[01:38] <jbicha> duflu: hi, could you go ahead and propose your synaptics changes to gnome-control-center? 3.28.0 only touched translations & the Network panel
[01:39] <duflu> jbicha, gladly, but are you saying we have the synaptics support in mutter landed?
[01:40] <duflu> or patched
[01:41] <jbicha> um, I guess do mutter too :)
[01:41] <duflu> jbicha, mutter upstream rejected it last week
[01:41] <jbicha> mutter 3.28.0-1 is already in bionic-proposed
[01:41] <duflu> so I expected distro patching
[01:42] <jbicha> I mean I'm not making a decision on whether we should patch or not, but I believe you said you didn't want to work on it until 3.28.0 was released
[01:42] <jbicha> so I was letting you know that happened :)
[01:43] <duflu> jbicha, indeeed. I think I just need to move the blocked card in Trello to remind me later
[01:43] <jbicha> ok
[01:44] <duflu> jbicha, done. I don't remind on my own memory. I rely on Trello and similar systems to tell me what to do later :)
[01:44] <duflu> Ugh
[01:44] <duflu> I don't rely on my own memory
[01:47]  * duflu tries more coffee
[01:54]  * jbicha looks suspiciously at gnome-bluetooth
[06:36] <duflu> tjaalton, do we still need to patch mesa since mutter 3.27.92? https://github.com/GNOME/mutter/blob/master/NEWS
[06:47] <tjaalton> duflu: I think so, othe clients are affected too, like clutter
[06:48] <tjaalton> and the patch just flips the default on gallium
[06:49] <duflu> tjaalton, No problem. The login animation corruption issue is probably simple, given enough time to find an analyse the code that copies the framebuffer from gdm into the session
[06:49] <duflu> *and analyse
[06:49] <tjaalton> yeah
[06:49] <tjaalton> I can't repro it on my kbl
[06:49] <duflu> Hmm
[06:49] <duflu> tjaalton, what model screen/laptop?
[06:50] <tjaalton> thinkpad t470s
[06:52] <duflu> tjaalton, that's 8-bit right?
[06:52] <duflu> find /sys/devices -name edid -exec edid-decode '{}' \;
[06:52] <tjaalton> should be, hdr screens are only now being added as an option
[06:53] <duflu> yeah
[06:53] <tjaalton> Supported color formats: RGB 4:4:4
[06:54] <duflu> Look for this or similar: 10 bits per primary color channel
[06:54] <tjaalton> 6 bits per primary color channel
[06:54] <duflu> Oh. That's the 1080p panel then?
[06:54] <tjaalton> yes
[06:55] <duflu> Yeah I heard there was a big difference between t470s's two panel types
[06:55] <duflu> Hmm. Maybe still != 8 bit is the trick
[06:55] <duflu> I didn't test a 6-bit panel
[06:56] <duflu> Not that it should matter too much. You should be getting 8-bit framebuffers, and have done for years
[06:57] <duflu> Maybe we have an uninitialized channel during the framebuffer copy.
[06:57]  * duflu is assuming there is a framebuffer copy, and has never looked
[06:58] <oSoMoN> good morning desktoppers
[07:08] <himadri_> hello again
[07:09] <himadri_> i could fix the issue with purple screen before the splash screen
[07:09] <himadri_> thanks to jamesh :)
[07:09] <himadri_> i have observed one more thing
[07:10] <himadri_> I see a purple screen after entering login password till the desktop launches
[07:10] <himadri_> why is it happening ?
[07:10] <himadri_> the same purple blank screen shows up for a second when i press the lock screen button
[07:13] <duflu> I'd hazard a guess forcing gdm3 to use Xorg (not Wayland) might work around it
[07:13] <duflu> Morning oSoMoN
[07:14] <oSoMoN> good afternoon duflu
[07:15] <himadri_> I'm using lightdm with pantheon greeter and I have elementary desktop installed
[07:17] <duflu> himadri_, if you have a fix that works in gdm3 with unity-greeter and gnome-shell then please comment in bug 1753776
[07:17] <ubot5`> bug 1753776 in xorg-server (Ubuntu) "Graphics corruption (or distortion?) in login animation to Xorg sessions" [Medium,Confirmed] https://launchpad.net/bugs/1753776
[07:24] <didrocks> hey guys
[07:33] <duflu> Hi didrocks
[07:34] <duflu> doesrock,s
[07:34] <himadri_> hey people
[07:34] <duflu> Hello himadri_
[07:34] <himadri_> can anybody help me to fix my issue
[07:34] <himadri_> ?
[07:36] <duflu> himadri_, I don't think that's a bug. The purple screen is meant to be there during transitions
[07:36] <himadri_> can I change with something else of my choice ?
[07:37] <duflu> Because the login screen and your actual session run on different virtual terminals, possibly under different display servers. It's an attempt to make the transitions look seamless
[07:38] <duflu> I *think* that might be the Ubuntu/greeter theme. Not sure if you can change it without modifying code
[07:38] <himadri_> it looks bizarre to have a purple screen in between
[07:38] <duflu> himadri_, yes it's designed to match the default Ubuntu experience, not a customized system.
[07:39] <duflu> himadri_, But maybe didrocks can help you
[07:39] <himadri_> okay
[07:40] <himadri_> didrocks: any comments ?
[07:40] <himadri_> duflu: i have one more question if you don't mind
[07:44] <duflu> himadri_, yes?
[07:45] <himadri_> duflu: I have seen on my friend's laptop that on pressing lock button from top right corner, a screen with logo shows up and then lock screen is displayed. On my system I see the same purple blank screen between the transition. How can I customize it ?
[07:47] <duflu> himadri_, that's the same as your first question. You need to modify code in the gdm3/unity-greeter packages, or replace them with other packages I think. didrocks will know
[07:48] <himadri_> duflu: okay
[07:58] <didrocks> (back in ~15min)
[08:24] <didrocks> hum, no more himadri
[08:28] <icey> Ran into a fun issue this morning when trying to log into a VPN: network-manager-openvpn-gnome can start a VPN connection, but will not show it as up, even when it is, and will make noise about the connection being active if you try to start it again
[08:30] <icey> ah, and it will randomly stop working, and start dumping stack traces into syslog
[08:30] <icey> https://pastebin.ubuntu.com/p/Bj5VFRfCvF/
[08:50] <tjaalton> duflu: libinput 1.10.3
[08:50] <duflu> Sounds good
[08:51] <tjaalton> I'll upload it to debian now
[08:53] <tjaalton> I'm not sure about the ubuntu crashers, so will leave them out from the changelog
[08:57] <duflu> Unfortunately the good stuff is still only in master, I think. So v1.11.0
[08:57] <seb128> koza, duflu, do we have anything to discuss for bluetooth this week?
[08:59] <willcooke> morning
[08:59] <tjaalton> duflu: sure, I don't mind having that in bionic, but maybe if you write the FFE report ;)
[08:59] <tjaalton> once it's out
[08:59] <seb128> hey willcooke
[09:01] <duflu> tjaalton, yeah I can justify that, when a release exists. Although we may only need a patch or two on 1.10
[09:01] <duflu> Although upstream could be convinced to do that for us in a real release
[09:01] <duflu> Morning seb128, willcooke, no nothing blue to talk about
[09:02]  * duflu assumes koza is not in the hangout
[09:02] <Laney> moin
[09:03] <willcooke> duflu, we're in HO, but nothing totalk about
[09:03] <Laney> oh yeah I was going to try a bluetooth keyboard wasn't I
[09:03] <Laney> where did I put that thing
[09:03] <seb128> good morning Laney
[09:04] <didrocks> hey willcooke, Laney!
[09:04] <willcooke> Laney, I got one of those speakers for koza
[09:04] <seb128> duflu, we had a 30 seconds round of "nothing? ok, good, let's wrap then" :)
[09:04] <duflu> willcooke, seb128, yeah sorry I was too slow in plugging in peripherals
[09:04] <seb128> Laney, did koza managed to figure out the issue with bluetooth on your laptop?
[09:04] <seb128> duflu, no worry, there was not much to discuss anyway
[09:05] <duflu> Too slow? You need Bluetooth! :D
[09:06] <duflu> Morning Laney
[09:07] <Laney> hey seb128 didrocks willcooke
[09:07] <Laney> willcooke: nice one
[09:07] <Laney> seb128: nope, there's some wiki page with steps to get logs so I will file a bug
[09:08] <seb128> k
[09:08] <willcooke> oh, Stephen Hawking died. :(
[09:12] <duflu> willcooke, also Rex Tillerson was sacked via Twitter
[09:12] <duflu> Great day
[09:13] <willcooke> Can you do that now?  They're really diversifying their business
[09:13] <duflu> Twitter Diplomatic and Personel Services
[09:14] <duflu> Also Personnel
[09:25] <seb128> duflu, bug #1753776, no idea if that's Ubuntu specific, there are less users testing current GNOME on other distributions and most of upstream people are probably on wayland sessions so wouldn't notice that one
[09:25] <ubot5`> bug 1753776 in xorg-server (Ubuntu) "Graphics corruption (or distortion?) in login animation to Xorg sessions" [Medium,Confirmed] https://launchpad.net/bugs/1753776
[09:26] <duflu> seb128, yes good point. I think it would be easish to debug, but I've never looked at the code before. Also trying to finish the cards I already started instead of starting more
[09:26] <duflu> easy-ish
[09:27] <seb128> duflu, that's bug fixing so that can come later in the cycle
[09:27] <seb128> so if you are busy on other things it's fine to delay that one
[09:27] <duflu> I think everything I'm on is bugfixing now?
[09:28] <seb128> right, though the libinput code changes are border line features
[09:28] <seb128> so better to land that before other changes since that might need more testing
[09:29] <duflu> I think my libinput work is done. Just a matter of waiting for upstream to merge it with other changes that are only in master
[09:29] <seb128> duflu, could you at least open an upstream report about that one with your comment on 8bits/10bits?
[09:29] <seb128> those might be interesting to them
[09:29] <duflu> seb128, it's pure theory. I wouldn't log a bug without more investigation and having looked at the code
[09:29] <duflu> Also disproven by Timo above
[09:30] <seb128> not sure you need to be that strict with bugs filing
[09:30] <seb128> upstream are often interested ot know issues with their code/product
[09:30] <seb128> even if we don't have the time to investigate/debug the code before filing
[09:31] <seb128> anyway, if you don't filing one that's fine
[09:31] <seb128> somebody else can do it
[09:32] <duflu> seb128, yeah I think you're right. There's a chance it's everyone's problem....
[09:32] <duflu> I was talking about my theory of the cause, not the existence of the bug
[09:37] <duflu> seb128, upstream report done
[09:37] <seb128> duflu, thx!
[09:37] <duflu> Aside from anything else I don't have time today to learn the code. May as well see what upstream says overnight
[09:38] <seb128> yeah, I think it's usually a good thing to wait a few days
[09:38] <seb128> sometime upstream just know by reading a report what the issue is/what code change is needed
[09:38] <didrocks> andyrock: hey, have you seen my question about livepatch enable/disable (how to detect it?)
[09:39] <seb128> so no hurry to spend a day to learn the code/come with a solution that has no real benefit for them
[09:42] <seb128> didrocks, what andyrock did on https://code.launchpad.net/~azzar1/software-properties/canonical-livepatch is
[09:42] <seb128> +  def is_livepatch_enabled(self):
[09:42] <seb128> +    file = Gio.File.new_for_path(path='/var/snap/canonical-livepatch/common/machine-token')
[09:42] <seb128> + return file.query_exists(None)
[09:42] <didrocks> jibel: hey! as you were away yesterday as I pinged you late, I'm unsure you have seen the ping on "I tested ubiquity with my latest changes, and all good to me"
[09:43] <didrocks> seb128: excellent, thanks!
[09:43]  * didrocks adds to telemetry
[09:43] <jibel> Hey didrocks, I didn't sorry. I'll havea look
[09:44] <didrocks> jibel: no hurry, not on the critical path anymore :)
[09:44] <didrocks> thanks
[09:45] <seb128> didrocks, I supposed that's going to be replaced by a better api at some point but it should do for now
[09:46] <didrocks> seb128: yeah, won't be hard to change anyway, but it's good enough to me for now :)
[09:46] <seb128> cool
[09:49] <duflu> Wow, bionic has sufficient tooling to build upstream gtk now
[09:49] <duflu> That's handy
[09:49] <duflu> Maybe that's why Jeremy pushed meson?
[09:50] <seb128> no, he did it by mistake
[09:50] <seb128> he was just updating GNOME to 3.28 and though it was part of that when it was not
[09:50] <duflu> Anyway, we can build gtk4 now, apparently
[09:50] <seb128> good
[09:50] <seb128> so at least the ffe was useful at the end
[09:50] <seb128> that update just didn't follow the process and we had to fix that (https://bugs.launchpad.net/ubuntu/+source/meson/+bug/1754578 has the details)
[09:51] <ubot5`> Ubuntu bug 1754578 in meson (Ubuntu) "FFe: New meson version including features uploaded without ffe" [High,Fix released]
[09:54] <seb128> Laney, do you have any "bionic autopkgtest env" using 'machine isolation' set up when you can log into easily to test something for me?
[09:56] <seb128> Laney, if so could you try to run hostnamectl there and see what it says?
[09:57] <seb128> I should really try to get an env configured but I need to change my disk, my 80G ssd just doesn't have enough space for vms & extra lxd stuff
[10:18] <andyrock> good morning
[10:18] <andyrock> hey good morning
[10:18] <andyrock> ops
[10:18] <andyrock> didrocks: yeah what duflu suggested is hacky but okay
[10:19] <didrocks> I think it was seb, but yeah, doing this!
[10:23] <andyrock> didrocks: where do you need this?
[10:25] <didrocks> andyrock: telemetry
[10:25] <andyrock> kk
[10:26] <jbicha> good morning
[10:26] <seb128> hey jbicha, you wake up early!
[10:27] <Laney> seb128: you can get those locally with the cloud VMs... so I can just run the test, which one is it?
[10:27] <jbicha> yeah, as long as I don't stay up too late at night, it works ok :)
[10:27] <Laney> seb128: you can probably also use canonistack for this with the SSH runner, although I didn't try that myself
[10:27] <seb128> Laney, udisks2 from bionic-proposed
[10:28] <seb128> the error looks like hostnamectl erroring out
[10:28] <jbicha> seb128: could you take a look at component-mismatches? presumably we are good to drop some of our apps to universe that have switched to snaps
[10:28] <seb128> jbicha, adding to my list but that doesn't seem like of any use to do now, that's usually tided up at the end of the cycle which is good enough
[10:28] <Laney> those should be seeded into supported imho, we still need them for upgraders
[10:29] <seb128> +1
[10:29] <jbicha> oh ok, I needed to know whether to add the dh_translations hack to gnome-characters for instance (if it was in universe, it's just extra clutter for Debian)
[10:30] <Laney> ah man, some lxd upgrade has lost my dev environment :((((((((((((((((((((((((((
[10:30] <Laney> laney@raleigh> lxc start bionic                                                                                 ~
[10:30] <Laney> error: not found
[10:31] <seb128> :/
[10:33] <seb128> jbicha, some people are going to still use the debs so it's still useful to be able to fix translations issue through langpacks
[10:33] <seb128> jbicha, gegl added a patch that fixes the armhf issue to the bug
[10:34] <jbicha> seb128: are you suggesting we add X-Ubuntu-Use-Langpack to a bunch of GNOME universe stuff?
[10:34] <seb128> yes
[10:34] <seb128> we traditionally did that
[10:34] <jbicha> I wish you had thought of that before 3.28.0 ;)
[10:34] <seb128> well, if you had asked earlier... :)
[10:35] <seb128> anyway that's a nice to have and I'm not even sure it's needed/better
[10:35] <seb128> so whenever we change those is fine
[10:36] <jbicha> hmm
[10:37] <seb128> it's probably fine to have those out of launchpad translations, I guess most users are going to use the snaps anyway since those are default/uptodate/etc
[10:47] <jbicha> I added a Trello card for it
[10:47] <seb128> thx
[10:49] <seb128> though I'm unsure how much it fits there, it's sort of a ~canonical-wont-do
[10:49] <jbicha> I added it so I can get a clear answer from y'all :)
[10:49] <seb128> I don't think it's especially up to us
[10:50] <jbicha> if you're delegating the decision to me, I'm not particularly interested in doing it at this time
[10:50] <seb128> though as L_aney said if we still have those supported for upgraders then it makes sense to keep them in main/langpacks
[10:55] <Laney> seb128: hostnamectl doesn't have that field in its output
[10:55] <Laney> that's why you'll get a failure
[10:55] <seb128> ah
[10:55] <seb128> thx
[10:56] <jbicha> what's even calling hostnamectl? I didn't see it by grepping udisks source
[10:56] <Laney> grep harder
[10:56] <Laney> laney@raleigh> rgrep hostnamectl                                                             ~/temp/udisks2-2.7.6
[10:57] <Laney> src/tests/integration-test:    out = subprocess.check_output('hostnamectl status | grep "CPE OS Name"', shell=True).decode().strip()
[10:57] <seb128> what L_aney said
[10:59] <Laney> dunno why they didn't use /etc/os-release for this
[10:59] <oSoMoN> popey, flexiondotorg : can you please comment on https://forum.snapcraft.io/t/request-auto-connection-of-removable-media-for-libreoffice/2660 ?
[11:00] <seb128> Laney, jbicha, that was added recently in https://github.com/storaged-project/udisks/commit/dd1c5c85
[11:00] <jbicha> helps if I grep udisks2 instead of udisks 🤦
[11:00] <popey> oSoMoN: done
[11:01] <Laney> seb128: ya, looks like that get_distro_info thing needs fixing
[11:01] <seb128> Laney, jbicha, I'm going to have a look at sending an upstream pr for it
[11:01] <Laney> nice
[11:01] <seb128> Laney, do you have any idea offhand why our hostnamectl isn't including that info?
[11:01] <seb128> if not I can have a look to that as well
[11:01] <Laney> never heard of it
[11:02] <oSoMoN> popey, thanks. I'm not sure what the process is now that the request has a sufficient number of +1s ? does Jamie need to push a button for this to be effective?
[11:02] <popey> yes
[11:02] <popey> he's on vacation, so will be done when he's back
[11:02] <oSoMoN> ack
[11:02] <popey> unless it's urgent and emily can help
[11:02] <oSoMoN> not that urgent
[11:02] <oSoMoN> but bus factor :/
[11:03] <seb128> Laney, https://www.freedesktop.org/software/systemd/man/os-release.html says it's optional
[11:03] <oSoMoN> popey, mind approving my joining https://launchpad.net/~ubuntu-wiki-editors ?
[11:04] <Laney> seb128: think you can probably use ID and VERSION_ID for this
[11:04] <seb128> right
[11:16] <didrocks> speaking of VERSION_ID, it doesn't change once we have .1, .2, .3, correct?
[11:19] <Laney> no, only the full name
[11:19] <Laney> VERSION and PRETTY_NAME
[11:19] <didrocks> ok, that was my bet, but better to confirm, thanks Laney!
[11:23] <oSoMoN> davidcalle, would you mind approving my request to join ~ubuntu-wiki-editors ?
[11:24] <davidcalle> oSoMoN: done
[11:24] <oSoMoN> cheers!
[11:25] <oSoMoN> jibel, https://wiki.ubuntu.com/Process/Merges/TestPlans/libreoffice , I would say that's a first cut, feel free to add tests or reword to your liking
[13:00] <kenvandine> jbicha, freetype solved the emoji issue
[13:00] <kenvandine> jbicha, fixed in the platform snap from the edge channel
[13:00] <jbicha> kenvandine: great! bigon mentioned freetype in his reponse at https://www.reddit.com/r/debian/comments/83lyf2/color_emoji_on_debian_9/
[13:01] <kenvandine> jbicha, ah, cool
[13:01] <kenvandine> i backported all 3 of those packages :)
[13:03] <seb128> we wouldn't have those issues if we were building on bionic!
[13:03] <jbicha> yup, I confirm it works now :)
[13:03] <kenvandine> seb128, i know...
[13:03] <kenvandine> jbicha, cool, thx
[13:03] <kenvandine> i need to do a round of testing of the other snaps against it on 16.04 and bionic before i promote it
[13:03] <jbicha> do you want me to close the LP bug now?
[13:04] <kenvandine> jbicha, sure
[13:04] <kenvandine> well, no
[13:04] <kenvandine> wait :)
[13:04] <kenvandine> jbicha, what was the bug number?
[13:04] <seb128> kenvandine, I'm going to open a trello card about "needs to build our default snaps on bionic" if it's ok for you
[13:04] <jbicha> LP: #1753146
[13:04] <ubot5`> Launchpad bug 1753146 in gnome-characters (Ubuntu Bionic) "Characters Snap doesn't have color emoji" [High,Fix released] https://launchpad.net/bugs/1753146
[13:04] <kenvandine> seb128, that's fine
[13:04] <seb128> thx
[13:05] <kenvandine> seb128, i'll keep nagging folks until i can do that
[13:05] <seb128> :)
[13:06] <jbicha> seb128: https://trello.com/c/1R0lryqL/183-gnome-328-platform-snap
[13:07] <seb128> jbicha, I know about this one, that's about the platform though, not about rebuilding apps
[13:07] <seb128> we could build on bionic using the xenial based platform today
[13:14] <jbicha> oh, I don't know much about how snaps work yet
[13:19] <kenvandine> yeah, we really want to build on bionic with a core18 snap, so all bionic
[13:26] <seb128> jbicha, snap don't have (yet) platform-dev snaps, so GNOME snaps are basically built on launchpad on a stock xenial chroot + a backport ppa
[13:26] <seb128> the backport ppa is what is used to build the platform
[13:27] <kenvandine> i hope to get that build snap work done soonish
[13:27] <seb128> and those are currently stacked on 16.04 core
[13:27] <kenvandine> so in the future, post 18.04, we don't need backport ppas
[13:28] <seb128> so basically when you start a GNOME snap on bionic the env is a 16.04 core + GNOME from bionic backported to xenial through a ppa + app itself built on xenial
[13:28] <seb128> we want to change that to be core 18 + GNOME built from bionic + app built on bionic
[13:28] <seb128> having a core 18 we can use is snapd team work
[13:29] <seb128> the second item is the existing card
[13:29] <seb128> the third one is the one I'm going to create
[13:38] <kenvandine> seb128, good news, sergiusens has a PR adding the bases support :)
[13:38] <seb128> kenvandine, I saw :)
[13:38] <seb128> \o/
[13:42] <popey> oSoMoN: just fyi. As you're in ~canonical, you didn't need to join ~ubuntu-wiki-editors. Both have access to edit the wiki
[13:45] <oSoMoN> popey, ack. it's weird that I wasn't allowed to edit that page initially, though
[13:46] <popey> The wiki is weird. 😃
[13:46] <oSoMoN> oh well, now I'm doubly allowed :)
[13:59] <jbicha> seb128: good news, I discovered that Deja Dup uses ulimit which seems to be why it's crashing
[14:00] <jbicha> https://salsa.debian.org/gnome-team/deja-dup/blob/debian/master/data/org.gnome.DejaDup.Monitor.desktop.in
[14:00] <jbicha> the autostart comment points to LP: #1302416 :(
[14:00] <ubot5`> Launchpad bug 1302416 in deja-dup (Ubuntu) "deja-dup monitor was taking 6GB of memory" [High,Confirmed] https://launchpad.net/bugs/1302416
[14:01] <seb128> jbicha, good news of some sort, it means webkit is eating more than this limit?
[14:01] <seb128> so still a bug on their side
[14:02] <jbicha> no, that's a security feature in webkitgtk
[14:02] <seb128> "feature"
[14:02] <jbicha> complicated aslr stuff. I didn't really understand it
[14:02] <seb128> well they have a bug in any case
[14:02] <seb128> it should at least exit with a clear error
[14:02] <seb128> not segfault
[14:02] <jbicha> I didn't get a degree in Computer Science ;)
[14:02] <seb128> me neither :p
[14:02] <jbicha> mcatanzaro says Deja Dup should stop using ulimit
[14:03] <jbicha> interestingly, Deja Dup doesn't directly link against webkit but it's pulled in because of GNOME Online Accounts
[14:04] <seb128> well, still looks like they are trying to not fix their bug
[14:05] <seb128> and they should comment on the bugzilla to say so
[14:05] <jbicha> comment is coming. You're welcome to reply there
[14:05] <jbicha> I talk to mcatanzaro on #epiphany on irc.gnome.org
[14:10] <doko> desktop-common seeds libu2f-host, which doesn't have a MIR
[14:12] <doko> seb128, willcooke: ^^^
[14:12] <doko> oSoMoN: LO has two new font recommendations: http://people.canonical.com/~ubuntu-archive/component-mismatches-proposed.svg
[14:13] <seb128> doko, what is libu2f-host?
[14:13] <doko> ENOCLUE. see http://people.canonical.com/~ubuntu-archive/component-mismatches-proposed.svg
[14:17] <seb128> doko, http://bazaar.launchpad.net/~ubuntu-core-dev/ubuntu-seeds/platform.bionic/revision/2162
[14:17] <seb128> doko, talk to xnox
[14:17] <doko> xnox: ^^^
[14:37] <oSoMoN> doko, if they are recommends, do they need to be in main too?
[14:44] <oSoMoN> I guess if it's on that graph then the answer is yes…
[14:45] <oSoMoN> I'll file MIRs
[14:49] <xnox> doko, it is udev rules, to make Yubikeys work and other U2F tockens.
[14:50] <xnox> for e.g. ssh / gpg / OTP on the desktop, with these dongles. These udev rules used to be shipped in src:systemd, but now they are shipped by that source package.
[15:29] <jbicha> kenvandine: how often is 'snap refresh' run in the background?
[15:29] <kenvandine> 4 hours
[15:29] <jbicha> wow ok
[15:37] <didrocks> hum
[15:37] <didrocks> not 6 hours?
[15:38]  * didrocks looked this morning for other reasons on the timer unit
[15:41] <seb128> I think it's more complex than that and they are discussing tweaks
[15:42] <seb128> like having a "minimal delay" and then do a random scheduling in the n-following-hours
[15:42] <seb128> like not before 2 hours and then a random schedule in the next 4 hours
[15:42] <seb128> the initial delay would be to not autorefresh temp instances and generate more traffic than needed
[15:44] <didrocks> yeah, I was only talking about the basic refresh rate in /lib/systemd/system/snapd.refresh.timer
[15:44] <seb128> that timer is going away in .32 afaik
[15:44] <seb128> they have an internal timer now
[15:45] <seb128> anyway
[15:45] <seb128> jbicha, what are you trying to figure out exactly?
[15:46] <jbicha> I was just curious how quick users would get updates if they did nothing
[15:52] <jbicha> I pinged mterry about the deja-dup issue, let's see how he wants it handled
[15:58] <seb128> thx
[16:49] <Laney> Trevinho: https://paste.ubuntu.com/p/G6gfkF2fNw/ got the nautilus thing again
[16:50] <Laney> I'm guessing it's doing something for everything in my maildir for some reason
[16:50] <Laney> laney@nightingale:~/Mail$ find -type f | wc -l
[16:50] <Laney> 372915
[16:51] <Trevinho> Laney: mh ok thanks
[16:51] <Trevinho> Laney: it's just parsing the results from locate
[16:52] <Laney> if you can, you probably want to cancel when the overview isn't visible any more?
[16:52] <Laney> also maybe limit the number you process given that only 3(?) are going to be shown anyway? :-)
[16:52] <Trevinho> Laney: it should cancel
[16:52] <Trevinho> Laney: well, that'sa  thing to be done at different lelve
[16:53] <Trevinho> Laney: since the same search is used also for normal nautilus searches
[16:53] <Laney> ok, well the cancel didn't work then!
[16:53] <Trevinho> Laney: so it would be wrong to limit it search engine level
[16:53] <Trevinho> but we can do it in the shell search provider
[16:53] <Trevinho> it's also true that it might process eleemtns with better priority
[16:53] <Trevinho> so... like you find a file you accessed more recently or matching better
[16:54] <Trevinho> so it could be shown first
[16:54] <Trevinho> unfortunately is a bit hard to predict what you want to see
[16:54] <seb128> Trevinho, oh, seeing you talk here reminded me ... did you see duflu's review comments on your theme changes?
[16:54] <Trevinho> seb128: I saw them, but not went thorugh
[16:54] <seb128> k, just a reminder you might want to do that in the next days then :)
[16:54] <Trevinho> https://www.irccloud.com/pastebin/2EUj3wSz/
[16:54] <Trevinho> seb128: ^
[16:55] <seb128> oh, and sorry that daniel is being picky ;p
[16:55] <Trevinho> good to be, I also am. It's even true I won't be bale to fix everything probably
[16:57] <Laney> (gdb) print error
[16:57] <Laney> $1 = (GError_autoptr) 0x0
[16:57] <Laney> this thing can't be cancelled properly
[17:06] <Trevinho> Laney: mh that's pretty weird though
[17:07] <Trevinho> Laney: so it's crashing when something strdup'ing...
[17:07] <Trevinho> but, that memory area shouldn't have been free'd yet, isn't it?
[17:08] <Laney> it's not crashing
[17:08] <Laney> I just connected while it was spinning the cpu to see what was happening
[17:08] <Laney> and it's going wild on my home directory
[17:13] <Laney> can't see _stop being called when you close the overview
[17:17] <Trevinho> mh, stop might not being called if start has not being advertized
[17:17] <Trevinho> I've noticed that in other cases
[17:17] <Trevinho> but, iirc running is always set when starting
[17:17] <Trevinho> Laney: ah, in the mean time, can you please check https://bileto.ubuntu.com/#/ticket/3196 ?
[17:17] <Trevinho> (check = publish)
[17:19] <Trevinho> Laney: so you connected with nautilus and its's just a random place, right? isn't hanging there...
[17:21] <Laney> Trevinho: it's not hanging anywhere, it is looping over my whole home directory processing each file
[17:22] <Laney> it'll probably finish at some point
[17:22] <Laney> after using 50% of my battery
[17:22] <Laney> published
[17:22] <jbicha> seb128: https://bazaar.launchpad.net/~ubuntu-core-dev/ubuntu-seeds/ubuntu.bionic/revision/2648
[17:22] <jbicha> so gnome-calculator is in component-mismatches for re-promotion to main now
[17:28] <Trevinho> Laney: mh, the normal shell doesn't check files that aren't in the real home, but that makes the tool not usfeul
[17:29] <Trevinho> but in general, if you close the overview the search should stop
[17:29] <Trevinho> like when you search in nautilus
[17:29] <Trevinho> as soon as you type something and then hit escape, everything is cancelled.
[17:29] <Laney> I did a G_MESSAGES_DEBUG=all run and you can't see _stop being called
[17:30] <Laney> there's https://paste.ubuntu.com/p/92vDkDjGmN/ and you see it spamming for each file after that
[17:31] <Laney> before that there are some _stop calls but I think they are from previous searches, like if you type more text it aborts the old one
[17:47] <Trevinho> Laney: mh, ok... this might be more related to the activity search though
[17:47] <Trevinho> I'll check that
[17:48] <Trevinho> as I've mostly debugged it using normal search, as the shell is reusing almost the same thing
[17:48] <Laney> it is the activity search
[17:48] <Trevinho> anyway... Laney give that a try too and see if  you can reproduce (i.e. just ctrl+f on the nautilus home folder)
[17:49] <Trevinho> yeah, yeah... just see if you get the same when the non activity search is used
[17:49] <Trevinho> so we can isolate better
[17:50] <Laney> ok, so yes it happens there too
[17:50] <Laney> I press ctrl-f, type "ti" and now I have a force quit dialog on nautilus
[17:50] <Laney> it's spamming this for each file https://paste.ubuntu.com/p/wcCDRbm5gh/
[17:50] <Trevinho> mh that's weird though since also my updatedb db is quite big but it doesn't show such issues
[17:51] <Laney> probably because "ti" is a substring of "nightingale" which is the name of my laptop and is in the filename of all emails
[17:51] <Trevinho> yeah, very likely
[17:52] <Laney> so the problem is when there are tons of matches
[17:53] <Trevinho> right, but that shouldn't block nautilus.. mhmh
[17:53] <Trevinho> Well, or maybe it does since this processing is happening in the main thread at this point...
[17:54] <Trevinho> in the debug you don't get anything related to the locate search engine though
[17:54] <Trevinho> mh
[17:58] <Laney> in the ctrl-f case it is stopped but then it blocks when trying to add the results to the view
[18:03] <Laney> Trevinho: https://paste.ubuntu.com/p/RCdjkr9wmw/
[18:03] <Laney> thats from ctrl-f
[18:04] <Laney> got to go, hope that is enough information
[18:08] <flocculant> hi peeps - over in Xubuntu land we've been scratching our heads over a 75s timeout - which possibly is something to do with pulseaudio/bluetooth - digging a bit further I can see the same issue booting the Ubuntu live session (except it only times out once for 25 secs) bug 1754836
[18:08] <ubot5`> bug 1754836 in pulseaudio (Ubuntu) "Xubuntu: ERROR:dbus.proxies:Introspect error on org.bluez:/org/bluez: dbus.exceptions.DBusException: org.freedesktop.DBus.Error.TimedOut: Failed to activate service 'org.bluez': timed out (service_start_timeout=25000ms)" [Undecided,Confirmed] https://launchpad.net/bugs/1754836
[18:08] <flocculant> bluesabre: ^^ as discussed in x-dev in ping to me ;)
[18:11] <willcooke> jibel, could be relevent to our interests? ^ :)
[18:12] <willcooke> flocculant, jibel is probably EOD now, but he's been looking at some slowness in booting the live session.  May or may not be related, but I'll speak to him in the morning and see if he's seen anything similar
[18:14] <willcooke> and with that, I need to go an paint a door
[18:14] <willcooke> night all
[18:41] <oSoMoN> good night all
[20:32] <kenvandine> seb128, still around?
[20:46] <robert_ancell> kenvandine, I guess not!