted plaintext messages https://gitlab.com/dekkoproject/ubuntu-ui-plugin/tree/master/plugins/core/mail/webview
 Thanks!
 https://gitlab.com/dekkoproject/ubuntu-ui-plugin/blob/master/plugins/core/mail/webview/userscripts/font_user_script.js being the one that applies the css.
 @padraic7a, 👍
 @samzn, what do you have in mind?
 Any plans for supporting Librem 5?
 This could be great for promoting Ubuntu Touch
 Just curious to know!! Who gets benefited from the librem 5 is it manufacturer or kde? Who gets more users?
 We've contacted them. Else than that, no comment.
 @neothethird, Anything to fill the app gap and bring more users in
 But as long as it has a practical use
 And the community decides which app is more valuable to them
[18:48] <xorpad> the app gap will be filled by the time the product is marketable and worthy of being shipped with devices
[18:48] <xorpad> which is a long time from now
 I think a model that trickles down can work well and incentive people from all levels to try to participate
 It's complicated xorpad
[18:49] <xorpad> my experience has been that people will do the same work wether you pay them or not when it comes to opensoruce software
[18:50] <xorpad> I've seen people do projects that there was rewards for finishing, and turn down the money
 Sure, but incentives can always bring new people in or make the people already here more interested.
 I think it's definitely something to continue thinking about.
[18:50] <xorpad> so put bounties on features
[18:51] <xorpad> people can contribute to they bounty pool for features they want
[18:51] <xorpad> so that the money for implementing it gets bigger as more people want it
[18:52] <xorpad> Most FOSS coders just do whatever they want, they don't care about a reward
[18:52] <xorpad> there is surely some freelancers who do
[18:53] <xorpad> but the majority of people who do stuff like this don't do it for money and would rather see the project have the money for something else
[18:53] <xorpad> like, I'd rather see a developer get a new phone to port ubuntu touch to, than me get paid for work I do
[18:53] <xorpad> and i'm not alone in that philosophy
 Yes, a lot of people agree with you.
[18:56] <xorpad> man, another day goes by...
[19:04] <xorpad> You know what we need
[19:04] <xorpad> a power saving mode
 xorpad: patches welcome ^^ But what would you want to disable?
[19:07] <xorpad> @NeoTheThird disable cores, scale down cpu frequency, optionally restrict network traffic from apps not in the foreground, lower brightness, close apps that are switched away from... We can add lots of features and a checklist for which ones the user wants to have the switch enable
[19:07] <xorpad> just a tought
[19:08] <xorpad> I mean, android and ios do most of those things, with the exception of closing background apps
[19:08] <xorpad> also, background apps can be hibernated to NAND storage
 xorpad: sounds good ;) feel free to send a pr
[19:09] <xorpad> and pulled back into ram and re-executed where they left of
 Ubuntu Touch aggressively kills background apps, does it not?
 we don't have background apps either way, though
 @UniversalSuperBox, yeah
 @xorpad, I think ZRAM does that
 Ubuntu touch also freezes it a lot
 Zram is compressed swap in ram
 xorpad, I agree with you. Power saving needed.. 👍
 I had a look at yunit and i was happily surprised everything was QML
[19:10] <xorpad> compressed swap imo is a negative for power saving, because you have to use cpu power to compress and decompress it
[19:10] <xorpad> it sort of limits the enhancements you get
[19:10] <xorpad> it probably helps
 @samzn, Would you like me to disappoint you?
[19:10] <xorpad> but not as much as it could if it was not compressed
 @UniversalSuperBox, sure
 @samzn, Canonical started a project to rewrite the QML in C++ for performance reasons.
 Well, it's canned either way
 @UniversalSuperBox, which makes sense for many parts
 During winter break I want to rewrite it to wayland over qtwayland or even base it off lipstick
 at least to get minimal functionality
 Mir is a dead end even if we use wayland shimming
 Mir isn't dead... It's still being massaged in some special use cases. See latest ask Noah show from Ubuntu rally
 I'm aware of it, but it's going to be used in embedded cases, not general computing scenarios
 Right
 Which is what ubports is :P
 unifying to be as close to mainline ubuntu & halium as possible should be a goal
 it'll save a lot of engineering time down the road imo
[19:15] <xorpad> if mir is abandoned officially, then we can adopt it
 Even whith Mir living on, trying to move to wayland makes sense. After all, we're building an operating system with convergence as a goal, so we should try to stay close to whatever GNU/Linux on the desktop is doing. … Either way, it should be an interesting experiment, @samzn. Please keep us in the loop ;)
[19:24] <xorpad> xfree86
[19:24] <xorpad> lulz
 Mir is neither abandoned, nor a dead end. It is an open source project where the major contributor is focussing on certain usecases. But not to the exclusion of everything else.
 I should rephrase
 I would like if Mir gets broad support for general use cases but if it doesn't, I think it's another layer of engineering that needs to be done for future libraries and programs to support Mir, maybe doable for a team of canonical engineers but tough for a small team of semi-volunteer developers
 Unity 8 currently delivers undoubtly the best foss handset UX
[19:31] <xorpad> It does, and unlike Ubuntu's desktop Unity it isn't overly bloated with stuff it doesn't need
[19:31] <xorpad> It could still use optimizing, but it's not full of features that are obviously a waste of resources
 I worked hard on Nemo Lipstick UX with a group of talented people and it's still missing a lot of what could be considered essential for a normal user
 (Unity 8 is missing N9 gestures though :) )
 Much of what U8 delivers is based on the support Mir *already* provides for "general use cases". It would be tough for a small team of semi-volunteer developers to implement that on another engine.
 good point
 @alan_griffiths, Sure, and that is appreciated by us of course. Since Ubuntu Touch is so closely tied to it at the moment, even if there were immediate plans to remove it from our system, we could not do it. But in comparison to its goals, UBports is very limited in what its developers can pull off on their own, so it does make sense for us to at least look into ways to take some load off of our own shoulders. I know you are still doing great work
[19:34] <tgBot> on mir, and you have some interesting stuff cooking as well :)
[19:34] <xorpad> There's no reason we have to only have one UI
[19:35] <xorpad> if they run on the same graphical environment it should be trivial for users to swtich
 @neothethird, Understood. I just want Mir to be viewed as a friendly upstream, not an abandoned dead end.
 Sorry, I used poor words
 @alan_griffiths, Yeah, we should probably change our vocabulary in that regard. If anything, a good long-term solution with Mir will be a lot better and easier than a bad one without. Keep up the good work :)
 What's the best cheap low latency miracast device I can buy today?
 Microsoft's?
 Receiver or broadcast?
 Microsoft's is the only one that's likely to work since everyone else got the standard wrong.
[20:08] <xorpad> I use miracast from windows to a Zoomtak 8 Android TV box
[20:09] <xorpad> it's amazing hardware, h265 decoding accellerated
[20:09] <xorpad> but it has android 4.4
[20:09] <xorpad> lulz
[20:09] <xorpad> it's funny to put such old android on such nice hardware
 funny they put android on such nice hardware
[20:10] <xorpad> I also use the app moonlight(root) for remote display using Nvidia GameStream
[20:10] <xorpad> amazing over gigabit ethernet
 kinda asking mostly for ubuntu touch
[20:10] <xorpad> oh
[20:10] <xorpad> well, I'm not porting it to a tv box
 if I'm going to use it full time I'd love to be able to use convergence + usb
 which seems to be only possible with miracast on hammerhead due to the hw :p
 I see
 hm there is a software only miracast too..
[20:32] <tgBot> VolkerDormeyer was added by: VolkerDormeyer
 welcome :)
 Hi
[21:02] <ubports_bot> Project daily-bacon build #296: SUCCESS in 11 min: http://ci.ubports.com/job/daily-bacon/296/
[21:03] <ubports_bot> Project ota-push build #1144: SUCCESS in 56 sec: http://ci.ubports.com/job/ota-push/1144/
[21:19] <tgBot> widecurio64 was added by: widecurio64
 @samitormanen, Have you tried to type this command in the terminal? It normally solves the problem...  … rm .config/connectivity-service/config.ini.lock
 Welcome Emanuele