[00:38] <robert_ancell> RAOF, the faffing comes with the old technology :)
[00:38] <RAOF> :)
[06:28] <seb128> good morning desktopers
[06:32] <duflu> Hi seb128
[06:40] <seb128> hey duflu, how are you?
[06:40] <duflu> seb128, slightly unwell but otherwise good. How are you?
[06:40] <seb128> oh :( get better then!
[06:40] <seb128> I'm good, having still a bit of a cold still though :/
[06:41] <duflu> :\
[06:53] <seb128> cyphermox, hey, do you plan to upstream those recent plymouth patches you upload or do you need help with that? (I'm happy to create gitlab.freedesktop.org PRs for those)
[07:23] <didrocks> hello!
[07:27] <duflu> Hello didrocks
[07:27] <didrocks> hey duflu
=== pstolowski|afk is now known as pstolowski
[08:54] <willcooke> morning gang
[08:55] <willcooke> rebooting, brb
[09:01] <Laney> hey thereeeee
[09:02] <willcooke> morning Laney
[09:02] <didrocks> hey willcooke, Laney
[09:02] <duflu> o/  willcooke, Laney
[09:04] <Laney> OH MAN
[09:04] <Laney> WHAT A LINEUP
[09:04] <Laney> HEY WILLCOOKE DIDROCKS DUFLU YEEEEEEEEEEEAAAAAAHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
[09:04] <duflu> I want we he's having. Half dose
[09:05] <willcooke> Someone's had their Wheatabix
[09:05] <willcooke> * Weetabix
[09:05] <willcooke> (TM)
[09:05] <willcooke> So, I'm trying to verify this one: https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/1784363
[09:05] <ubot5> Ubuntu bug 1784363 in grub2 (Ubuntu Bionic) "High delay when booting (grub waits with a blank purple screen for 10 seconds before booting the kernel)" [High,Fix committed]
[09:06] <willcooke> but it talks about installing grub-pc, and I don't have grub-pc installed
[09:06] <willcooke> is that right?!
[09:07] <Laney> that's the bios one isn't it?
[09:07] <willcooke> ah
[09:07] <willcooke> that would make sense
[09:07] <seb128> hey Laney willcooke
[09:08] <willcooke> I now have grub2 2.02-2ubuntu8.9 and it looks fixed now
[09:08] <seb128> nice
[09:09] <duflu> Speaking of which... seb128 have you looked into the seamless UEFI booting that Red Hat was doing?
[09:09] <seb128> does anyone here has usb-c to help verifying bug #1800715?
[09:09] <ubot5> bug 1800715 in bolt (Ubuntu Bionic) "Prompt for credential when it shouldn't" [Undecided,Fix committed] https://launchpad.net/bugs/1800715
[09:09] <duflu> I've been thinking for years it would be nice
[09:10] <seb128> duflu, yeah, funny you mention it, I was co-working with Hans from RH who works on that a week ago
[09:10] <seb128> he got his changes merged upstream for different component, just not for grub because grub... we should perhaps look at backporting that
[09:11] <seb128> willcooke, oh also, he suggested upstream changes to plymouth to improve the "plymouth is showing for a second" issue, https://gitlab.freedesktop.org/plymouth/plymouth/issues/64
[09:11] <gitbot> plymouth issue 64 in plymouth "[PATCH] Update splash_delay based on previous boot's time" [Bugzilla, Opened]
[09:11] <seb128> I didn't know by today plymouth has a fixed 5 seconds delay before kicking in
[09:11] <seb128> to avoid extra transitions on fast machines, like you don't get it at all
[09:12] <duflu> I was about to ask about that too. Rumors for years have been that plymouth slows boot down
[09:12] <seb128> but then if you boot takes 8 secondes you have it for 3 seconds which is worth than having it for 8 seconds
[09:12] <seb128> that patch he suggested would adapt according to the recorded boot time from the machine
[09:12] <willcooke> oooh
[09:12] <willcooke> nice
[09:14] <duflu> seb128, surely the cleanest solution is to skip plymouth then? For initial boot at least
[09:14] <willcooke> seb128, @ usb-c - no, fraid not.  I will ping YC and/or Jerry and ask them to find someone to test it
[09:15] <duflu> We could just animate below the UEFI/BIOS logo screen until gdm is ready
[09:15] <Laney> willcooke: https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/1784363/comments/14
[09:15] <ubot5> Ubuntu bug 1784363 in grub2 (Ubuntu Bionic) "High delay when booting (grub waits with a blank purple screen for 10 seconds before booting the kernel)" [High,Fix committed]
[09:15] <Laney>    Installed: (none)
[09:15] <willcooke> ?!?!!?!?
[09:15] <willcooke> what did I do
[09:16] <Laney> probably grub2-common is the one
[09:16] <willcooke> yes
[09:16] <willcooke> thank you
[09:16] <willcooke> updated#
[09:17] <seb128> duflu, skip in which cases? also you can't really skip it, it's not only the splash screen but also handling prompting for disk password and stuff
[09:18] <duflu> seb128, I forgot about that. I just mean there are too many transitions. So yes if we can make plymouth not appear at all on fast unencrypted systems then that would be some improvement
[09:19] <willcooke> seb128,  @ bolt - I've pinged OEM
[09:19] <seb128> I tried the fc29 iso on my inspiron, it feels a bit weird. what they do to avoid the transitions is to keep the vendor logo on screen for the boot
[09:19] <willcooke> yeah I saw that too
[09:19] <seb128> it feels weird to me to have the dell logo sitting there for 15 seconds
[09:19] <duflu> Unless plymouth can be ported to render within the bottom half of the UEFI BIOS splash screen
[09:19] <willcooke> ditto
[09:19] <andyrock> morning!
[09:19] <seb128> hey andyrock!
[09:19] <oSoMoN> good morning desktoppers
[09:19] <duflu> Hi andyrock
[09:19] <willcooke> I asked Dell about that too, they said that one /could/ set one's own BIOS logo thingy
[09:19] <andyrock> seb128: are you sure you pushed the branch ubuntu/cosmic yesterday?
[09:20] <seb128> duflu, right, well if your system boots in less than 5 seconds you should already be not seeing it
[09:20] <duflu> seb128, unfortunately I am around 7-9 seconds
[09:20] <willcooke> seb128, also Windows doesn't do that, it has a Windows logo (afaicr)
[09:20] <seb128> andyrock, https://git.launchpad.net/~ubuntu-desktop/ubuntu/+source/gnome-control-center/log/?h=ubuntu/cosmic
[09:20] <seb128> andyrock, unsure why it's not listed on https://code.launchpad.net/~ubuntu-desktop/ubuntu/+source/gnome-control-center/+git/gnome-control-center
[09:20] <seb128> maybe a question for #launchpad
[09:21] <seb128> duflu, you can tweak the timeout in by creating a /usr/share/plymouth/plymouthd.defaults with
[09:21] <seb128> [Daemon]
[09:21] <seb128> ShowDelay=10
[09:21] <duflu> OK
[09:21] <seb128> if you want to try how it feels like
[09:22] <seb128> but yeah, I'm going to nag about resuming the discussion on https://gitlab.freedesktop.org/plymouth/plymouth/issues/64
[09:22] <gitbot> plymouth issue 64 in plymouth "[PATCH] Update splash_delay based on previous boot's time" [Bugzilla, Opened]
[09:23] <duflu> Right. So as I understand it, eventually a fast system will go from the BIOS logo straight to the login screen... ?
[09:23] <duflu> Although it would be nice if some Ubuntu stuff appeared below the BIOS logo
[09:24] <andyrock> seb128: asking
[09:25] <seb128> andyrock, thx, I guess you can probably work with the repo and mp without issue, just an UI bug
[09:25] <andyrock> yeah
[09:26] <seb128> duflu, right, the current implementation does keep the bios logo on screen during the whole boot, which feels a bit weird, but they do plan to add a spinner/animation under the logo
[09:26] <seb128> so it would be
[09:26] <duflu> Yeah. So the machine looks alive
[09:26] <seb128> bios logo -> logo/bg doesn't change but a spinner get added when you reach plymouth -> gdm
[09:26] <willcooke> seb128, duflu - I remember something else Dell said - the background colour on those BIOS logos probably won't be transparent, so merging the logo and the Plymouth output might not be doable
[09:26] <duflu> willcooke, I imagine there's an Intel UEFI spec for it
[09:27] <willcooke> xnox knows the spec :)
[09:27] <willcooke> (if he was here he would be sad now)
[09:27] <seb128> lol
[09:27] <seb128> yeah, there is a spec
[09:28] <seb128> hat described boot path is only on machines which provides what is needed to fetch the logo etc
[09:28] <seb128> the other day Hans was having fun dealing with screen rotations on some devices with small screen which default to portrait mode
[09:29] <duflu> Yeah tablets are often wired natively for portrait
[09:32] <seb128> willcooke, https://fedorapeople.org/~jwrdegoede/flickerfree-videos/workstation-normal.webm is a video of how it currently looks
[09:32] <willcooke> yeah I saw that one.  I don't like it as much as our one with the Ubuntu logo etc
[09:32] <willcooke> it looks like it's not doing anything
[09:33] <willcooke> IMO
[09:33] <seb128> I'm not really convinced that without the animation it's a better experience than what we current have, sure there are less transitions but there is also less feedback
[09:33] <seb128> it looks stucked on bios for a while to me
[09:33] <willcooke> same
[09:33] <duflu> seb128, I agree
[09:33] <seb128> k, so seems like we are in agreement to not try to squeeze that in disco yet
[09:33] <duflu> But I also think plymouth should be reusing the lower part of the screen and not creating extra fullscreen transitions
[09:33] <willcooke> that would be nice
[09:34] <willcooke> seb128, I agree
[09:34] <duflu> Unless it can nicely animate the BIOS logo away
[09:34] <willcooke> or even blank the screen without flickering/changing modes etc
[09:34] <seb128> willcooke, duflu, ah, you can also add "plymouth.splash-delay=0" to the kernel options in grub instead of editing the config, if you want to see the difference without timeout
[09:35] <willcooke> I'll try that later
[09:35] <seb128> duflu, right, I'm unsure animating the logo can be easily done
[09:35] <andyrock> Laney: gnome-control-center 3.30.2-1ubuntu1 is in disco but the git branch is not updated https://git.launchpad.net/~ubuntu-desktop/ubuntu/+source/gnome-control-center/log/
[09:36] <Laney> terrible
[09:36] <duflu> It only depends on whether you can read/write the whole screen, which I imagine you can
[09:36] <seb128> willcooke, duflu, at least one of the changes to to have grub not change the framebuffer if it's not displaying the menu, that bit should land this cycle through upstream updates&co (and maybe we need to backport a grub patch as well) and be a small improvement
[09:37] <duflu> Yes, sounds like a good first step
[09:37] <willcooke> that will be a nice improvement
[09:37] <willcooke> thanks seb128
[09:37] <seb128> np
[09:37] <Laney> there you go
[09:37] <duflu> Though I can't tell people to "Press Esc when you see the blank purple screen" which may be a problem
[09:37] <andyrock> Laney: nice!! thx for the quick fix
[09:38] <Laney> no worries
[09:38] <seb128> andyrock, Colin replied to you on the other channel btw, so yeah UI refresh bug on launchpad
[09:39] <seb128> let's not worry about it, it should pick up when we commit some changes
[09:42]  * Laney did that rescan thing
[09:43] <seb128> which works too
[09:43] <seb128> good than some people are less lazy than me, I would just have waited for the next commit :)
[09:43] <Laney> yeh :p
[09:57] <andyrock> mmm for the SRU (g-c-c/cosmic) should I merge with ubuntu/1%3.30.2-1ubuntu1 or with debian debian/1%3.30.2-1
[09:57] <andyrock> there is a patch that needs to be dropped (the one that lowercase the username)
[09:57] <andyrock> *lowercases
[09:58] <seb128> usually we don't do merges for SRU but stick to the minimal diff
[09:58] <seb128> now if merging the update and taking out some changes is easier for you feel free to do that
[09:58] <seb128> I think that's the result which counts, the way you get to it is up to you, doesn't matter much
[10:00] <andyrock> well it's a new upstream release (3.30.1 to 3.30.2) so we need some merging
[10:00] <seb128> no we don't
[10:00] <seb128> you could gbp import-orig the tarball on the cosmic branch
[10:01] <seb128> no?
[10:01] <seb128> though the import might complain about the pristine-tar already having it
[10:01] <andyrock> let me try
[10:01] <seb128> so yeah, you probably want to at least merge the changes up to the import of the orig from master
[10:02] <seb128> andyrock, ^
[10:02] <seb128> that's probably easier
[10:02] <seb128> then if there are not other extra packaging changes etc it's probably fine/easier to just import ubuntu/1%3.30.2-1ubuntu1
[10:06] <andyrock> so uscan will not work because there is already an upstream tag 'upstream/3.30.2'
[10:06] <andyrock> I guess that merging with 'upstream/3.30.2' would be enough
[10:06] <andyrock> + minimal changes to debian to make it work if required
[10:10] <seb128> andyrock, I would merge eb3c501e
[10:11] <seb128> and then take out the uppercase patch
[10:11] <seb128> 9a196f3f is probably noise for a SRU and going to make the diff more difficult to review so I would keep that out
=== alan_g is now known as alan_g_
[11:24] <andyrock> would be nice to hide changes to .po files in launchpad
[11:24] <andyrock> every diff got 5000+ lines just for that
[11:43] <jbicha> I'm just glad more stuff uses meson (or cmake) so we don't have the autotools clutter in the new release diffs
[11:56] <andyrock> Laney: hi updated https://bugs.launchpad.net/ubuntu/+source/mozjs60/+bug/1796238 to make it SRU compliant
[11:56] <ubot5> Ubuntu bug 1796238 in gjs (Ubuntu Cosmic) "[SRU][regression] mozjs60 crashes with SIGSEGV on gnome-shell exit, in GetPropertyOperation() from Interpret() from js::RunScript()" [Medium,In progress]
[11:57] <andyrock> new gjs + new mutter should make `gnome-shell -replace` work again
[11:58] <andyrock> someone should prepare the upload (using the same packages from disco)
[12:03] <Laney> andyrock: ok, can you make a bug for mutter too please? or convert existing ones
[12:03] <Laney> can upload those later probably
[12:04] <andyrock> kk
[12:05] <Laney> merci :>
=== Class7_ is now known as Class7
[12:52] <andyrock> Laney: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1804448
[12:52] <ubot5> Ubuntu bug 1804448 in mutter (Ubuntu) "SRU Mutter 3.30.2-1 to cosmic" [Undecided,New]
[12:52] <andyrock> I could not find a better bug
[12:52] <andyrock> *good
[12:53] <andyrock> so I created one
[13:04] <cyphermox> seb128: I haven't had time to do it yet, but I will today
[13:16] <seb128> cyphermox, good, thx, let me know if you need help (also would be nice to update to 0.9.5 in disco, that might fix some of our segfault issues who knows, it would for sure make easier to upstream reports than having them coming from an older version with a stack of patches)
[13:19] <Laney> #plymouth exists btw - I found it a good place to discuss when I was writing those other patches
[13:20] <Laney> they ended up better than I would have made them when working in isolation
[13:20] <oSoMoN> willcooke, have you had a chance to install network-manager 1.10.14 on bionic from https://launchpad.net/~osomon/+archive/ubuntu/nm-lp1754671/+packages ?
[13:29] <seb128> cyphermox, ^ in case you are not reading the channel, L_aney comment might be useful to you
[13:31] <willcooke> oSoMoN, installed it, need to reboot
[13:31] <willcooke> which I will do now
[13:34] <willcooke> oSoMoN, well, I'm online :)
[13:34] <willcooke> I will test with some wifis
[13:35] <willcooke> but I'd prefer to give it at least until the start of next week before I declare it OK
[13:35] <oSoMoN> willcooke, yeah, let's not rush this
[13:36] <willcooke> hrm, first problem (but might be my end) I can't authenticate to a wifi network
[13:43] <willcooke> oSoMoN, Yeah, seems like connecting to a wifi network from Settings generates a "Failed to authenticate" notification the first n times, and then goes away.  Feels like maybe it's converting some settings from old to new, and then works afterwards (pure speculation).  I will need to install it on my test machine from fresh again and see.
[13:44] <willcooke> To be clear: it's now working fine though
[13:45] <oSoMoN> hrmm, it should have worked at the first attempt though, that's concerning
[13:46] <willcooke> it might still be me, I was messing with my wifi access points recently
[13:46] <willcooke> so I;d like to try from scratch
[14:08] <cyphermox> seb128: well, I'm all setup for gitlab now; so I'll be submitting the patches/PRs. do you care to be CCed or otherwise notified?
[14:08] <seb128> cyphermox, no, it's fine, I can follow up/subscribe there
[14:09] <cyphermox> ok
[14:10] <seb128> cyphermox, I'm just trying to make sure we do upstream our changes, especially on packages that are not easy to maintain ... we got bitten by old n-m patches that were never applied and where we don't have anyone understanding the code enough to update them and I don't want that to happen again if we can avoid it :)
[14:10] <seb128> also working upstream usually makes sense, they know their code and can help fixing things the right way or spot errors
[14:25] <cyphermox> of course
[14:26] <cyphermox> fwiw the NM patches you're referring to are completely unapplicable, and I don't understand the code any better to make them work
[14:26] <cyphermox> seb128: !10, !11, !12
[14:29] <seb128> ?
[14:29] <seb128> cyphermox, and yeah, I was not especially blaming you for those, just saying that not upstreaming them at the time was an error which has been bitten us back later
[14:30] <cyphermox> *shrugs* it happens; discussion on IRC didn't lead at the time to a patch that was quite ready to upstream
[14:31] <cyphermox> ok, already closed one of the plymouth PRs, it's well covered by Hans' PR, and I like the solution better
[14:32] <seb128> :)
[14:34] <seb128> cyphermox, thx for upstreaming those :)
[14:36] <cyphermox> I really don't understand why the other guy still has crashes
[14:36] <cyphermox> I spent a whole lot of time yesterday looking at the ASM for the functions, and figuring out exactly how it's blowing up
[14:37] <cyphermox> I can't make sense of it, and it looks a bit like broken because of compiler optimization, but I'm not sure
[14:37] <seb128> fun :/
[14:38] <cyphermox> then, the asserts make even less sense.
[14:38] <seb128> ask him to try building in -O0?
[14:38] <cyphermox> *or* it's just another symptom of the same race, and potentially fixed by the keyboard deactivation patch I sent
[14:38] <seb128> also maybe it would help to upstream if we rebase on 0.9.5/lower our patches, at least we can upstream the issue and see if they have a clue then
[14:39] <seb128> ah, I see you asked already about -O0
[14:39] <cyphermox> the patches apply cleanly
[14:39] <cyphermox> the code isn't that meaningfully different that I can tell to explain the issues
[14:39] <cyphermox> I mean, yeah, a newer plymouth would be good
[14:40] <cyphermox> anyway, it's my plan to update it, just don't have the time just now
[14:42] <cyphermox> seb128: I trade you a plymouth update for testing and finishing casper a11y? :)
[14:42] <seb128> haha
[14:42] <seb128> nice try :)
[14:43] <cyphermox> I'm talking about https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1796275
[14:43] <ubot5> Ubuntu bug 1796275 in ubiquity (Ubuntu Disco) "Screen reader is not auto-enabled on first login if enabled when installing" [High,Triaged]
[14:44] <cyphermox> I have the "revert" in progress, I could get you a casper that's 90% of the way there, but I don't know the gsettings as well as I used to, which wasn't much to begin with ;)
[14:44] <seb128> cyphermox, I'm happy to finish that work for you if you want
[14:46] <seb128> so if you want to hand me that over and update plymouth instead we can do that deal :)
[15:02] <cyphermox> that's what I meant
[15:05] <seb128> cyphermox, k, let's do that
=== pstolowski is now known as pstolowski|afk
[18:30] <willcooke> night all
[18:46] <xnox> seb128, i had a ping from you, somewhere but i cannot find it
[18:46] <xnox> only saw it on my phone.
[18:46] <xnox> where was that?
[18:46] <xnox> ah devel
=== caravena_ is now known as caravena
[20:27] <cyphermox> seb128: https://code.launchpad.net/~cyphermox/ubuntu/+source/casper/+git/casper/+ref/ubuntu/devel