[00:09] <Eickmeyer[m]> If it would help, should I go back and mark the sddm SRU as verification-needed?
[00:10] <Eickmeyer[m]> That could at least delay some things.
[00:10] <arraybolt3[m]> I don't know. Personally (and this is just my opinion), I would mark it as staging so that it's clear that the verification worked but it's not supposed to migrate yet. But I don't know if that's really the best idea, that's just my idea, and so far I've had a lot of bad ideas today, so :P
[00:10] <Eickmeyer[m]> That's not valid in an SRU tag.
[00:11] <arraybolt3[m]> ```
[00:11] <arraybolt3[m]> To stage an upload, follow the usual process but additionally add a block-proposed-<series> tag to at least one of the SRU bugs together with a comment explaining the reason for the staging.
[00:11] <arraybolt3[m]> ```
[00:11] <arraybolt3[m]> What am I missing?
[00:12] <Eickmeyer[m]> The upload is far beyond the staging stage at this point. It's completely in proposed.
[00:12] <arraybolt3[m]> Oh.
[00:12] <arraybolt3[m]> Hey I'm learning all sorts of new things today! :D
[00:13] <Eickmeyer[m]> Happens. SRUs are tricky logistically.
[00:15] <Eickmeyer[m]> Ok, I stand corrected: "...his can be done at any time before an SRU is released. If you do so, please make sure that you add a bug comment that explains the reason."
[00:15]  * Eickmeyer[m] does that
[00:18] <Eickmeyer[m]> https://bugs.launchpad.net/ubuntu/+source/sddm/+bug/1998254/comments/9
[00:18] -ubot93:#lubuntu-devel- Launchpad bug 1998254 in sddm (Ubuntu Kinetic) "[SRU] sddm with multiple monitors can result in screen overlays and insets" [High, Fix Committed]
[00:20] <arraybolt3[m]> Thank you for your patience. I can't believe something as silly as one missing feature could cause this big of a problem.
[00:21] <teward> behold
[00:21] <teward> tsimonq2: now get off my ass about this one :P
[00:22]  * arraybolt3[m] starts an emergency
[00:22] <Eickmeyer[m]> The Grumpasaurus Rex has arrived.
[00:22] <Eickmeyer[m]> Oh, nvm. It's just teward .
[00:22] <teward> arraybolt3[m]: you have incoming missiles :P
[00:22] <arraybolt3[m]> It's lubot [matrix] teward, that's different. He's much more dangerous than teward.
[00:23] <Eickmeyer[m]> Oh, so Grumpasaurus Rex with Missile Launchers. Gothcha.
[00:23]  * Eickmeyer[m] programs his Ubuntu Studio setup with the brown note
[00:23]  * teward assigns shit for arraybolt3 to do
[00:24] <Eickmeyer[m]> "assigns shit" : "brown note" Yep, that checks.
[00:25] <teward> *shoves Eickmeyer[m] out the airlock*
[00:25] <teward> i've had an evil day
[00:25] <Eickmeyer[m]> Gladly. o/
[00:25] <teward> at least every message I post here pings me in Matrix xD
[00:25] <teward> Eickmeyer[m]: you try waking up without water in the apartment
[00:25]  * arraybolt3[m] sucks Eickmeyer back into the airlock and accidentally triggers the Big Bang
[00:25] <teward> and then having to ask your PARENTS to let you use their shower instead
[00:25] <teward> and steal breakfast
[00:25] <arraybolt3[m]> OK hold on things aren't quite working, uno momento por favor
[00:25] <teward> ... okay not steal but still
[00:26] <Eickmeyer[m]> Oh, yeah, that sucks.
[00:26] <teward> *pokes arraybolt3[m] with a stikc*
[00:26] <arraybolt3[m]> Oh nice, Matrix.org is trying to hold me hostage!
[00:26] <arraybolt3[m]> Bah, I don't want my Encrypted Messages anyway.
[00:26]  * arraybolt3[m] signs out and loses data because why not
[00:27] <teward> arraybolt3[m]: you'll lose your current connections so :)
[00:27] <teward> use a private window to use a separate web session
[00:27] <Eickmeyer[m]> The bridge, at it again.
[00:28] <teward> oooops i broke it :)
[00:28] <arraybolt3> ?!
[00:28] <teward> welcome back bot!
[00:28] <teward> soi soi soi soi soi
[00:28] <arraybolt3> And it's giving me a hard time setting the homeserver...
[00:28] <teward> arraybolt3[m]: that might be your client specifically
[00:28] <arraybolt3> I'm using Element Nightly.
[00:29]  * Eickmeyer[m] uses Matrix and sees a matrix relay bot and is trying to figure out what the hell....
[00:29] <teward> Eickmeyer[m]: because IRC be bitchy and we already have an irc bridge bot and to sate Simon's hatred we're simply using the existing bridge we have for our stuff?
[00:29] <arraybolt3> Eickmeyer[m]: Dedicated Lubuntu Development matrix room, bridged to IRC, then showing up on Matrix. Welcome to the Wormhole.
[00:30] <arraybolt3> Sooo... I guess Element Nightly must not be compatible with the new homeserver. /me nukes Element Nightly and installs Element Stable
[00:30] <arraybolt3> But then again app.element.io didn't do it either so I'm wondering if our own personal GFW is blocking things again.
[00:31] <teward> arraybolt3[m]: here's your test.  https://matrix.lubuntu.me/ in a web browser
[00:32] <teward> if you can reach it good if not then we have to debug your stuff
[00:32] <teward> the federation tester shows it works though so
[00:32] <arraybolt3> I, in Missouri, cannot access it.
[00:33] <arraybolt3> teward: https://ibb.co/zR3sM9d
[00:35] <arraybolt3> teward: Odd thing, I can get a web page from matrix.lubuntu.me.
[00:35] <arraybolt3> The web page?
[00:35] <arraybolt3> That is a web page.
[00:35] <arraybolt3> (maybe Lubuntu Lunar is messing with me)
[00:35] <arraybolt3> Yes, it says Synapse is running.
 everything shows here too :)
[00:37]  * arraybolt3 tries using the envs.net Element
[00:37] <arraybolt3> Nope, same error.
[00:38] <arraybolt3> eickmeyer: What did you do exactly?
[00:38] <arraybolt3> You seem to have been able to shake it just right.
[00:38]  * arraybolt3 tries using Firefox rather than Chrome
[00:39] <arraybolt3> Fail.
 arraybolt3: eickmeyer didn't register on the lubuntu matrix server
 you can keep your previous matrix server just join #lubuntu:lubuntu.me
 you don't *have* to use the Lubuntu homeserver
 and worst case I just spawn you with a default password and give you that to login with
 as for why it can't get the server configs I have *no idea*
 fresh Element here can
[00:40] <arraybolt3> Ah, OK.
[00:40] <teward> it might be failing to get homeserver config for some obscure A F reason hence why i wanted you to try the web app, hit F12 and watch the dev console for output
[00:40] <teward> because that gives you useful information
[00:40] <teward> indeed
[00:41] <arraybolt3> Oh, OK, I'll do that.
[00:41] <teward> it's how i am working - my own homeserver, connected to Lubuntu's space
[00:41] <arraybolt3> I missed you asking me to use F12.
[00:41] <teward> *evil grin*
[00:41] <teward> ffs nodejs takes an eon to install >.>
[00:42] <Eickmeyer[m]> That's because it's YUUUUUUGE.
[00:43]  * arraybolt3[m] uploaded an image: (1830KiB) < https://libera.ems.host/_matrix/media/v3/download/matrix.org/bMKdJMXuGqWAIeglkgmjINwk/image.png >
[00:43]  * arraybolt3[m] uploaded an image: (1833KiB) < https://libera.ems.host/_matrix/media/v3/download/matrix.org/LblhltiFKkyhaOhtSKVreKsv/image.png >
[00:43] <Eickmeyer[m]> arraybolt3: /join #lubuntu-devel:lubuntu.me 
[00:43] <Eickmeyer[m]> Or click that, whatever...
[00:44] <arraybolt3> teward: There's more than just that, but that's a lot of it.
[00:44] <teward> oh i freaking HATE cors
[00:44] <arraybolt3> Eickmeyer[m]: Matrix.org is *so buggy* for me, so I'd really like to make this work.
[00:44] <Eickmeyer[m]> It's not Matrix.org that's buggy for you. It's the libera bridge that's the buggy part.
[00:44] <Eickmeyer[m]> And that has nothing to do with matrix.org.
[00:45] <teward> so one thing i know by the way arraybolt3 
[00:45] <teward> is Chrome uses web isolation
[00:45] <teward> and that breaks CORS
[00:45] <teward> that might affect the element client too
[00:46] <arraybolt3> Eickmeyer[m]: It's not just the Libera bridge. My whole E2EE setup is in some sort of weird limbo state, and I'm locked out of a space I made.
[00:46] <arraybolt3> teward: It also failed with the same error in Firefox, FWIW.
[00:46]  * arraybolt3 tries to get a debug log from there too
[00:46] <teward> firefox does it too
[00:46] <teward> i'll; do some poking
[00:46] <arraybolt3> Ah. Well then I'll get some other Matrix client then.
[00:46] <Eickmeyer[m]> Spaces aren't a Matrix thing, that's an Element thing.
[00:51]  * arraybolt3 installs Nheko
[01:00] <arraybolt3> Yay! Nheko works!
 hi arraybolt3
 so
 firefox and chrome have web security enabled by default
 so for Firefox you have to set this to false in about:config - security.fileuri.strict_origin_policy
 Chrome is harder, you have to disable it with a command line arg/flag
 *does a thing*
 i... think i fixed it...
 i need to test from another computer that doesn't have element on it
 *steals his desktop*
 but first, sushi dinner
 yep
 i had to dig into the configurations for matrix-synapse
 and get some things configured but I *think* i got that all done
 gonna learn from this and fix my own server i think too xD
 there we go
 that should behave now :)
[01:37]  * lubot [matrix] <arraybolt3> now tries Element Nightly
 hah its the same bridge we use for telegram to IRC so xD
 again most of our stuff is on the dedi so :)
[05:48] <Joo-Hwan> HI
[05:49] <Joo-Hwan> Is it possible to consult?
[05:50] <guiverc> Joo-Hwan, if you have a Lubuntu development issue/suggestion, providing details may help us to understand & know how to respond.  Does it relate to lunar?
[05:50] <Joo-Hwan> Wait
[05:51] <Joo-Hwan> What does the detail mean?
[05:51] <guiverc> This is the Lubuntu development channel, current development is on lunar (what will be 23.04 on release next April)
[05:52] <guiverc> channel topic has links on channel, read up your screen
[05:53] <Joo-Hwan> korea / Can't you make an inquiry about Luburn?
[05:53] <arraybolt3[m]> Joo-Hwan: We may be encountering a language barrier. guiverc, is there a Korean lubuntu support channel?
[05:53] <Joo-Hwan> I'm sorry. I didn't know
[05:53] <guiverc> I see now you were on #lubuntu, that is a better place to be for support  (sorry I missed that before)
[05:54] <Joo-Hwan> ok Thank
[12:30] -ubot93:#lubuntu-devel- Issue 5 in lubuntu-team/lubuntu-update-notifier "lubuntu-updater stucks on SDDM update prompt" [Open]
 "[matrix] <kc2bez> FYI https://..." <- If I'm not mistaken, ubuntu-update-notifier doesn't have the built-in option to ask for various options, overwrite config files... In that case it never ends and gets stuck
[13:24] <Roberalz[m]> * If I'm not mistaken, lubuntu-update-notifier doesn't have the built-in option to ask for various options, overwrite config files... In that case it never ends and gets stuck
 It happened to me once in a similar situation. (re @lubuntu_bot: (matrix) <kc2bez> Unti yesterday there was no bug on that. That was most of yesterday's topic of discussion.)
 I'll reread all the posts then, thanks (re @lubuntu_bot: (matrix) <kc2bez> Unti yesterday there was no bug on that. That was most of yesterday's topic of discussion.)
[14:28]  * lubot [matrix] <kc2bez> sends a tanker of Tim Horton's coffee.
 morning dayjob crap takes priority :P
[18:56] <tsimonq2> Good afternoon!
[19:08] <tsimonq2> EOUTOFTIME
[19:08] <arraybolt3[m]> Well then I'll come over here.
[19:09] <arraybolt3[m]> I have a complicated plan of attack for the lubuntu-update-notifier bug that you can probably find in the backlog.
[19:23] <tsimonq2> Here's a plan forward...... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/118cd7657b4950eaf74671ed29dfae2fd12c0abc>)
[19:26] <arraybolt3[m]> OK. For me it seemed like notes.lubuntu.me was working sorta well enough, but it was a very thrown-together method of doing things. And I do not know what software would be best to use for this task, and am too pressed for time to be investigating it - unless of course you'd like to take over making the recovery script (if that's even a thing we're doing), and we're willing to delay Debian LXQt a couple more days. Then I could
[19:26] <arraybolt3[m]> spend my time only searching for a task-tracking system so we can unravel this.
[19:28] <arraybolt3[m]> Preferably we'd be done with LXQt before the end of this month to make it in time for the Bookworm freeze, we need to finish our own LXQt 1.2.0 audit before the end of the cycle, and we have to unbreak lubuntu-update-notifier pronto, everything else can wait I believe.
[19:28] <arraybolt3[m]> Simon Quigley: ^
[19:28] <arraybolt3[m]> Would you like me to spend my time searching for a task-tracking system for us?
[19:29] <tsimonq2> I would personally say carry on for now, let's ping chaosadmin teward to see if he has ideas, since he'll be the one setting it up.
[19:29] <tsimonq2> https://discourse.ubuntu.com/t/lunar-lobster-release-schedule/27284
[19:30] <tsimonq2> We have just about 2 months for all of this
[19:30] <teward> nuke the world, nuke the world, 1 2 3 4. together we will nuke the world, 1 2 3 4
[19:30] <tsimonq2> That being said, there's also char holidays[10];, so 2 months will fly.
[19:30] <teward> tsimonq2: arraybolt3[m]: short term, would a kanban board / issue tracker work?
[19:31] <tsimonq2> teward: yes
[19:31] <teward> do you have any objection to me letting that live on Trello for now?
[19:31] <teward> with me as godadmin
[19:31] <teward> until i can set up Taiga on our stuff
[19:31] <teward> (which is basically open source trello)
[19:32] <tsimonq2> I hate the idea. I really do. But we need something, and it's under the hope of "godadmin is in the process of actually setting up Taiga"
[19:32] <tsimonq2> Also
[19:32] <tsimonq2> How easily can you migrate tasks from Trello to Taiga?
[19:32] <teward> takes me about 5 minutes.
[19:33] <arraybolt3[m]> What is Trello?
[19:33] <teward> https://trello.com/en
[19:33] <teward> basically SaaS kanban boards and issue tracking
[19:33] <tsimonq2> *RMS voice* "it's non-free software"
[19:33] <teward> truth
[19:34] <tsimonq2> I've used it before
[19:34] <arraybolt3[m]> Ugh. Non-free and non-cost-free. Oh well.
[19:34] <tsimonq2> heavily
[19:34] <teward> arraybolt3[m]: it's actually got a free version
[19:34] <teward> it's just not OSS
[19:34] <teward> i'm working on deploying Taiga now though
[19:34] <teward> *shoves kc2bez out the airlock until the coffee debt is paid*
[19:34] <arraybolt3[m]> Oh.
[19:35]  * genii twitches
[19:35]  * tsimonq2 pays kc2bez's coffee debt
[19:35] <tsimonq2> carry on :P
[19:35] <arraybolt3[m]> Well we use GitHub so I won't gripe too much.
[19:35] <tsimonq2> GitHub has always been a mirror tbh :P
[19:35]  * arraybolt3[m] returns to writing a 90% comment 10% code script that hopefully won't give the SRU team a heart attack
[19:36] <tsimonq2> We've only used it recently because Phab is obsolete
[19:38] <arraybolt3[m]> Sap. xdotool is not installed by default. I need that for my script.
[19:39] <arraybolt3[m]> Maybe we can add it as a depends?
[19:40] <tsimonq2> So, recovery script, or an actual modification to lubuntu-update-notifier allowing for continuation of interrupted apt processes?
[19:40] <tsimonq2> Except, what you said before is true, apt isn't killed...
[19:40] <tsimonq2> I'm confused as to what your approach will be
[19:40] <arraybolt3[m]> Recovery script.
[19:41] <arraybolt3[m]> Basically, it's a bunch of long, convoluted logic to detect if lubuntu-upgrader and apt need to be restarted at all. If it does, it can kill python3 /usr/sbin/aptd (I think?), kill lubuntu-upgrader,and restart it with a special script.
[19:42] <arraybolt3[m]> Same approach as last time, except for much better logic so that, if the user isn't updating using lubuntu-upgrader, the script just silently exits.
[19:42] <arraybolt3[m]> Simon Quigley: ^
[19:43] <arraybolt3[m]> The reason I wanted xdotool is because that would let me detect where the lubuntu-upgrader window is at (I think), so that if the user is doing something like SSH with X forwarding and running lubuntu-upgrader through that, the window pops up in the right spot.
[19:43] <arraybolt3[m]> (Basically trying to handle multiple X displays.)
[19:43] <arraybolt3[m]> But without xdotool, I can't do that, which means that I have to figure out some other way to make it work, or we have to install xdotool.
[19:43] <arraybolt3[m]> Or I have to find a different tool to figure out which X display a window is on.
[19:44] <arraybolt3[m]> (It's not really that convoluted, the comments just make it look that way.)
[19:44] <tsimonq2> I say install xdotool, just be careful that it's a Pre-Depends if it needs to be.
[19:45] <tsimonq2> If it works, I'm not against it. Just make sure you test it, really really well :P
[19:45] <tsimonq2> https://launchpad.net/ubuntu/+source/lubuntu-update-notifier/0.5
[19:46] <arraybolt3[m]> OK. I'll need to make one more change to the update notifier to be able to repair the busted installation once it gets restarted, that will turn into version 0.5.1, right?
[19:47] <tsimonq2> It can, yeah :)
[19:50] <arraybolt3[m]> Simon Quigley: Oh my gosh, I just had a way easier idea.
[19:51] <arraybolt3[m]> We'll make the script detect if lubuntu-upgrader is being used to do the upgrading. If it is, it will just kill it and aptd... and leave it. The user can then be like "Oy, what happened?", restart the upgrade themselves, and bam, problem solved. No xdotool, no undocumented "fix it" switch, just blow everything up and make it easy for the user to clean up.
[19:52] <arraybolt3[m]> I think that will be way easier to implement, without causing too much trouble. Worst case scenario, the user will miss an update and then everything will be fixed the next time a new update comes through.
[19:52] <tsimonq2> I'm open to seeing your code :)
[19:52] <arraybolt3[m]> It will be the same as "let it freeze, the user can restart it themselves", only we break through the freeze.
[19:53] <arraybolt3[m]> OK, will submit soon-ish.
[19:55] <arraybolt3[m]> (This does have the significant downside of leaving a possibly CVE-riddled system if the user fresh-installed 22.04.1, updates, and the updater self-nukes and then lets the user pick up later. It will be fixed at the next update, but that's a pretty significant risk.)
[19:55] <arraybolt3[m]> (So maybe the self-resurrecting updater is a better idea.)
[19:56] <tsimonq2> arraybolt3[m]: Check what actually launches lubuntu-upgrader
[19:56] <tsimonq2> Maybe piggyback off of that slightly?
[19:56] <arraybolt3[m]> OK.
[19:56] <tsimonq2> Also, that's not the way to do it, in hindsight :P should be a systemd timer... how exactly it should be implemented, no clue at this current moment ;)
[19:58] <arraybolt3[m]> Hey, maybe we just install a systemd timer during the lubuntu-update-notifier update?
[19:58] <tsimonq2> Now that is too complicated for an SRU :P
[19:58] <arraybolt3[m]> Eh, maybe too much for an SRU.
[20:01] <teward> tsimonq2: taiga is pending, i need IS to update the DNS first before I can deploy
[20:03] <tsimonq2> teward: did you file an RT yet :P
[20:05] <tsimonq2> teward: throw me an IP, I'll point /etc/hosts >:P
[20:05] <teward> did i file an rt.  yes i filed the RT. numpty
[20:06] <teward> and that won't work because i haven't enabled the site yet in the config so SNI is a thing.
[20:06] <tsimonq2> XD
[20:06] <teward> ;)
[20:06] <teward> i need the DNS for LetsEncrypt
[20:06] <teward> so
[20:06] <teward> you can wait until tomorrow
[20:06] <arraybolt3> Who says we need HTTPS?
[20:06]  * arraybolt3 runs
[20:07] <teward> *trips arraybolt3*
[20:07]  * arraybolt3 falls over and knocks teward into a pool of lava
[21:01] <teward> lol
[21:01] <teward> right-o, so
[21:01] <teward> IS got the DNS done
[21:01] <teward> i got the cert generated
[21:01] <teward> i'm spinning Taiga up
[21:01] <teward> i require more caffeine
[21:26] <teward> oops i broke taiga.  btrfs on zfs is probably the problem.
[21:26] <teward> *rebuilds*
[21:49] <tsimonq2> teward: here have some Karuba Gold
[21:53] <arraybolt3[m]> Simon Quigley: I can see a flaw in my script. What happens if:... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/78ce874678b33226a1d2ef0296ac99a9062f49b1>)
[21:55] <arraybolt3[m]> The question in any of those scenarios is, which X display should be used? If there's only one X display, easy. If there's only one window, easy. If there's multiple windows on multiple X displays, the script is left with the question of which display to use.
[21:55] <arraybolt3[m]> Perhaps in that edge case the script should kill everything and leave it to the user to figure out that they need to relaunch lubuntu-upgrader? That way the user decides which X display is used.
[21:56] <arraybolt3[m]> (One assumes that a user who can use SSH with X forwarding knows what they're doing enough to be able to restart an upgrade.)
[22:04] <arraybolt3[m]> K, learning from our mistakes with "not implementing things for this one edge case", I'm now handling the multiple-windows-on-multiple-X-displays scenario.
[22:21] <arraybolt3[m]> Simon Quigley: This is how far I've gotten so far: https://termbin.com/8ebc7 The end is not actually functional, but I'm brainfried.
[22:23] <arraybolt3[m]> Oh, and this whole thing is almost entirely untested.
[23:52] <arraybolt3[m]> Made more progress - now lubuntu-upgrader is terminated, restarted, and disowned, and then aptd is terminated. That should let control transition smoothly to the new upgrader, and hopefully the user should notice nothing more than a window suddenly vanishing and reappearing.
[23:52] <arraybolt3[m]> I still have a tiny bit more logic to work out for dealing with the highly unlikely "windows scattered everywhere" scenario, but it's almost done!
[23:53]  * arraybolt3[m] takes a break