| lubot | [telegram] <kc2bez> Thanks for the update @teward001 | 00:07 |
|---|---|---|
| lubot | [telegram] <teward001> arraybolt3: that's just for the initial deployment. unless Lubuntu wants to fund the purchase of a dedicated NUC or something (#NotCheap) any new hardware is at least $250 and i'm running out of space to put infra (my rack is already full...) | 00:59 |
| lubot | [telegram] <teward001> (also i have a lot of VPS credit at the providers i use so :P) | 01:00 |
| lubot | [telegram] <teward001> (and the VPS is not expensive by CPU or RAM usage, it's just $10/month) | 01:00 |
| lubot | [telegram] <teward001> i mean, if Lubuntu *wants* me to spend all our money on bare metal equipment I can but... :P | 01:01 |
| arraybolt3[m] | @teward001: I guess it's hard to beat $10 a month (would take over 4 years to outdo the cost of my build), and we're probably not going to use up all the bandwidth, and we can easily live with the build speed. I thought it would be cheaper and more efficient. | 01:01 |
| * arraybolt3[m] hates all things subscription :-P but in this instance it makes sense | 01:03 | |
| ArrayBolt3Quasse | Eickmeyer: Well, I did the Quassel test (that's what I'm typing from now), and while it is set to libera.chat by default, it also joined the #quassel network by default, not #ubuntu. | 01:11 |
| Eickmeyer[m] | Weird. | 01:12 |
| arraybolt3[m] | (ArrayBolt3Quasse is me - I tried to name myself ArrayBolt3Quassel, fail!) | 01:12 |
| Eickmeyer[m] | Did you change it to the UbuntuIRC network as opposed to libera? | 01:13 |
| ArrayBolt3Quasse | I didn't see that option - it gave me secure irc.libera.chat, and non-secure irc.libera.chat. | 01:13 |
| Eickmeyer[m] | Hmmm... | 01:13 |
| * arraybolt3[m] uploaded an image: (240KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/EIQveOxvzPSbhQklSUgHWWdd/image.png > | 01:14 | |
| arraybolt3[m] | Eickmeyer: ^ | 01:14 |
| lubot | [telegram] <teward001> oh i mean, we have to discuss disk IO really (re @lubuntu_bot: (irc) <arraybolt3[m]> @teward001: I guess it's hard to beat $10 a month (would take over 4 years to outdo the cost of my build), and we're probably not going to use up all the bandwidth, and we can easily live with the build speed. I thought it would be cheaper and more efficient.) | 01:16 |
| lubot | [telegram] <teward001> case inpoint the existing infra is a little tiny bit slow due to it being spinny disks | 01:17 |
| lubot | [telegram] <teward001> VPSes are all running on SSDs | 01:17 |
| lubot | [telegram] <teward001> i have a mix of disks in my infra but not enough compute for everything | 01:17 |
| lubot | [telegram] <teward001> (disk is not the issue, compute is) | 01:17 |
| arraybolt3[m] | lubot: My build included a 128GB NVMe SSD. | 01:17 |
| arraybolt3[m] | But I like using SSDs for our compute. And I like the low cost of the VPS, that's awesome. | 01:18 |
| Eickmeyer[m] | arraybolt3: Yeah, I see what's going on. It's defaulting to quassel's hardwired defaults because they have a channel on libera already. It's conflicting with the defaults I wrote, nullifying ours. @Leokolb Dan Simmons | 01:18 |
| arraybolt3[m] | @teward001: Forgive my ignorance, but wouldn't just one server be enough to cope with most of the stuff we have? I mean, the website isn't that heavyweight, is it? Nor is Discourse, the manual, or even Git. If you've already filled up one rack, couldn't you just merge multiple features into one server and then use a free server to do Drone? (Asking partially because I'm curious, partially because I want to be helpful, and | 01:22 |
| arraybolt3[m] | partially because I want to learn how to use servers one day) | 01:22 |
| Eickmeyer[m] | I'm not really sure there's much we can do, but it's better than sending people to OFTC/#debian | 01:23 |
| arraybolt3[m] | We had this working before though, when it pointed to #lubuntu:libera.chat . | 01:23 |
| Eickmeyer[m] | Yeah, that's true. | 01:23 |
| * Eickmeyer[m] inspects a few things | 01:23 | |
| Eickmeyer[m] | Aha! Found the problem. | 01:25 |
| Eickmeyer[m] | This doesn't look good.... | 01:25 |
| Eickmeyer[m] | I"m going to have to upload yet another fix because I have to 86 a line from the same file, perhaps with a different patch. | 01:26 |
| Eickmeyer[m] | SRU team is going to kill me. | 01:26 |
| arraybolt3[m] | Eickmeyer: Hey, that's why we test. I'm sure you already know this, I'm just reminding you, but we can always use a PPA for preliminary testing. | 01:27 |
| lubot | [telegram] <teward001> arraybolt3: if Simon tries to hammer CI like he used to with Jenkins or whatever it was, then no, one server is not enough at only 2 builds at a time | 01:28 |
| arraybolt3[m] | @teward001: Then how is RAMNode going to cope? | 01:28 |
| lubot | [telegram] <teward001> it's capped at 2 builds at a time | 01:29 |
| lubot | [telegram] <teward001> the point is | 01:29 |
| arraybolt3[m] | Oh. | 01:29 |
| lubot | [telegram] <teward001> we can't be rebuilding the entire Lubuntu packageset daily when there's no updates | 01:29 |
| lubot | [telegram] <teward001> we can do one off build triggers, but we can't do daily triggers for 30 packages on 4 different releases | 01:29 |
| lubot | [telegram] <teward001> or we can do on pull/push requests to the repos do a build | 01:30 |
| arraybolt3[m] | Good grief, I didn't know we used to do that... | 01:30 |
| lubot | [telegram] <teward001> but Simon had the thing targeting LTS and Devel and Current Interim Release for like 30 packages and repos | 01:30 |
| arraybolt3[m] | (Why would we do that?) | 01:30 |
| lubot | [telegram] <teward001> ask Simon why he did that | 01:30 |
| lubot | [telegram] <teward001> because that was Simon's baby | 01:30 |
| arraybolt3[m] | And Drone just builds EVERYTHING at once? | 01:30 |
| lubot | [telegram] <teward001> nope | 01:30 |
| lubot | [telegram] <teward001> think of it like GH Actions. You have to *do something* to trigger a build on a given repo | 01:30 |
| lubot | [telegram] <teward001> each repo is configured with its own .drone.yml for what tests/builds/etc it has to run for each repo | 01:31 |
| arraybolt3[m] | So we just used to keep triggering it to build everything at once? | 01:31 |
| lubot | [telegram] <teward001> so if you do that for a-la the welcome screen greeter component and push a change into the repo, it'll trigger a Drone build | 01:31 |
| lubot | [telegram] <teward001> for that package/repo per the config | 01:32 |
| lubot | [telegram] <teward001> Simon had cronjobs set up to run everything daily on Jenkins | 01:32 |
| lubot | [telegram] <teward001> it literally hammered any system it was on | 01:32 |
| lubot | [telegram] <teward001> and had way too many artifacts, etc. so it just exhausted everything | 01:32 |
| teward | so the issue here was | 01:32 |
| teward | CI wasn't being used the way CI should be | 01:32 |
| teward | if you're going to do daily rebuilds of the repos, you pay Canonical for that shit :P | 01:33 |
| teward | or shell out lots of cash for dedicated systems/metal | 01:33 |
| teward | which SImon couldn't do for... reasons he knows | 01:33 |
| arraybolt3[m] | If we can cap the builds with Drone on VPS, can't we cap them on Drone on physical hardware? | 01:33 |
| lubot | [telegram] <teward001> yes we can. but we still need the hardware and need to pay for it | 01:33 |
| arraybolt3[m] | (If you get tired of this, just tell me, I'm interested, but I don't need to suck your time away if you have other things to do.) | 01:33 |
| lubot | [telegram] <teward001> it's 9:33PM i'm literallyj ust watching an artist friend of mine stream (i'm a mod in her twitch chat) and am here | 01:34 |
| lubot | [telegram] <teward001> so i have not much else to really do at the moment | 01:34 |
| arraybolt3[m] | My system can build a package in a matter of minutes and it's an old clunker of a system I got for free when a business upgraded. Why wouldn't something like that work as a Drone server? | 01:34 |
| arraybolt3[m] | (If we use it right?) | 01:34 |
| kc2bez[m] | <Eickmeyer[m]> "I'm not really sure there's much..." <- I agree | 01:38 |
| Eickmeyer[m] | <kc2bez[m]> "I agree..." <- Well, I figured out what to do, but I'd have to do further patching. | 01:44 |
| arraybolt3[m] | @teward001: I guess even cheap physical hardware is pricier than the VPS setup you have, plus old cheap hardware will eventually be outmoded and cost us money to replace, whereas RAMNode will upgrade their hardware as needed for us. | 01:45 |
| kc2bez[m] | Eickmeyer[m]: I read that further down. Keep us posted and we will give it a test. | 01:47 |
| Eickmeyer[m] | arraybolt3: Can you do me a favor and post your findings on the SRU bug report for quassel and mark as verification failed? That might smooth the process. Attach your screenshot if you can. | 01:48 |
| arraybolt3[m] | Eickmeyer: np | 01:49 |
| arraybolt3[m] | Eickmeyer: Anything special I have to do to mark as verification failed? | 01:50 |
| Eickmeyer[m] | arraybolt3: Up top in the tags, switch verification-needed-jammy and verification-needed to verification-failed-jammy and verification-failed and then in the comments explain why (going to #quassel instead of #ubuntu as intended in the patch) | 01:52 |
| lubot | [telegram] <teward001> arraybolt: don't get me wrong if I upgrade my LLC's infra eventually I'll happily put the server in an environment where it loves to eat memory and such | 01:53 |
| lubot | [telegram] <teward001> but right now compute is expensive (worldwide chips shortage) | 01:53 |
| arraybolt3[m] | And we probably don't want to put our whole entire CI infra on a used system that may kick the dust at any time. (If we didn't mind using non-open-source Git, we could just use GH Actions, it's free for public repos.) | 01:54 |
| arraybolt3[m] | Eickmeyer: If you're wondering why the bug report still hasn't been updated, I trashed my VM before you asked for me to do the report, and I wanted a better screenshot (that of the opening screen of Quassel where it says it's only going to join #quassel), so I'm building a new VM, which takes a while on my Internet connection. | 02:19 |
| Eickmeyer[m] | arraybolt3: Don't worry about the quality of screenshot, just grab the one from up above. | 02:27 |
| arraybolt3[m] | If you say so, but I'm just about done. | 02:28 |
| arraybolt3[m] | I just got the good screenshot, it is OK to use it, or do you prefer if I use the other one? | 02:30 |
| arraybolt3[m] | Actually, I'll just put both. | 02:30 |
| Eickmeyer[m] | Meh, doesn't matter. I'm just at a point where I need to reupload for kinetic, and I'm going to have a hard time justifying that until I have something with a verification failed at this point. | 02:32 |
| arraybolt3[m] | Hold on, using Matrix as my screenshot maker... | 02:33 |
| * arraybolt3[m] uploaded an image: (218KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/AFhRfCSJZgXUTAawkJYPYErE/image.png > | 02:33 | |
| Eickmeyer[m] | Did you change the tags yet? | 02:33 |
| arraybolt3[m] | Not yet, about 30 seconds from doing that... | 02:34 |
| Eickmeyer[m] | kk | 02:34 |
| arraybolt3[m] | Do I also put (verification-failed-jammy)? | 02:34 |
| Eickmeyer[m] | Yes. | 02:34 |
| arraybolt3[m] | Both tags changes. | 02:35 |
| arraybolt3[m] | Er, chaned. | 02:35 |
| Eickmeyer[m] | Cool | 02:35 |
| arraybolt3[m] | AGH! Changed. (Good grief, I'm typo-happy tonight.) | 02:35 |
| Eickmeyer[m] | This is like those autocorrect typo issues on reddit.com/r/ihadastroke | 02:35 |
| arraybolt3[m] | Oh good grief. Those are so much worse. | 02:37 |
| arraybolt3[m] | INMEANY CALL UPO... CALL IOU... UOU... YOU | 02:38 |
| Eickmeyer[m] | HAHAHAHAHA | 02:38 |
| Eickmeyer[m] | (Breathe, you can do it) | 02:39 |
| arraybolt3[m] | Reminds me of the behavior of one of my coworker's Samsung phones - if you used the drag-to-type feature just right (start your finger at the C, move up to 6, release), it would spit out a UUID. | 02:45 |
| Eickmeyer[m] | Amazing if you need to do a quick fstab. | 02:47 |
| Eickmeyer[m] | Ok, uploaded new quassel, this time with an all-new patch. Hopefully the SRU team has mercy on me. | 03:03 |
| arraybolt3[m] | Eickmeyer: Any clue on whether that Quassel build is done? I'm not even seeing it in Launchpad. | 03:54 |
| arraybolt3[m] | s/clue/ETA/, s/whether//, s/is done// | 03:55 |
| arraybolt3[m] | (Phrased message poorly at first) | 03:55 |
| Eickmeyer[m] | Jammy or Kinetic? | 03:55 |
| arraybolt3[m] | Jammy | 03:55 |
| Eickmeyer[m] | It won't until someone from the SRU team let's the upload through the queue. We got lucky when Robie hadn't ended his day earlier. Lukas is next. | 03:56 |
| arraybolt3[m] | Ah. OK, will wait, should I test on Kinetic now? | 03:57 |
| Eickmeyer[m] | That might still be in kinetic proposed, give it a couple hours. | 03:57 |
| arraybolt3[m] | 👍️ | 03:57 |
| arraybolt3[m] | Simon Quigley (Developer): Can you (or anyone else here) please tell me what I'm doing wrong here? I'm trying to do a Lintian override for very-long-line-length-in-source-file, and no matter how I try, I can't get it to take it. It either gives me a Lintian error and doesn't block the messages, or it fails silently. lxqt-panel is throwing me an onslaught of these messages. | 04:06 |
| lubot | [telegram] <teward001> arraybolt3: you can probably ignore that safely, from a Debian perspective experimental and pedantic level tags are not really ones you need to worry about really chasing down | 04:16 |
| arraybolt3[m] | OK. Maybe it's just that it's a pedantic message that makes it seemingly impossible to override? | 04:17 |
| lubot | [telegram] <teward001> technically speaking every tag is overrideable, the quesiton is where are you doing it and are you doing it properly. however for pedantic level tags you can just let those lintian notices go | 04:20 |
| lubot | [telegram] <teward001> i can't even see what that tag *means* because lintian.d.o is fubar | 04:20 |
| lubot | [telegram] <teward001> > The source file includes a line length that is well beyond the normally human made code line length. | 04:21 |
| arraybolt3[m] | I did verbatim what the docs told me to do (lxqt-panel source: very-long-line-length-in-source-file). Fail. Also failed if I removed "source" or added an * at the end. I'm doing it in a file simply named "lintian-overrides"... and now it just clicked what I'm probably doing wrong. | 04:21 |
| arraybolt3[m] | I probably needed a different filename. | 04:22 |
| lubot | [telegram] <teward001> if it's a source package lintian warning then i think you did it right but i need sleep so i can't look at it right now | 04:22 |
| arraybolt3[m] | Anyway, I'll ignore it, but I bet that was the problem, and that will help me in the future (if that was in fact the problem). | 04:22 |
| lubot | [telegram] <teward001> long story short though i'd ignore it | 04:22 |
| arraybolt3[m] | OK. | 04:22 |
| arraybolt3[m] | Thank you for your help! | 04:22 |
| lubot | [telegram] <teward001> pedantic severity lintian tags aren't going to ding you in upload, etc. yes, packages should be lintian clean, but pedantic is really not a level of severity that needs to be worried about | 04:23 |
| lubot | [telegram] <teward001> long term if you REALLY Want to chase it | 04:23 |
| lubot | [telegram] <teward001> but you don't really have to in the short/mid term | 04:23 |
| lubot | [telegram] <teward001> i'd love to see your logs but i'm lazy too so | 04:23 |
| lubot | [telegram] <teward001> (the logs might help to indicate what really needs to be done for the override) | 04:23 |
| arraybolt3[m] | Eh, we can worry about it tomorrow. | 04:23 |
| lubot | [telegram] <teward001> (but for now, sleep is required) | 04:23 |
| lubot | [telegram] <teward001> yup | 04:24 |
| teward | okay yeah so what you have to do is actually | 04:24 |
| teward | *digs into systemd source* | 04:26 |
| arraybolt3[m] | Man, how fast do you type? I usually know it's you talking because I hear my computer ding at me rapidfire LOL | 04:26 |
| arraybolt3[m] | (I can do 80-90+ WPM, you seem to be in the 120+ range) | 04:26 |
| teward | so d/source/lintian-overrides - `pkgname source: very-long-line-length-in-source-file` | 04:27 |
| teward | at least, if you're using Sid's lintian | 04:27 |
| teward | oh i am a fast typer, but i also use the enter key excessively | 04:28 |
| arraybolt3[m] | Did that, fail. I'm using Kinetic's Lintian, which I know has differences. | 04:28 |
| teward | yeah that might be a factor | 04:28 |
| teward | but i would also point out that right now Lintian has no maintainer in Debian, sooooooooooooooo | 04:28 |
| teward | there are bugs in tests :p | 04:28 |
| arraybolt3[m] | Yikes. Seems like a pretty important package to maintain. | 04:28 |
| teward | well it's not *abandoned* but they want help | 04:29 |
| teward | https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1012289 | 04:29 |
| ubot93 | Debian bug 1012289 in wnpp "RFH: lintian -- Debian package checker" [Normal, Open] | 04:29 |
| teward | and technically 'orphaned' bu i digress | 04:29 |
| arraybolt3[m] | Part of me wants to try and adopt it, but I probably don't know the right coding language. | 04:30 |
| teward | *cough* perl is painful | 04:30 |
| teward | and there's a whole team in Debian for it so | 04:30 |
| teward | adoption is unlikely :P | 04:30 |
| teward | RFH is Request For Help so | 04:30 |
| arraybolt3[m] | (I'm good with C# and PowerShell (yeah, I used Windows when I learned those), semi-OK with Bash, newbie with Javascript.) | 04:30 |
| teward | pretty sure it's mostly Perl | 04:31 |
| teward | but i haven't checked myself | 04:31 |
| teward | *yawns* ok seriously, bedtime good night | 04:31 |
| arraybolt3[m] | OK, good night! | 04:31 |
| lubot | [telegram] <teward001> (also i just took a look at Debain tags, this triggers most often on binary files where you can't *control* line length so... depending on what source file is too long and what that file is, it's possibly "not fixable and totally ignorable" - shows up thousands of times without overrides in Debian it seems | 04:33 |
| lubot | [telegram] <teward001> *goes off to sleep*) | 04:33 |
| arraybolt3[m] | Simon Quigley (Developer): Finished LXQt Stage V! PR list is: | 05:07 |
| arraybolt3[m] | * https://github.com/lubuntu-team/lxqt-panel-packaging/pull/1 | 05:07 |
| arraybolt3[m] | * https://github.com/lubuntu-team/lxqt-powermanagement-packaging/pull/1 | 05:07 |
| arraybolt3[m] | * https://github.com/lubuntu-team/lxqt-runner-packaging/pull/1 | 05:07 |
| ubot93 | Pull 1 in lubuntu-team/lxqt-panel-packaging "Backports/jammy" [Open] | 05:07 |
| ubot93 | Pull 1 in lubuntu-team/lxqt-powermanagement-packaging "Backports/jammy" [Open] | 05:07 |
| ubot93 | Pull 1 in lubuntu-team/lxqt-runner-packaging "Backports/jammy" [Open] | 05:07 |
| * arraybolt3[m] needs to get some food, will be back shortly | 05:08 | |
| arraybolt3[m] | OK, I'm back. | 05:16 |
| arraybolt3[m] | Had a small but managed emergency, so I'm only just now going to bed. I intend to be here for the sprint at 10 A.M. today (so in about 3 hours from the time of this message). See you in a bit! | 11:52 |
| * arraybolt3[m] shuts down | 11:53 | |
| arraybolt3[m] | 🥱👋 | 14:58 |
| arraybolt3[m] | Just got Debian Salsa access this morning, so that should help with the sprint getting things upstreamed. | 14:59 |
| arraybolt3[m] | Good morning! | 14:59 |
| Rober[m] | arraybolt3[m]: Good morning :) | 15:01 |
| arraybolt3[m] | Simon Quigley (Developer): How'd the last bit of the backports go? Anything I need to fix? | 15:08 |
| arraybolt3[m] | libsysstat debian/experimental in Salsa was last updated 4 years ago (?!), should I sync everything with debian/sid first before getting started? Or is there something special I should do here? | 15:34 |
| tsimonq2 | <arraybolt3[m]> "Simon Quigley (Developer): How'd..." <- Will sprint a bit later | 15:47 |
| tsimonq2 | <arraybolt3[m]> "libsysstat debian/experimental..." <- Yes merge debian/sid over | 15:48 |
| lubot | [telegram] <Leokolb> 1 | 15:57 |
| arraybolt3[m] | @Leokolb 👋 | 15:58 |
| LeoK[m] | ready to run quick test on quassel | 16:01 |
| arraybolt3[m] | Leo K: That's what I'm doing rn too. | 16:01 |
| LeoK[m] | 1 | 16:02 |
| arraybolt3[m] | Eickmeyer: Whenever you get a chance (or whenever is a good time), can we bump Quassel through into Jammy? | 16:04 |
| arraybolt3[m] | (I guess I can test in Kinetic, actually...) | 16:04 |
| Eickmeyer[m] | arraybolt3: I've already pinged the necessary person, which is all I can do. | 16:04 |
| arraybolt3[m] | Ah, OK. (SRUs sound like they're annoying.) | 16:05 |
| Eickmeyer[m] | Lukasz is on SRU shift today. | 16:05 |
| Eickmeyer[m] | Luckily, it looks like 22.04.1 is being bumped by two weeks, so that buys a little extra time. | 16:06 |
| arraybolt3[m] | Simon Quigley (Developer): One (hopefully) last question - since I already did the work of backporting libsysstat from Kinetic to Jammy, the Jammy version is much more cleaned up than the Kinetic version, can I merge from it into Debian rather than from Kinetic into Debian? Or do I need to start with Kinetic, make it all nice, and then sync that into Debian? | 16:15 |
| lubot | [telegram] <Leokolb> Quassel good in Kinetic - still await update in Jammy | 16:27 |
| arraybolt3[m] | Simon Quigley (Developer): Just so I don't mess anything up, is it OK for me to directly submit PRs straight into Debian, or is that going to potentially cause problems? | 17:32 |
| Eickmeyer[m] | @Leokolb It's literally the same exact patch file, so I doubt it will be any different. Still nice to test once someone gives it an ack. Don't know what's taking so long, I guess Lukasz is busy. | 17:41 |
| arraybolt3[m] | Simon Quigley (Developer): Also, for whenever you see this, I'm stuck on qtxdg-tools, trying to get it into Debian. There isn't any source code for the repo in Debian, though there is an empty repo. Should I download a source tarbal for the latest version of qtxdg-tools, unpack it, git add, git commit, make a Debian directory, finagle the Ubuntu packaging into Debian packaging, then sbuild it with sid and if it comes back Lintian | 19:57 |
| arraybolt3[m] | clean, consider it done? | 19:57 |
| arraybolt3[m] | s/tarbal/tarball/ | 19:57 |
| tsimonq2 | <arraybolt3[m]> "Simon Quigley (Developer): Also,..." <- /me puts a Debian LXQt hat on | 20:09 |
| tsimonq2 | You want to use GBP for this | 20:10 |
| arraybolt3[m] | I want to? | 20:10 |
| arraybolt3[m] | OK. | 20:10 |
| * tsimonq2 puts a Lubuntu hat on | 20:10 | |
| arraybolt3[m] | STAY AWAY FROM GBP! | 20:10 |
| tsimonq2 | GBP is complicated and has too much overhead never use it | 20:10 |
| arraybolt3[m] | Did I get it right? | 20:10 |
| tsimonq2 | See the differences in hats? ;) | 20:10 |
| arraybolt3[m] | Yep. | 20:10 |
| tsimonq2 | <arraybolt3[m]> "Simon Quigley (Developer..." <- Go for it. Add me as a reviewer so you have a 0% chance of grumpiness | 20:11 |
| arraybolt3[m] | I like it. Should I try to learn GBP? | 20:11 |
| * tsimonq2 puts Debian LXQt hat back on | 20:12 | |
| tsimonq2 | Yeah, you're gonna need it | 20:12 |
| tsimonq2 | Especially working with packaging in Salsa | 20:12 |
| * tsimonq2 puts Lubuntu hat back on | 20:12 | |
| arraybolt3[m] | OK. Sad, but OK. | 20:12 |
| tsimonq2 | You'll only need it for Debian :P | 20:12 |
| tsimonq2 | The reason I'm so against it is not because I hate it as a tool. It can be a useful workflow | 20:13 |
| tsimonq2 | I just think it's not very digitally economical to store the entire upstream Git history in the packaging repo :P | 20:13 |
| tsimonq2 | Or even the upstream changes for that matter | 20:13 |
| arraybolt3[m] | Ah, that makes sense to me. | 20:13 |
| tsimonq2 | Especially as we scale, you'll notice the difference between GBP and not when it comes to cloning new repositories | 20:14 |
| tsimonq2 | GBP repos take waaaaaaay longer especially if the source tar is big | 20:14 |
| tsimonq2 | Even on a fast internet connection, GBP repos can scale up to be a few gigs | 20:14 |
| arraybolt3[m] | Oh dear. That may be a problem for me on my not fast Internet. | 20:15 |
| arraybolt3[m] | But hopefully it shouldn't be a problem here. | 20:15 |
| tsimonq2 | LXQt is pretty light anyway | 20:16 |
| tsimonq2 | Especially if you're only updating individual components at a time | 20:16 |
| === NotEickmeyer is now known as Eickmeyer | ||
| arraybolt3[m] | Good grief, just trying to wrap my mind around gbp is making my head spin... | 20:53 |
| arraybolt3[m] | (Which one is actually the master branch? master? pristine-tar? upstream? And how exactly do you import an external Git repo?) | 20:54 |
| tsimonq2 | No idea how you import an external repo | 20:57 |
| tsimonq2 | debian/sid or similar should be packaging | 20:57 |
| arraybolt3[m] | You just use the tarball method? | 20:57 |
| arraybolt3[m] | I'm working off this wiki: https://wiki.debian.org/PackagingWithGit | 20:57 |
| tsimonq2 | pristine-tar / upstream is just the source code | 20:57 |
| tsimonq2 | arraybolt3[m]: Yeah, can depend on the repo tho | 20:57 |
| arraybolt3[m] | If we're using the upstream repo, we can git fetch the new release tag into the repo, in the upstream branch. We need to git merge this branch into master, and to commit the new tarball with pristine-tar. The patches can be rebased with gbp-pq, and git itself can be used to resolve any conflicts that have been created. | 20:58 |
| arraybolt3[m] | Help?! | 20:58 |
| arraybolt3[m] | They lost me at the second sentence. | 20:58 |
| arraybolt3[m] | * If we're using the upstream repo, we can git fetch the new release tag into the repo, in the upstream branch. We need to git merge this branch into master, and to commit the new tarball with pristine-tar. The patches can be rebased with gbp-pq, and git itself can be used to resolve any conflicts that have been created. | 21:01 |
| tsimonq2 | ok so you have two remotes | 21:03 |
| tsimonq2 | Salsa and upstream Git | 21:03 |
| tsimonq2 | upstream tracks, well, upstream Git | 21:04 |
| tsimonq2 | And that's also pushed to Salsa | 21:04 |
| tsimonq2 | (See the redundancy?) | 21:04 |
| arraybolt3[m] | Yeah. | 21:04 |
| arraybolt3[m] | But it's pristine-tar that's blowing my mind up (well, that and the advanced git-fu the manual uses). | 21:05 |
| tsimonq2 | arraybolt3[m]: I *think* that may have to do with DFSG stripping? | 21:06 |
| arraybolt3[m] | Well, they say to experiment around, so I guess I'll try it. Worst case scenario you can tell me what to do if I'm really lost. | 21:07 |
| arraybolt3[m] | Simon Quigley (Developer): OK, so I think what I'm supposed to do is make a branch called "upstream", then "git pull" upstream's code into that branch. So I did "git init", "git checkout -b upstream", "git remote add origin <upstream URL>.git", "git pull origin 3.9.1". I now get the following mess on my screen: | 21:34 |
| arraybolt3[m] | https://pastebin.com/83NaJPuX | 21:34 |
| arraybolt3[m] | I tried using special options in git remote, which failed in other varying and interesting ways. | 21:34 |
| arraybolt3[m] | Am I doing this all wrong? | 21:34 |
| arraybolt3[m] | (It's almost like the Git repo being empty is making Git grumpy.) | 21:42 |
| arraybolt3[m] | Oh great. Looks like I found it. Classic case of XY problem. | 21:44 |
| tsimonq2 | Got it? | 22:25 |
| arraybolt3[m] | Ish...? Not sure yet, we'll find out! | 22:27 |
| tsimonq2 | Cool :) | 22:28 |
| arraybolt3[m] | I'm trying to figure out how to create the Git structure the way GBP wants - it's brain-bending, but it looks like it should hopefully work. | 22:29 |
| tsimonq2 | Cool | 22:31 |
| arraybolt3[m] | Simon Quigley (Developer): OK, am I supposed to have all of upstream's commit history in the upstream branch? Or is it supposed to be squashed? | 22:36 |
| arraybolt3[m] | Man, this is crazy. I think I'm supposed to have a "master" branch at some point in all this craziness, but right now I'm doing the initial commit AND I think it only goes in debian/experimental. How does that work? | 22:55 |
| tsimonq2 | arraybolt3[m]: Rename branches around | 22:56 |
| tsimonq2 | And debian/gbp.conf | 22:57 |
| tsimonq2 | Check existing LXQt packages | 22:57 |
| arraybolt3[m] | Oh OK, that makes sense for getting the conf file. | 22:57 |
| tsimonq2 | tsimonq2: I'm willing to bet copying this from an existing LXQt package will solve everything | 22:57 |
| arraybolt3[m] | I'm still trying to wrangle my Git-fu into some semblance of usability with this thing. | 22:58 |
| tsimonq2 | Carry on. Let me know if you need me to toss you a beach floatie :) | 22:58 |
| tsimonq2 | oh and watch out for tewardshark in the water, he likes legs | 22:58 |
| tsimonq2 | XD | 22:59 |
| arraybolt3[m] | OK, thanks. Hopefully I'll have figured out enough that by the time we're all sprinting I'm sprinting rather than sloughing through quicksand. | 22:59 |
| tsimonq2 | Believe me... we'll be in full swing by very early tomorrow morning :) | 22:59 |
| * arraybolt3[m] throws EMI at teward to throw off his sense of where I am | 23:00 | |
| tsimonq2 | arraybolt3 @arraybolt3:matrix.org: Try to see if you can create your own account on notes.lubuntu.me or if teward can give you access. Worst case scenario find a generic CodiMD (preferred) or Etherpad instance and paste your existing PRs there | 23:00 |
| arraybolt3[m] | Oh, nice! Do I just submit PRs straight into Salsa like I would with GitHub? | 23:01 |
| tsimonq2 | That way I can just bookmark your page and go back to it | 23:01 |
| tsimonq2 | arraybolt3[m]: Yes | 23:01 |
| tsimonq2 | So you fork it and then push to your fork | 23:01 |
| tsimonq2 | At that point you submit a Merge Request | 23:01 |
| arraybolt3[m] | Nice. | 23:01 |
| tsimonq2 | Merge Request = Pull Request | 23:01 |
| arraybolt3[m] | notes.lubuntu.me has no create account button. | 23:02 |
| teward | tewardshark was replaced with mechanical sharks lol. i'll create a codimd account after dinner. hunger is more important lol | 23:02 |
| arraybolt3[m] | Just a sign in. | 23:02 |
| arraybolt3[m] | teward: Thanks! | 23:02 |
| arraybolt3[m] | Yeah, definitely get food first. I'm in no hurry. | 23:02 |
| tsimonq2 | I've been running on very little sleep and very much caffeine the last several days | 23:03 |
| arraybolt3[m] | I got 2 1/2 hours last night, so I feel you. | 23:03 |
| tsimonq2 | That will likely continue as the weekend goes on | 23:03 |
| tsimonq2 | Luckily last night I got away with like 4.5... but the night before was 1 hour in one nap and 2 hours in another | 23:04 |
| arraybolt3[m] | What's with us and almost no sleep? It's almost like we like it or something. | 23:04 |
| tsimonq2 | <arraybolt3[m]> "What's with us and almost no..." <- Excuse my talk to texans but I have a habit of taking on several large projects at once for several unrelated parts of life so I'm always busy but I like keeping busy | 23:12 |
| arraybolt3[m] | Simon Quigley (Developer): OK, so about that pool noodle. I got all my branches set up I think (debian/sid, upstream/latest, upstream/3.9.1), got my gbp.conf, tried to build the package, and NOW I'm told that I'm missing pristine-tar. Looking at pristine-tar in other repos, it looks like it's a parallel universe in the same Git repo. How do I add it in this late in the game? In fact, how do I add it in at all? | 23:12 |
| tsimonq2 | I'd like to bring to a council vote that we officially call all maintenance scripts tewardshark XD | 23:13 |
| arraybolt3[m] | tsimonq2: Me thinks you'll get at least one council vote against you. | 23:14 |
| tsimonq2 | arraybolt3[m]: Honestly on this one I couldn't tell you off the top of my head | 23:14 |
| tsimonq2 | arraybolt3[m]: LMAO | 23:14 |
| arraybolt3[m] | Simon Quigley (Developer): Well, do we know how to make it in the first place, even if I have to start from square 1? | 23:14 |
| arraybolt3[m] | (I might be able to make a new branch, hard-reset it to the very beginning so the folder looks blank in that branch, then add in the stuff, but I have no clue how pristine-tar even works, so even if that's the right approach...) | 23:15 |
| tsimonq2 | arraybolt3[m]: Again I probably knew at 1 point off the top of my head but that time is not now | 23:15 |
| arraybolt3[m] | Ah. OK. | 23:16 |
| arraybolt3[m] | Aha, there's actually a gbp command that does this specific thing. So it's not my git-fu that was too horrible, it's my gbp-fu now. | 23:17 |
| * arraybolt3[m] cries in Debian | 23:17 | |
| arraybolt3[m] | (That's a joke, I'm loving this job.) | 23:17 |
| kc2bez[m] | arraybolt3[m]: It pays really well. | 23:18 |
| arraybolt3[m] | kc2bez[m]: Actually, it really does. It pays with a free, lightweight, powerful operating system. | 23:18 |
| guiverc | :) | 23:19 |
| kc2bez[m] | Don't forget the knowledge and networking. | 23:19 |
| arraybolt3[m] | And it pays with a fantastic experience learning skills and making friends. This is the first project of this scope and scale I've ever taken on, and it's been amazing from square 1. | 23:19 |
| arraybolt3[m] | guiverc: (which, by the way, you pretty much were square 1, so thank you for giving me an awesome entry to an awesome project!) | 23:20 |
| guiverc | You're most welcome & it's great to see you having fun, and contributing so much ! | 23:22 |
| arraybolt3[m] | WOOT! First gbp run that DIDN'T instantly do something wrong! (First it failed, then it failed, then it failed, then it tried to build with Ubuntu's packages (?!), now it's working with Sid as it should. I bet I forgot to specify the build dep resolver and extra repo. Great. So much for that.) | 23:30 |
| arraybolt3[m] | So now you know what it looks like when my brain goes from SUCCESS! mode to "Oh... 🤦♂️ | 23:35 |
| arraybolt3[m] | while I'm still writing. | 23:35 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
