[00:06] adding a .xscreensaver file to a 22.10 system APPEARS (to me thus far) to allow xscreensaver to open; it can't be blank (ie. `touch` won't help) but even adding a debian .xscreensaver causes it to open (ie. any ~valid file) [00:07] That's weird. It didn't used to require that, did it? [00:09] we've got the xscreensaver issue on the tracker (I assume in release notes too, sorry i forget) ... (why I was trying to go on discourse) .. nah I don't think it did but I can't be certain [00:09] guiverc: Gah! I think I know what is going on [00:09] * arraybolt3 was thinking of attempting a git bisect, but it sounds like kc2bez has it covered === Eickmeyer5 is now known as Eickmeyer [00:12] issue didn't get on release notes; only in the 22.10 tracker [00:13] https://github.com/lubuntu-team/lubuntu-default-settings/blob/ubuntu/kinetic/src/usr/lib/X11/app-defaults/XScreenSaver [00:14] Sounds like we need to remove that ^ [00:16] [telegram] Lubuntu website question. 22.04 shows as the NEW version on the front page. Should that be updated or is the LTS supposed to be featured? [00:17] I'm not on kinetic any longer here, can't look what I have locally.. My primary focus is doc/workaround for 22.10 users we can documented on discourse; or can a SRU attempt to correct issue? [00:17] @franksmcb: What website are you visiting? [00:17] Lubuntu.me is the official site. [00:17] Good question, we should update it [00:17] [telegram] Lubuntu.me [00:17] Oh on the front page. Right, that's outdated on our end. The Downloads page will have 22.10 at the top. [00:18] [telegram] I’m like a calendar, right once a year. [00:18] @franksmcb: Thanks! [00:20] :) [00:22] I also note a jammy daily; thanks tsimonq2 [00:38] WordPress and I are not friends I guess. I can't seem to edit the main page. [02:16] *burps* [02:16] the git site is up [02:16] * arraybolt3 throws an angry python at teward's server [02:16] *noms the python* [02:16] *was hungry* [02:16] * arraybolt3 gets tangled up in said python and dies [02:17] OK I guess then it's back to migration work. Thank you! [02:45] kc2bez[m]: need help with the site? [02:45] or is it busted [08:05] "Dan Simmons: need help with..." <- Yeah, the editor wouldn't launch. I got logged into the dashboard but couldn't use Elementor. [09:22] is breakage of Discourse forum something to report here or on #lubuntu ? [09:42] per https://lists.ubuntu.com/archives/lubuntu-users/2022-October/012682.html it is fully broken in general [13:34] Good morning [13:34] Sup [13:34] Discourse broken it seems? [13:35] Yes, needs teward love sprinkled on it. [13:36] Sounds good [13:38] Simon Quigley: I'm not sure if you read through the screensaver issues or not. It looks like we need to drop our customization stuff possibly. [13:39] kc2bez[m]: I did read through it but only partially understood: what is the root bug we're chasing? [13:41] The application crashes making it impossible to adjust the settings. [13:41] Oh, wow... did the XDG paths change maybe? [13:43] I think XscreenSaver updated and there may be an incompatibility with our config. [13:44] https://bugs.launchpad.net/ubuntu/+source/xscreensaver/+bug/1989163 [13:44] -ubot93:#lubuntu-devel- Launchpad bug 1989163 in xscreensaver (Ubuntu) "Unable to run screensaver when selected from menu" [Undecided, Confirmed] [16:38] o/ [16:38] Alright, so now it's time to finish the migration work. Sheesh, I hate $PERSONAL_LIFE's insanity sometimes. [17:23] Simon Quigley: Alright, someone tell me what I'm doing wrong here. I'm fighting with Phabricator's Conduit API. I'm trying to pass a JSON blob that looks like this:... (full message at ) [17:24] If I could make a JSON blob as a string and somehow pass that, I bet it would work. [17:30] But for some reason it wants form-encoded data (whatever that means, I'm sure it means something but I sure don't know), so it won't take a JSON string as a data parameter, and if I try to pass a JSON string as a json parameter, it tells me that sending a JSON-encoded body isn't supported. [17:30] (╯°□°）╯︵ ┻━┻ [17:53] I just don't get it. If I do all of the dictionary building in a variable and the json.dumps() the variable to look at the result, it comes out looking like exactly the format the web console is telling me to use, so Idk what it's grumpy about. I find it severely unlikely I've found a flaw in Python Requests, but I'm starting to think it may be quicker if I just use curl to short-circuit the process since there's an example for [17:53] how to do exactly what I'm trying to do using curl directly in the web console. [17:57] aHA! I am victorious! Alright, so, note to self, Phabricator outputs JSON, it does not like taking JSON. Format your dictionary elements like you were putting them in data parameters of curl, and it works. [18:39] s/severely/extremely/ [18:39] teward: Have we already snapshotted Gitea? I'm just about done writing the Phabricator exporter, and then will be writing the Gitea importer. I find it highly unlikely that anything will go wrong since I'm not going to overwrite anything that already exists, but better safe than sorry. [18:43] it's been taking too long so i'm going to copy it from the lxd infra box to my local system so give me a bit [18:43] it's transferring a copy now but isn't the fastest [18:43] No prob. I won't do any write operations until I have a thumbs-up from you. [18:57] Simon Quigley: I think this reddit support request is good now. https://www.reddit.com/r/Lubuntu/comments/y9kbc3/lxqt_panels_custom_command_widget_missing/ [19:12] "I think XscreenSaver updated and..." <- I say fix the compatibility issue instead of throwing the baby out with the bathwater :) [19:13] "aHA! I am victorious! Alright..." <- Thanks for sticking through that :) [19:14] "Simon Quigley: I think this..." <- Thank you very much! Well done :) [19:14] tsimonq2: That is fair. [19:15] BRING OUT THE LIST! [19:15] roaring in the distance [19:16] tsimonq2: Can I just say I really wish the API docs hadn't spent so much time telling me how to use JSON as input when it turned out that if you're not using their special client, you don't use JSON?!? [19:16] tsimonq2: ```... (full message at ) [19:16] I am not sure we need/want our customizations though. The dialog has changed. It needs some testing. [19:16] > <@tsimonq2:linuxdelta.com> BRING OUT THE LIST!... (full message at ) [19:16] Sorry, that was a reply [19:16] arraybolt3[m]: Let's be fair here... Phab has been deprecated for a bit now and will show more signs of bitrot the more we use it :) [19:17] tsimonq2: Fair enough. Glad we're about to be done with it. [19:17] * ```... (full message at ) [19:17] > <@arraybolt3:matrix.org> ```... (full message at ) [19:17] (Changed the list a few times, re-review it) [19:17] kc2bez[m]: Good to look at, right about now in the cycle :) let's do it [19:18] Channel topic done? [19:18] * arraybolt3[m] points at channel topic, it no longer says Kinetic, it will need changed one more time once LL is named [19:18] So I guess not really done... [19:18] Thanks for the clarification :) [19:19] * ```... (full message at ) [19:19] Alright, another list update. [19:19] * arraybolt3[m] slaps this onto CodiMD so we can all see and edit [19:19] arraybolt3[m]: Thomas has op powers now [19:19] Nice 👍️ [19:22] Alright what does everyone think of this? https://notes.lubuntu.me/huOk59_iRSaAMZDl_my8bw [19:23] That way we can add, move, reprioritize, whatever. I prioritized it according to what I think is right, but obviously it might not be all correct yet. [19:23] Will look when I'm back at my computer. I've been running errands [19:32] Btw, you'll be delighted to hear that Gitea is fully supported by ubot93 wrt issue, pull request, and commit snarfing - as opposed to Phab previously. [19:32] krytarik: \o/ [19:35] Oh also somewhere around here I have a running bug list of things that need taken care of at some point, I think it's on my Chromebook which I don't have near me right now, but I should also upload that and present it for destruction. [19:45] arraybolt3: re: bug 1989203 - I subscribed ~ubuntu-sponsors in case I don [19:45] -ubot93:#lubuntu-devel- Bug 1989203 in manuskript (Ubuntu Kinetic) "Manuskript crashes on start" [High, Confirmed] https://launchpad.net/bugs/1989203 [19:46] * I don't get to it [19:46] Simon Quigley: Thanks, I was just thinking of asking about that. [19:46] you get a 6th sense for those kind of things eventually ;) [19:50] Another thing to mention... [19:50] Does anyone have a laptop with an Ethernet port? Try connecting to WiFi and plugging in the ethernet cord [19:50] * arraybolt3[m] has multiple laptops like that [19:50] One of them is even free for testing at the moment. [19:50] The ethernet connection doesn't even come up for me. `ip addr` to test [19:51] Jammy? Kinetic? [19:51] Kinetic [19:52] K, guess that will be the next test once I'm finished fighting the Artificial Perpetually-grumpy Intelligence... er. API. [19:53] > <@tsimonq2:linuxdelta.com> Another thing to mention... [19:53] > [19:53] > Does anyone have a laptop with an Ethernet port? Try connecting to WiFi and plugging in the ethernet cord [19:53] I can test that tonight. [19:53] kc2bez[m]: Sounds good [19:55] The manual workaround I have so far is to, in this order: [19:55] To me that sounds like either a severe Network Manager bug or a hardware problem. [19:55] 1. Disable WiFi via nm-tray [19:55] 2. Run `sudo dhclient -r NEW_NAME_FOR_eth0` to *release* the DHCP client process on that interface [19:55] 3. Run `sudo dhclient NEW_NAME_FOR_eth0` to restart the DHCP client process [19:56] Also how old is your laptop? Should I be doing this on my ancient Elitebook8570p, or should I boot my KFocus with a live ISO? [19:56] Both have an Ethernet port. [19:57] I've seen other reports in the mist, if it is a hardware problem I'd be curious to see "of what nature", if it's a kernel problem, we're screwed, and if it's nm territory, we need to see if we're the only flavor with this problem [19:57] It's a Thinkpad T480s [19:57] There was a recent (today?) kernel update addressing WiFi issues. [19:58] For 5.19 [19:58] ...oh? [19:58] $ uname -r [19:58] 5.19.0-23-generic [19:58] https://9to5linux.com/ubuntu-22-10-gets-first-kernel-security-update-to-address-recent-wi-fi-stack-vulnerabilities [19:59] tsimonq2: Heh, my systems hit on either side of that one's age. One is a 3rd Gen Core i5, the other 11th Gen, yours is 7th Gen... I guess I'll test both and see what happens. [19:59] *7th or 8th gen [20:00] I shouldn't patch the kernel myself... I shouldn't patch the kernel myself... every time I say that it becomes less convincing :P [20:00] tsimonq2: Then livepatch the kernel yourself? [20:01] I don't know all the details of how using Linux's livepatching features works, but that might actually work. [20:01] Well, one simple way to see if it's a local or a real issue would be to see if anyone else can reproduce it [20:01] If you want to make a patch for just your own machine. [20:01] Simon Quigley: Another thing you might try is falling back to the normal kernel - it just dawns on me you use the mainline one. [20:02] tsimonq2: ( https://www.scientificamerican.com/article/the-universe-is-not-locally-real-and-the-physics-nobel-prize-winners-proved-it/ ) [20:02] arraybolt3[m]: I switched back after hearing about the LED screen stuff [20:02] Just for the time being, heh [20:02] Ah. [20:02] I think that was actually in 5.19, and was fixed? Don't remember for sure. [20:03] Might've been 5.18 [20:04] I dunno, I saw the headline and panic-switched back :P [20:05] I do still use a LUKS + btrfs setup though [20:05] It was 5.19.12, I believe? it was fixed in 5.19.13. [20:05] https://arstechnica.com/gadgets/2022/10/linux-5-19-2-kernel-can-flicker-and-damage-screens-on-some-intel-gpu-laptops/ [20:07] ¯\_(ツ)_/¯ I've been seriously considering switching back [20:07] Anyway, here's the perspective I take on this... [20:08] If it's only my issue, I want to take the time to see why the issue is caused, to perhaps reveal an underlying problem [20:13] "I do still use a LUKS + btrfs..." <- Rebel. [20:20] If someone else can reproduce it, and it's a real bug, we can definitely consider patching it for everyone [20:23] "Rebel." <- ikr heh [20:24] Well, maybe I should just grab the mainline kernel, hm :P [20:24] If it starts working right with the mainline kernel, then you get to bisect it! [20:25] https://thumbs.gfycat.com/AbleMelodicHapuka-size_restricted.gif [20:35] Followup on that issue...... (full message at ) [20:35] s/aquire/acquire/ [20:36] nm regression, bet you. [20:36] Also now my snaps won't load lol [20:36] XPCOMGlueLoad error for file /snap/firefox/1959/usr/lib/firefox/libmozgtk.so: [20:36] /snap/firefox/1959/gnome-platform/usr/lib/x86_64-linux-gnu/libgdk-3.so.0: undefined symbol: XIFreeDeviceInfo [20:36] Couldn't load XPCOM. [20:36] * arraybolt3[m] quotes the bit about mainline kernels being unsupported, guess now we know it's true! [20:37] hahahahaha [20:37] let me see if I can get it to work, and report back on my steps if I do [20:37] Sounds good. Might also be interesting to see how Debian LXQt behaves (I think it uses ConnMan rather than NetworkManager?). [20:38] Yeah, it does [20:39] also if other n-m flavors exhibit this bug, we know it's n-m [20:39] Especially if the same version of NM on non-Ubuntu distros does this (perhaps give Arch a shot?). [20:39] Anyway... I'll boot back into the supported kernel so I can get work done, until this bug is fixed :P [21:12] * arraybolt3[m] is going to vanish for a bit if all goes well - rebooting after a system upgrade and also installing KDE updated from -proposed [21:12] (Yes, on my production system. I like playing with fire sometimes I guess.) [21:56] Alright so that was weird. I now know how to make Gitea repos from the API, and push to them using Git manually. Only thing I don't think I have the brains to automate right now is the username and password entry for each git push, so I'm just going to finish up here, and then plug in my username and password several times while it migrates everything over. [21:57] (And yeah an SSH key would probably have automated even the authentication but I don't have the brains for that right now :P) [22:00] Simon Quigley, teward, Dan Simmons: Alright, I am about to pull the "Migrate" lever, are we ready? [22:01] Or do I hand the script over to someone else at this point and they do it? [22:06] This is the script: [22:06] * arraybolt3[m] sent a code block: https://libera.ems.host/_matrix/media/v3/download/libera.chat/8c6da1f6afb31328014f7b3946c2c98b5370cade [22:06] And yes, I am a bit embarrassed that it took me this long to come up with something that looks this simple, but at least it's ready. [22:06] (I believe.) [22:09] Heh, I forgot to double-check some of the names, one moment... [22:10] "is going to vanish for a bit..." <- stable-proposed isn't devel-proposed, I think you're safe :) [22:10] just do it when you're ready its why i made a backup of gitea [22:11] * ```... (full message at ) [22:11] Alright, had one name wrong, fixed it. [22:11] "Alright so that was weird. I now..." <- git clone https://GH_USER:KEY_NOT_YOUR_PASSWORD!!!/etc. [22:12] "And yes, I am a bit embarrassed..." <- I mean I guess some of those Phabricator JSON monsters were tough cookies to crumble so it's not that simple but... [22:12] tsimonq2: Oh nice. [22:13] Key being an API key or a personal access token? [22:13] "And yes, I am a bit embarrassed..." <- I mean, here's the thing... I didn't give you a lot of info to start this project and you built it in less than a week. Don't be sorry for anything :) [22:13] arraybolt3[m]: Latter [22:14] > <@arraybolt3:matrix.org> ```... (full message at ) [22:14] Odd, Gitea doesn't even have personal access tokens... [22:14] That I can tell. [22:14] tsimonq2: #!/usr/bin/env python3, right? [22:14] arraybolt3[m]: That syntax is how you'd feed it info [22:14] That's what I see used a lot. [22:15] GiTea may not support those types of tokens... but you may be able to use your password or an API key? [22:15] My password works, I just noticed you said "not my password" so that made me a bit gunshy. [22:15] "Also now my snaps won't load lol" <- *cough* Flatpak 😂 [22:15] arraybolt3[m]: Not good practice. Works though :) [22:16] kc2bez[m]: *someone* on the team has to actually use snaps, I won't pawn that off on anyone else :P [22:17] * arraybolt3[m] likes snaps but doesn't use them since Chrome works better... [22:17] ¯\_(ツ)_/¯ [22:17] https://gitpython.readthedocs.io/en/stable/ [22:17] Also arraybolt3 @arraybolt3:matrix.org... [22:17] "git clone https://GH_USER:..." <- Eh, I'd have to tear up too much code to do that, considering the number of repos involved. [22:18] tsimonq2: Oh that looks nifty. [22:18] Alright, activating migration in three... two... crud wait I have to make sure these go under the right organization. [22:19] tsimonq2: I kid. I have some snaps that I use. [22:21] Umm... ok so it turns out I can *create* a repository as the Lubuntu organization, I just can't delete it thereafter. So now I have "arrayboltAPITest" stuck in Lubuntu... [22:21] Hopefully someone can delete that. [22:21] I think my all time favorite snap is LXD [22:21] I should be able to delete stuff [22:22] kc2bez[m]: Yes, definitely kidding, sure ;) [22:24] Alright, carefully starting migration... (just gonna make sure I don't make a ton of garbage repos and then fail to fill them with anything) [22:25] Oh lovely I missed a couple of slashes, glad I proceeded carefully! [22:28] * kc2bez[m] uploaded an image: (52KiB) < https://libera.ems.host/_matrix/media/v3/download/matrix.org/lKjJpsdFSdapeURyYgMsmHvi/Screenshot%20from%202022-10-26%2018-27-49.png > [22:28] Cue Kenny Loggins [22:29] Yeah I don't see that section when I look there. [22:29] I think it needs Admin privs. Want me to delete it? [22:29] Dan Simmons: Yeah, and also lubuntu-meta while you're at it since my code boffoed and made that one and then errored out. [22:30] ok [22:30] (Failed to increment a counter :P) [22:32] Poof They are gone. [22:32] Thanks! [22:32] Alright, here goes! [22:33] Shoot, now Git is giving me errors I've never seen before... [22:33] Hmm, whatcha got? [22:33] Probably a failed chdir somewhere that I tested first and then botched once it was all the way in the code. [22:34] Oh, that would do it. [22:34] 🤦‍♂️ Sure enough. Alright, let's just make the repos this step and then populate them a second step. [22:36] Sap, that won't work. [22:38] Alright, it's working this time. [22:39] what in the world... now Gitea just timed out during a git upload.... (full message at ) [22:40] Sigh. I have all five repos locally and all five made remotely, let's just finish this with some git pushes and call it a day. [22:43] Dan Simmons: OK so it looks like Gitea doesn't want to take lxqt-l10n-packaging. It fails with a Payload Too Large error as described above. [22:43] I thought it was a timeout, but I don't think so after all. [22:44] We don't need that anymore [22:44] Oh good. [22:45] Wow. So all that work to migrate a whole whopping four repos, three of which ended up having to be done halfway manually. [22:45] :P Well hey, at least we're done. [22:46] That is awesome! [22:46] Anyway, there's now an lxqtmeta-packaging repo that could use deleted, other than that I think we have conquered the Git repo migration. [22:46] And I learned a lot, so that's good! [22:47] Thank you guys for your patience with me, and hey, we're done! [22:47] arraybolt3[m]: **thank you** for this :) [22:47] 🎉 [22:49] Final notes for myself after this. One, just because you tested it in the Python REPL and it worked, doesn't mean it will work once you plug it into the source code. Something may be different and mangle everything. Two, if your only good way of testing involves running something in production, find some other way to do it. Three, if you really can't find some other way to do it, save yourself the headache and run the whole thing [22:49] in a debugger for at least the first loop. [22:50] Had I followed all three of those (or even just two of the three!), the whole migration would have been automatic except for that one bit with lxqtmeta-packaging. [22:51] * ```... (full message at ) [22:51] And also I went back and edited that script I posted first with all of the fixes. [22:52] Thankfully between teward's snapshot and Dan's willingness to help with deleting garbage repos, nothing was lost except for time. So woot! [22:52] Thank you guys! [22:53] I'm glad you learned something, thank you to Dan and Thomas for assisting :) [23:29] "Anyway, there's now an lxqtmeta..." <- Er... lxqt-l10n-packaging needs deleted. Wow, typo. [23:39] Ok, just got back. I can do some cleanup.