[00:28] Simon Quigley (Developer): So, after grabbing Ubuntu's symbols and plopping them into Debian's package (for that one part of libqtxdg), I decided to build the Debian package in a PPA just in case (did a quick hack to make it look like a Kinetic package), then looked at the build logs, and GUESS WHAT... s390x has one extra symbol that the other architectures don't have! Should I add that to the Debian symbols? [00:41] "Simon Quigley (Developer): So..." <- What's the symbol? [00:42] Simon Quigley (Developer): (arch=s390x)_ZNSt6vectorISt10unique_ptrI14XdgDesktopFileSt14default_deleteIS1_EESaIS4_EE17_M_realloc_insertIJS4_EEEvN9__gnu_cxx17__normal_iteratorIPS4_S6_EEDpOT_@Base 3.9.1 [00:42] I'll c++filt it real quick... [00:42] Wow, that's not much better. [00:43] Simon Quigley (Developer): void std::vector >, std::allocator > > >::_M_realloc_insert > >(__gnu_cxx::__normal_iterator >*, [00:43] std::vector >, std::allocator > > > >, std::unique_ptr >&&) [01:05] Don't even worry about that one them [01:05] s/them/then/ [01:05] Seems pretty ephemeral tbh [01:06] OK. So it's normal for it to not exist? [01:06] yes [01:07] Weird. Is there some way I can tell that? (Just wanting to learn.) [01:08] Bunch of <> [01:08] Also, how ethical is it to use Canonical's servers for building Debian packages for testing? [01:08] shrug [01:08] (I'm worried maybe I did it wrong by hacking a Debian package into an Ubuntu lookalike.) [01:08] arraybolt3[m]: I mean, it's not possible :P [01:08] (Sure worked well, though.) [01:09] tsimonq2: That's what I did though. I took the Debian package, then for testing in a PPA, I added a changelog entry with a Ubuntu version number, then uploaded it to a PPA and lo and behold, success. [01:09] That's how I found that one weird s390x symbol. [01:10] Simon Quigley (Developer): https://launchpad.net/~arraybolt3/+archive/ubuntu/rnd/+packages (Look at the package that ends in ~ppahack1, that's actually a Debian package disguised as an Ubuntu package.) [01:11] But OK, the symbol can be left out. 👍️ [01:51] "That's what I did though. I took..." <- That isn't building against Debian though :) [01:52] I mean, yes, it is the Debian packaging [01:52] But if you check the build logs it's pulling Ubuntu sources :) [01:52] Anyway [01:53] Make any progress or do you need me to jump in? [01:53] Making progress, but my personal life has thrown a serious roadblock for a few minutes (hopefully not more than an hour), so bear with me. [01:54] Go for it man, life is what it is, let me know how I can help :) [01:55] How exactly would I build against Debian for other architectures other than amd64? I just want to make sure the symbols are working right after I fiddled with the Debian symbols. [01:55] (Hopefully would like a solution OTHER than an emulator.) [01:55] arraybolt3[m]: debci :) [01:55] Is that something I can access, or do I give you the files and ask for help? [01:56] Hmm, I don't know [01:56] I'd say try with a Debian schroot and hope for the best [01:56] It will let me log in with Debian Salsa (I think), but it might take 4 days before my Salsa account gets approved. [02:37] OK, feel free to just send me the changes elsewhere either in like a text file or by a get hove or something like that [02:37] Please excuse my talk to text I'm driving [02:38] OK. Sounds like a good idea. And thank you for not texting and driving at the same time. I can translate Google's (or Apple's) voice-to-text silliness. [02:39] (You should see what YouTube's voice-to-text auto subtitler does. "Holy moo" was one of the best.) [02:54] PMAO [02:54] * LMAO [02:54] The goal here is to be able to sync the version from experimental and drop all Ubuntu changes, if we can [02:55] Now, there are some areas where Ubuntu did it right and Debian didn't (Ubuntu using Standards-Version 4.6.1 whereas Debian was using 4.6.0), so I'm taking those changes from Ubuntu to Debian. So in the end, they'll be the same, but which one gets changed depends on who's right. Is that right? [02:56] (Also I just copypasta'd the symbols for libqt5iconloadersomethingorother from Ubuntu into Debian, which is why I'm anxious to test it in a PPA-like build, since you said you'd use Ubuntu's symbols.) [02:56] I guess so. [02:57] I've not PR'd or anything yet, all my changes are local, so you'll be able to see what I've done and accept/clean up/lecture me over everything before anyone else sees it. [02:57] If there is anything that should not be upstreamed, please tell me [02:57] * be upstreamed to Debian, please [02:58] Just wait until we get CI back up and running, you'll see why I want it back so bad :) [02:59] OK. I'll make a list of "things I think had better stay in Ubuntu" (it's mainly stuff like the XS-Debian-Vcs-Git field in debian/control, stuff where Ubuntu and Debian are actually different and I believe they should stay that way). [02:59] (You'll be able to push one commit and it will build on all arches + give you symbols + give you the Lintian output) [02:59] arraybolt3[m]: If that's the only change, we can drop it [03:00] * If that's the only part of the Ubuntu delta, we can drop it [03:00] Well, yes but we can't put "Vcs-Git: https://phab.lubuntu.me/source/libqtxdg.git" in Debian. (Right?) [03:00] Same with the changelog [03:00] arraybolt3[m]: True. Same thing, use Debian's and we can just sync it [03:01] Oh, OK. I guess to begin with this is more of a packaging cleanup, then I'll do the sync once everything's in a more sane state. [03:12] Simon Quigley (Developer): What about the Lubuntu Developer's copyright? I'm actually making changes to Debian's debian directory (pun not intended), so should I upstream the Lubuntu Developers copyright line? [03:13] (Also unsure of what to do with gbp.conf in Debian.) [03:25] Also, I didn't realize how much of Debian's changes I could sync into Ubuntu, so I probably have a lot of changelog entries and git commits that don't actually make sense anymore, so if I need to try again after this initial try, please just tell me, "Dude... you... what? OK, hold on. Start over from scratch and let's try again." [03:28] "Simon Quigley (Developer..." <- Nah it's all good [03:28] "(Also unsure of what to do..." <- Keep it in Debian [03:28] arraybolt3[m]: Yeah there you go :) just keep on chugging [03:31] I'm at home now btw if you wanna be like "hey am I on the right track?" [03:32] Simon Quigley (Developer): Should I wipe the original maintainer and copyright fields out of Ubuntu and just clobber it with Debian? Or should I preserve that stuff in Ubuntu? [03:32] (Talking about fields in debian/control) [03:32] And debian/copyright [03:33] Well, if we're keeping an Ubuntu delta and merging, then keep those. If we're going to upstream everything and sync it from Debian, nah, throw it away [03:33] Different repositories require slightly different goggles, so to speak [03:33] Versioning and maintainership are the big ones [03:34] Debian takes an approach of "everything goes in one main pool but I maintain my own special package" [03:34] Ubuntu takes the approach of "okay everyone in this particular group maintains these particular packages but we're only going to let some people have access to some things" [03:34] With Debian, once you're a DD it's no longer a matter of "access" it's a matter of "courtesy" [03:35] LOL man, that sounds chaotic. [03:35] Believe me, it definitely is [03:36] (Like having to work with upstream on every single thing, whereas in Ubuntu you only have to work with upstream every once in a while. Fewer words strike more "uh oh" into me in the Ubuntu world than "upstream bug report".) [03:36] Yeah, it's quite a bit different [03:37] Once you become a Lubuntu Developer, if you can get your GPG key signed by a DD, I'll sponsor you for Debian Maintainer [03:37] Which reminds me... [03:37] @teward001: As we're going through this process, lmk if you want upload access to LXQt. [03:37] (As a DD I can grant such permissions. :)) [03:39] tsimonq2: LOL "once I become a Lubuntu Developer", I'm definitely aiming there, but I'm still nervous about the Lubuntu Member interview... [03:40] Really thankful you all are here to walk me through. [03:40] guiverc: G'day! [03:41] "Well, if we're keeping an Ubuntu..." <- Uhh... which one should I be doing? I'm trying to **merge** right now. [03:45] arraybolt3[m]: The goal should be, first, to try to evaluate the diff to see if merging or syncing is the best option. [03:45] Are all of the changes applicable to Debian? [03:46] Here, let me show you exactly the diff I would be looking at... [03:48] $ pull-lp-source libqtxdg && pull-debian-source libqtxdg experimental [03:48] [...] [03:48] $ debdiff libqtxdg_3.9.1-1.dsc libqtxdg_3.9.1-0ubuntu1.dsc > delta.debdiff [03:48] * tsimonq2 posted a file: delta.debdiff (28KiB) < https://libera.ems.host/_matrix/media/r0/download/linuxdelta.com/xvpoyovJDQTBgkZEGCABCthC > [03:48] (Can you see that file I just uploaded?) [03:49] Yeah, looks a lot like what I've got (only I "diff -r"'d two git clone'd directories). [03:50] I think I should probably start over. I didn't even know about pull-debian-source until right this moment. A lot of my cleanup work was unnecessary, and looking at it now, I think a sync would be best, not a merge. [03:51] Here's what I'd upstream to Debian, specifically:... (full message at https://libera.ems.host/_matrix/media/r0/download/libera.chat/dee17f89ac27d971ab1dbd541c09f8cdf2785234) [03:51] Then do a `debuild -S -d -sa` [03:51] Oy, I downstreamed copyright. I'd definitely better start over. [03:52] Hmm, well maybe that's part upstream part downstream. [03:52] Like this should be kept: [03:52] ``` [03:52] - 2015-2022, Andrew Lee (李健秋) [03:52] ``` [03:52] (I saw more recent dates in Debian's copyright and was like, "OK, must be newer, they know what they're doing, yay!") [03:52] Yes and no :) [03:53] Always double-check if you're unsure [03:53] And be more unsure... [03:53] But yes, I don't think we have anything downstream that we can't upstream here. [03:53] This entire delta looks upstreamable. [03:53] (Obviously don't upstream changelog entries or maintainership...) [03:53] Agreed. I'm going to wipe everything and start over. [03:53] tsimonq2: But do downstream the whole changelog, right? [03:54] Cool. Take a crack at it and throw me a debdiff from the original Debian version to your Debian version when you're done. :) [03:54] arraybolt3[m]: Yeah, that changelog should be merged in with downstream *if* we plan on maintaining a delta. [03:54] In this case I don't think we will. [03:55] 👍️ OK, well, that was a lot of wasted work, BUT, I learned a lot. Thanks for teaching me even though it's slower than just doing it at this point. [03:56] Of course. I apologize if this has been frustrating for you at any point. You're going to experience that here and there, but the benefit of working with a team of credentialed people is that you're going to get good results eventually and learn a lot :) [03:56] Not frustrated, I'm having a lot of fun! I'm scared of frustrating you. [03:57] (You'll be able to learn something from anyone between Dan, who is simply an LD, to me who is an Ubuntu Core Dev and Debian Developer with about a medium experience level, to vorlon who is an absolute wizard and has been in the project longer than we've both been alive (ping him sparingly :)) [03:57] arraybolt3[m]: Nah man you're totally cool [03:57] I don't want you to get frustrated and give up, especially since you're on a roll :) [03:57] I have patience all day [03:58] Especially for someone who has proven themselves to be adaptable and quick learning ;) [04:11] Eickmeyer: And, in turn, a present for you: https://pad.lv/1980180 [04:11] Launchpad bug 1980180 in calamares (Ubuntu) "[SRU] Calamares 3.2 Series for 22.04 LTS" [Undecided, New] [04:16] Wow, this is going WAY faster now that I know more of what I'm doing! [04:21] OK, possibly silly, but I'm thinking of skipping the watch file upstreaming, since the difference in the watch file between Debian and Ubuntu is four spaces of indentation, and that's it. Is that reasonable, or should I go ahead and try to upstream the indentation too? [04:22] Meh, upstream the indentation too [04:22] OK. [04:30] Now, the copyright file in Debian *looks* pristine to me. I want to double-check it, but you kinda got me spoiled on GitHub's "compare" feature, so now I'm hoping Debian's GitLab will have the same thing. Any chance? [04:30] (That "compare" feature was SO HANDY for doing copyright updates, btw.) [04:30] Oh my gosh, I can probably get git to do it locally. [04:30] Hold on, nevermind. [04:35] Hey, sorry, just got liblxqt into Debian Experimental. [04:35] Sup? [04:35] Just asking questions and then having the answer hit me. (Just learned about git diff) [04:36] Cool :) [04:38] https://launchpad.net/ubuntu/+source/libqtxdg/3.9.1-1 [04:38] I'm going to sync from Experimental as I upload them unless I see a meaningful delta [04:39] So at this point you're working on debian/experimental branch upstream - making 3.9.1-2 [04:39] I'm almost done doing it, I'm just checking over the copyright file. [04:39] Cool, cool. [04:42] arraybolt3: FWIW, contributions upstream to Debian **do** count towards both Lubuntu Member and Lubuntu Developer imo, btw. [04:43] OK, that's fun! [04:43] (If you aren't already, start making a list of links to uploaded packages and generally your contributions and links... will be helpful soon.) [04:43] (I may have decided to diff a bit too far back...) [04:43] (Let me be specific, contributions to Debian LXQt...) [04:43] arraybolt3[m]: Oh yeah? :) [04:43] OK. I'm not keeping that close track of things, maybe I should. [04:43] (I've got like over 20 removed files and counting trying to check the copyright file...) [04:43] All good. I know you'll get it eventually :) [04:44] Remember to strip the upstream source code unless you're working directly from Debian git. [04:45] Oh... no. I've been working in the Debian git tree and just making modifications to the debian directory and committing happily, with all the source code still there... [04:45] Was I supposed to delete all the source code and make a prep commit first? [04:45] Meh, nah, you're good [04:45] Forget about it [04:45] In fact, I'd prefer it that way [04:46] Whew, ok. [04:46] If we're working with Debian git, unless we're maintaining it by ourselves, we should just keep with gbp like they've done it in the past [04:46] You can also just push that repo to GitHub and I can manually add it as remote locally [04:46] Like, your changes and stuff. [04:46] That's what I figured I'd do, since I don't have Debian Salsa access yet. [04:47] If I need to endorse you in any way, please do let me know. [04:48] OK. I think they just want to make sure I'm not a spammer, but if four days passes and I still don't have access, and the support gives me any resistance, I'll let you know. [04:48] Cool. [04:48] Let me know if you need any more help, otherwise I'll be looking forward to your repo! [04:48] I think I'm good! If I ever get to the end of this diff laughs with chagrin [04:48] bahahaha [04:52] Holy smoke, well that explains it. I've been keeping track of every removed file from src/tools/mat, when src/tools no longer exists in the Debian repo... [04:54] Oh, and you know what's funnier? My search kept wrapping and I didn't notice. [04:55] I have to show you a pastebin of what I just spend fifteen minutes doing... 🤦‍♂️ [04:55] https://pastebin.com/gXWW1tbi [05:00] Niceeee lmao [05:01] Simon Quigley (Developer): Alright, skim over this and tell me what you think. [05:01] https://github.com/ArrayBolt3/libqtxdg-debian [05:04] LGTM, one last request... [05:04] Is the Lintian output clean? [05:04] No clue, forgot to sbuild! One moment (hopefully) please... [05:05] Cool. Debian Sid sbuild preferrably. [05:05] 👍️ [05:05] I'd also enable experimental in your Sid schroot [05:06] This will be useful for later. [05:06] OK, this will actually take a bit, since I'm making a new schroot (I have Jammy and Kinetic, but not Sid yet). [05:06] Cool. [05:07] tsimonq2: And set aptitude as your build dep resolver [05:07] Uh, ok, can you tell me how to do those last two things? [05:08] (Never used aptitude before, also don't know how to enable experimental.) [05:08] (I'm guessing the latter is a matter of changing /etc/apt/sources.list?) [05:10] I can just look it up if you're busy (I think). [05:11] Yep, found it. [05:21] OK, looks like there's a couple of boffos I need to fix, but then it should be good. [05:34] Cool [05:35] Simon Quigley (Developer): I just got barked at by Lintian over my work being a "non-maintainer upload", and I think it wants me to use a version ending in -1.1, rather than -2. Is that right? [06:01] "Simon Quigley (Developer): I..." <- Nope. Proceed with -2. [06:24] Oh, OK. [06:25] Final sbuild, if this checks clean I'll git push one more time and we'll be done! [06:32] Sweeet [06:32] Only thing I see that may be worrisome is "E: libqtxdg changes: distribution-and-experimental-mismatch". I think that might just be because of my sbuild line (I ended it with -d sid-amd64). [06:33] Doesn't matter [06:35] Nice. Git push incoming. [06:36] Simon Quigley (Developer): https://github.com/ArrayBolt3/libqtxdg-debian [06:37] Revert 9d53688ef38ab3f6f5b9dfce3f4fff8955092a70 please [06:37] Otherwise you're good [06:38] Can I just make a new commit over the top, or do I need to reset to before then, and redo everything after? [06:40] `git revert COMMITID` [06:40] so new commit [06:40] Learn new commands every day! [06:40] :D [06:41] Done. [06:42] Out of curiosity, why was it supposed to not specify non-maintainer upload? Because I'm not uploading it? [06:48] Nah, cause I'm part of the Debian LXQt team and I'm your sponsor [06:48] Also, arraybolt3 - congratulations on your first upload to the Debian archive. :) [06:48] Thank you! [06:49] Whew, that was hard, and really fun. Next time around should hopefully be way easier. [06:49] Of course :) [06:49] Of course it'll be. [06:49] (Guess I didn't really know what "non-maintainer upload" was for, I just knew Lintian wanted me to put it there.) [06:50] Now, what I'd do from here is update the ubuntu/kinetic packaging to sync it with the archive (-1). Then apply -2 as a separate commit. Then, merge it into backports and prepare a -2~ppa1 upload. [06:50] arraybolt3[m]: All good. [06:50] * tsimonq2 is wondering if I can get liblxqt to upload to experimental... [06:51] See, Debian is weird... [06:51] If there are any binary NEW packages (new binary packages introduced via d/control), it makes you run a build and upload the amd64 deb file with the source [06:52] And then you have to upload again, because, well, Debian won't accept packages with locally built debs into testing [06:52] First time I just uploaded the source... wrong [06:52] Ohhhkayyyy, that's odd. Wonder why they do that. [06:52] Second time I uploaded the amd64 changes file that included the source, forgot to change the distribution from sid-amd64-shm [06:52] So hopefully this upload works. [06:53] (I just had to manually change it in amd64.changes and re-debsign + re-upload) [06:54] Also, all of the Debian scripts are via cron [06:54] It's not like Launchpad where uploads are processed instantly... [06:54] First, it actually has to upload to the build server [06:54] And then dinstall runs up to 6 hours later which actually installs those sources and debs into the archive [06:55] Ubuntu archive = 1 hour minimum for packages to land, not 6 [06:55] Odd. To me that just sounds like it would slam the build servers every so often, rather than letting them constantly chug away at a comfortable rate. Maybe it deters people from doing things that aren't well thought out? [06:56] Hold on, I just realized something I have to go fix (not in Ubuntu or Debian, personal life). BRB [07:02] OK, back. [07:02] (Forgot to let dog in earlier.) [07:02] "Odd. To me that just sounds like..." <- Perhaps [07:03] "Now, what I'd do from here is..." <- This is two uploads, right? One for applying -2 to ubuntu/kinetic, one for applying -2 to backports/jammy? [07:03] arraybolt3[m]: Correct [07:03] -2 isn't uploaded [07:03] I use a tool to sync it via Launchpad [07:03] Only -2~ppa1 is uploaded, to the PPA [07:04] You're just getting a base in ubuntu/kinetic [07:04] If your git-foo is strong you may consider cherry-picking commits individually that don't touch the source to save yourself some work. [07:04] (That works both ways, btw.) [07:05] Let's say you have commit abcd on debian/experimental, you can checkout ubuntu/kinetic and `git cherry-pick abcd` to just pull it over [07:05] I've used git cherry-pick before. [07:05] Sweet :) [07:05] I get it, that makes sense. [07:06] So, just to make sure I'm understanding, -2 is NOT going into kinetic? Is that something to be handled later, or is it just, "we're not doing that"? [07:06] You can also cherry-pick from ubuntu/kinetic over to debian/experimental :) [07:06] (By "going into kinetic", I mean into the release.) [07:06] arraybolt3[m]: -2 as-is from Debian automatically imports to Ubuntu [07:06] We don't manually upload it [07:06] Ah. click [07:08] So, sync with Debian by copy-pasting the debian directory I think. Then cherry-pick all my changes so that it updates to -2. Then change it to a ~ppa1, sbuild, send. Right? [07:09] So the goal here is to take all the packaging from Kinetic to Experimental, get it syncable or at least understand the delta, upload the delta to Kinetic if there is one and the original to Experimental, backporting Kinetic (perhaps inherently Experimental) to Backports [07:09] arraybolt3[m]: I wouldn't even do that first step [07:09] Just cherry-pick all your commits over [07:10] tsimonq2: I lost you at "and the origianl to Experimental". [07:11] s/origianl/original/ [07:13] Okay so you want to make the majority of changes in Experimentsl [07:14] Yes. [07:14] That may involve cherry-picking commits from Kinetic or Backports [07:14] So that is your "original" [07:14] Kinetic has a delta, if a delta is needed, or just straight inherited Experimental [07:14] OK. [07:14] Backports has whatever Kinetic has [07:14] Does that make sense now? [07:15] Yeah. [07:15] Before we just made whatever changes in Kinetic [07:15] But now we're basically Debian too lol [07:15] So I already made all the changes in Experimental. [07:15] So now I need to propagate those to Backports. [07:15] Right, and I don't think we need an Ubuntu delta on this one [07:15] Yes [07:15] But... [07:15] And Kinetic has no delta. [07:16] It shouldn't go Experimental -> Backports [07:16] It should always go Experimental -> Kinetic -> Backports [07:16] So Kinetic is never forgotten [07:16] Right. [07:16] That makes it very easy to add a delta if we ever need to [07:17] One thing I'm missing is how cherry-pick makes this easier. I already have everything in Experimental, seems like it would be easiest to "cp -R", "git add -A", "git commit", boom. [07:17] Are we cherry-picking since that would squash the commit history? [07:17] Hmm... [07:18] I think I was meaning it as a general concept for future reference in hindsight [07:18] (That being the cp, git add, git commit.) [07:18] In this case the appropriate tool is likely git merge [07:18] And favor all changes in debian/experimental [07:18] me man git-merge [07:18] * arraybolt3[m] * man git-merge [07:20] OK, there's some stuff that's still a bit unclear in my mind, but an error message or two will probably kick me into shape. I'm gonna try and see if I can get this right. [07:21] Simon Quigley (Developer): One thing I'm not sure about is, the Ubuntu packaging and Debian packaging are in entirely different Git repos. How am I supposed to git merge those? [07:22] arraybolt3[m]: Just add different remotes and fetch them [07:22] Oh, that's why I said cherry pick [07:22] We don't want gbp stuff [07:23] aHA! OK, that makes sense. [07:24] How does git cherry-pick handle significantly different files (like the files in Debian and Ubuntu)? [07:24] Will it replace the file entirely? [07:24] (That's what I'm hoping.) [07:24] Nah just apply the diffs [07:25] OK, lemme try. [07:30] Whew, that was a messy merge. Not sure if I did that right, but I ended up with only just the debian folder from debian/experimental, which I think is right? [07:31] Welp. Now I know how not to merge stuff in Git. [07:36] OK, my git-fu is failing me. [07:36] "Whew, that was a messy merge..." <- On ubuntu/kinetic, yes that's right [07:36] arraybolt3[m]: Sup? [07:36] How do I get the debian/experimental branch as a branch of libqtxdg-packaging so I can start cherry-picking? [07:37] git remote add salsa.git [07:37] I tried "git config pull.rebase false" then "git pull", and it went POORLY. [07:37] er sorry [07:37] Then I tried "git config pull.rebase true" and I couldn't figure out what I was doing. [07:38] Somehow I ended up with... corrupted? versions of the Kinetic files, and never could get the Debian files to show up properly (unless they showed up during the rebase catastrophe, which I doubt). [07:38] git remote add debian salsa.git... (full message at https://libera.ems.host/_matrix/media/r0/download/libera.chat/1995fcd2e94216be2ca68ff6d79e03b53c405851) [07:38] From a clean Kinetic clone [07:38] * arraybolt3[m] man git-fetch [07:38] That should get you what you need [07:39] Oh, so I'm using the totally wrong command. That explains the tantrum it threw. [07:41] Hey, that did it! \o/ [07:42] Simon Quigley (Developer): So now I go through the history, find the most recent time a file was modified, and then cherry-pick that, doing the same for every file in the package? [07:50] Simon Quigley (Developer): So, what I'm missing is, I'm supposed to apply the diffs, but a simply "git cherry-pick" doesn't do that - if I try to git cherry-pick, say, "Revert "Specify non-maintainer upload"", I get an instant conflict in Git. I think this is obviously to be expected, but where am I missing something? [07:50] Start with the oldest commit, and start with the one after I merged from upstream/latest [07:51] Oh, from "New upstream release"? [07:52] Nope. That can't be right. [07:53] Ah, start with the oldest (Debian) commit, and start with the one after I merged from upstream/latest (in Ubuntu). [07:54] The merge from upstream/latest should be entirely in the Debian packaging [07:54] You want to cherry-pick all packaging changes that happen after that point [07:55] OK, trying... [07:55] (I will figure this out eventually, thank you for your patience!) [07:55] And actually, in that process you can just lose all the Ubuntu changelog entries. You want a `git diff debian/experimental ubuntu/kinetic` to come up empty. That's your measure of success here. [07:56] OK. But... something's going awry here. [07:56] Yeah? [07:56] I've tried "New upstream release", I've tried "Add back symbols file", and I've tried "Update install file", all have come back with a merge conflict. [07:56] s/merge// [07:57] Okay, so resolve the merge conflict. :) [07:57] Pop open the file in the text editor [07:57] 🤦‍♂️ [07:57] Evaluate the scaffolding [07:57] right. [07:57] (Why didn't I think of that?) [07:58] Pick your changes, remove the scaffolding, git add, repeat until all conflicts have been solved, git commit when you're done [07:58] git status if you need a reminder on which files still have conflicts [07:58] For some reason, I thought I was supposed to have everything just work. But you're right, this is going to conflict, I just need to fix it every time it does. [07:58] ls [07:58] Nope, meant to put that in my terminal... [07:59] * arraybolt3[m] * apparently thinks Matrix == QTerminal [08:03] I would just wholesale import the Debian changelog instead of messing with the changelog [08:04] Or, if you read the manpages deep enough arraybolt3 @arraybolt3:matrix.org, you will find an option that will let you just favor debian/experimental changes [08:05] I think it's in git-merge [08:05] I've actually just been learning Git from experience, I've not read the man pages deeply at all. I might should sit down and just read Git Pro one of these days so I get a good handle on what I'm doing. [08:05] StackOverflow too :) [08:05] But yeah just wanted to make that suggestion to you [08:05] I learned Bash script from "help " and StackOverflow. [08:05] (And a couple of Googles of official docs, but not many.) [08:07] It's an experience [08:10] Looks like the last time this was synced was before Hiruste! I'm three commits in so far, but is this... right? [08:13] Sounds about right [08:13] :+1: [08:13] * Uh, how do I make this a thumbs up?! [08:14] 👍️ [08:14] (Got it.) [08:21] "I would just wholesale import..." <- Do I just cp it over? [08:28] "Do I just cp it over?" <- Yeah [08:29] I cherry-picked all the way from Severus Septimus's "Update copyright" just after Andrew Lee's "Releasing debian version 3.6.0-1", yet somehow my copyright file and parts of the control file are still different. Can I fix those with cp, too, or do I need to delve deeper into the Debian commit history? [08:33] I'm seeing before the commit I started at, there were some "Cherry-picked new upstream-release " commits with TONS of changes, including copyright file changes. [08:36] (And I don't think the commits I'm mentioning are safe to cherry-pick, since they include source code changes, too.) [08:38] I mean, you can just overwrite the debian dir and do one commit if it's that much of a delta... [08:38] Now you know the process if you do it in the future though [08:38] Nah, just two files. I did it. I even preserved the commit history with the cherry-picking. [08:47] Now is the package you have identical to debian/experimental? [09:10] Except for a single changelog entry for "Backport to Jammy." But... crud, I didn't make the one vital change in Kinetic. I'll have to back up. I'm not in a good position to do that at the moment, though, we may have to finish tomorrow. [09:37] I'm about to crash anyway :) [14:40] Simon Quigley (Developer): I can help with the SRU write-up if you need: https://bugs.launchpad.net/ubuntu/+source/calamares/+bug/1980180/comments/1 [14:40] Launchpad bug 1980180 in calamares (Ubuntu Jammy) "[SRU] Calamares 3.2 Series for 22.04 LTS" [Medium, New] [17:13] Eickmeyer @eickmeyer:matrix.org: Would you like to work together on a dual justification response? [17:16] Simon Quigley (Developer): Sure, we can probably collaborate in a g-doc or something like that. Robie's evaluation is correct, which does beg the question, "WTF, why hasn't this ever been updated, and why are we suddenly doing it now and trying to backport it?" So, we'd better be prepared to answer that. [17:17] I start work in about 45 minutes, FYI. [17:19] I think it's a good question to ask to gain further understanding but I disagree with the stance on it [17:19] Give me a sec to get to my computer, I'll start going through changes etc. [17:19] I also disagree with his stance. Callamares is only an installer, so it's not part of a permanent installation, so his stance is somewhat flawed. [17:20] Additionally, it's seeded by two flavors which are giving it a green light at this point. [17:21] Eickmeyer @eickmeyer:matrix.org: What's a good Google email? [17:21] Simon Quigley (Developer): As to avoid more spam than usual, I'll dm you. [17:21] Ack :) [17:38] Dan Simmons: You around? [17:44] "Dan Simmons: You around?" <- Yes [17:49] "Yes" <- https://bugs.launchpad.net/ubuntu/+source/calamares/+bug/1980180 - may be interesting for guiverc @guiverc:matrix.org Leo K to read too... [17:49] Launchpad bug 1980180 in calamares (Ubuntu Jammy) "[SRU] Calamares 3.2 Series for 22.04 LTS" [Medium, New] [17:53] I agree with both of you. It might be good if we were to compile a list of high profile bug fixes. [17:54] Cool. [18:14] https://tracker.debian.org/news/1341161/accepted-liblxqt-110-1-source-all-amd64-into-experimental-experimental/ [18:14] arraybolt3 @arraybolt3:matrix.org: liblxqt was accepted into experimental ^^^^ [18:18] As usual with Debian it will take upwards of a few hours for that to publish in experimental. From there I can continue [18:19] You're welcome to continue merging/syncing and backporting. I'll try to get as much as I can upstreamed to experimental [18:20] Simon Quigley (Developer): Nice! [18:21] I might not be on and doing things until later today, but I definitely intend to finish the sync from last night (just a matter of a hard reset and a couple of changes before the branch), and hopefully start backporting again. [18:22] Sweet. Let me know how I can help :) [18:46] This might be silly, and if the answer is "No, what are you thinking?!", I get it. I was wondering if I might be able to help with the SRU bug by writing some reasoning behind why I think including the new Calamares into Lubuntu 22.04 is a good idea. I have an argument on why this would directly, positively affect end users that appears to have been missed in the original response to Robbie. I would present it for your guy's [18:46] approval before posting anything on the bug report. Again, if this is ridiculous and I should stay out of it, no worries. [18:46] (New as in the version debated in the bug, not as in Calamares 3.3.) [18:47] Simon Quigley (Developer) Eickmeyer Pinging with above idea. [18:53] arraybolt3 (@arraybolt3:libera.chat): I'd say go ahead and write someting into a pastebin, link it here, and we can review it and see if it would be a good idea to throw it in there as a comment. [18:54] OK. Thanks! [18:56] [telegram] *shoves Simon into the wall* [18:56] * arraybolt3[m] shoves teward into /dev/kvm [18:57] Now escape THAT! [18:57] [telegram] @tsimonq2 if you want to give me upload rights in the event you're dead, feel free. but put me in uploaders if you do that, so i can satisfy Bage's needs if he or the LXQt team needs additional requirements for upload. This said, if you aren't the maintainer yet, and the team doesn't let me have upload rights, then i can't help you there :P [18:57] [telegram] i don't *need* upload privs ;P [18:57] [telegram] *shoves arraybolt3 into /dev/null where he can't do anything* [18:57] * arraybolt3[m] sudo rm /dev/null [18:57] Oh great, now I'm gone. [18:58] * arraybolt3[m] sudo cp /restore/dev/null /dev/null && sudo cp /dev/urandom > /dev/teward [18:58] * arraybolt3[m] * sudo cp /restore/dev/null /dev/null && sudo dd if=/dev/urandom of=/dev/teward [18:59] [telegram] you can't write to read-only medium :P [18:59] [telegram] media* [19:00] * arraybolt3[m] uses RowHammer CVE to escape /dev/null [19:01] * Eickmeyer[m] rolls a nat-20 [19:02] @teward001 Any Drone or GiTea updates? [19:08] guiverc @guiverc:matrix.org Leo K: Sorry for the middle of the night ping but perhaps we can get started on some sort of "new feature tour" to link to in the 22.10 release notes? Perhaps reach out to the community [19:08] lynorian: ^^^ maybe include in the manual [19:09] Simon Quigley (Developer) Eickmeyer: I have written a draft for the SRU bug response. Should I password-protect it, or is a public pastebin OK? [19:09] Meh, f*** it, public pastebin is cool [19:09] paste.ubuntu.com [19:10] Well, so much for word wrap, but here it is: https://paste.ubuntu.com/p/NMpRjry87q/ [19:10] Simon Quigley (Developer) Eickmeyer Pinging with above draft [19:11] Click "View raw", it makes it look better. [19:13] Oh great, I had to mess up the Calamares version, didn't i? [19:13] * arraybolt3[m] is REALLY GLAD this is a draft [19:14] I have a suggestion but give me a few [19:16] * kc2bez[m] > <@tsimonq2:linuxdelta.com> guiverc @guiverc:matrix.org Leo K: Sorry for the middle of the night ping but perhaps we can get started on some sort of "new feature tour" to link to in the 22.10 release notes? Perhaps reach out to the community [19:16] * kc2bez[m] thinks maybe I should start the release notes so we can just keep adding to it. 🤔 [19:17] Might not be until tomorrow for that though. [19:31] "Well, so much for word wrap, but..." <- I agree with your reasoning but I'd remove the last two lines :) otherwise mostly okay, I trust you're cleaning it up further [19:33] OK. I fixed my ?? in the version number, and removed everything after "Hoping this message finds you doing well." [19:36] Simon Quigley (Developer): Quick question, I called you and Eickmeyer "Simon and Eickmeyer", since those are the user names it appears you prefer. But I'm calling one of you by your first name and one by your last name, is that weird? [19:36] Eickmeyer is his IRC nick, tsimonq2 is mine [19:37] Right, but should I change it to "Simon and Erich", or "Quigley and Eickmeyer" (that just sounds wrong for some reason), or use your full names? [19:38] Simon and Erich would be fine IMO [19:39] Simon Quigley (Developer) Eickmeyer Second attempt https://paste.ubuntu.com/p/6sHhYFVmf3/ [19:40] (Can also provide a diff if that's easier) [19:40] Reviewed-by: Simon Quigley [19:40] Reviewed-by: Erich Eickmeyer [19:41] > <@tsimonq2:linuxdelta.com> Reviewed-by: Simon Quigley [19:41] > Reviewed-by: Erich Eickmeyer [19:41] I know you just made up that second email address. [19:41] I was waiting XD [19:41] I like Ubuntu Studio, and it does NOT always segfault. GNOME Boxes always segfaults when I close it when accessing it over SSH. [19:42] * ``` [19:42] Reviewed-by: Simon Quigley [19:42] Reviewed-by: Erich Eickmeyer [19:42] ``` [19:42] LMAO [19:42] * Eickmeyer[m] references the "Simon's Fault" meme [19:43] XD [19:43] I guess I should make a meme for myself (something like "NOOOO... well, crud, now what?!" or something. [19:43] (I have had that feeling so many, many times in the middle of doing something important...) [19:44] Don't worry, you'll get made fun of already as a n00b ;) [19:44] "NOOO! YOU CAN'T REVIVE A DEAD MULTIPURPOSE MULTIMEDIA CREATIVITY OPERATING SYSTEM!!!!!11!!!" Me: "Haha Ubuntu Studio go brrrrrrrrr" [19:45] harumph I was here first!! [19:45] * tsimonq2 shakes his cane at Eickmeyer [19:45] * tsimonq2 yells at a cloud [19:45] * tsimonq2 yells at the cloud [19:45] SPEAKING OF THAT [19:46] @teward001: it's broken [19:46] * arraybolt3[m] yells at RISC-V64 for being slow [19:46] What's broken, you may ask? [19:46] Well, my will to live for starters. (Gen Z joke, don't take it seriously. :P) [19:46] Eickmeyer[m]: And now I get to use an AWESOME Kubuntu-based music-centric distro for everything. I love it. Keep it up. [19:47] * tsimonq2 forces Eickmeyer to switch to LXQt [19:47] * arraybolt3[m] installs KWin on Lubuntu [19:47] [telegram] I have documented how to change the window manager [19:47] * arraybolt3[m] installs Plasma on Lubuntu [19:48] Anyway, what productive things do we have going on today? :P [19:48] lubot: Wow, that's awesome. [19:48] @lynorian: As I would have expected you to have already :) nice work as always [19:48] Simon Quigley (Developer): You were talking about a problem with GiTea, and I'm still waiting on what to actually put in the reviewed-by lines of the draft. [19:48] Eickmeyer @eickmeyer:matrix.org: you're being asked a question. :P [19:49] > <@tsimonq2:linuxdelta.com> ```... (full message at https://libera.ems.host/_matrix/media/r0/download/libera.chat/de13429e097e2c1e57cb1cb46ad12f086ba2fa54) [19:49] Simon Quigley (Developer): Right. [19:49] What's the question? My email address? Use eeickmeyer@ubuntu.com for that. [19:49] OK. [19:51] One more draft just to make sure. https://paste.ubuntu.com/p/jswF2NzkHn/ [19:52] ""NOOO! YOU CAN'T REVIVE A DEAD..." <- Off topic, but why the 11? [19:52] I hope this message finds you well. [19:52] Otherwise LGTM full send [19:52] [telegram] simon: you know that you need to be more specific about shit being broken right [19:52] arraybolt3[m]: ZOMGWTFBBQ!!!!!!1111111 [19:52] [telegram] failure to comply with that means I shove you into the abyss [19:52] @teward001: My will to live, I told you already :P [19:52] arraybolt3[m]: That's what happens when someone lets go of the shift too early [19:53] * tsimonq2 shoves teward into the gulag [19:53] lubot: It just doesn't work. That's always the problem with anything that breaks. [19:53] > * <@tsimonq2:linuxdelta.com> forces Eickmeyer to switch to LXQt [19:53] Done. Install Lubuntu then ubuntustudio-installer and select what you need. BOOM. [19:54] Eickmeyer[m]: That better not be some of that GTK garbage :P [19:55] tsimonq2: Sending? [19:55] I think so. Just waiting on Eickmeyer to approve as well. [19:55] tsimonq2: Actually, it's worse. TK/TCL. [19:55] arraybolt3: LGTM [19:55] 👍️ [19:56] Eickmeyer[m]: Oh gawd. Good lawd almaighty [19:56] tsimonq2: I want to do a rewrite utilizing kdialog or at least zenity and some shellscripting amazement. [19:57] Eickmeyer[m]: Qt 6 bro [19:57] Or GTFO :P [19:57] Shortly I will be taking the time to chug caffeine and port everything over [19:57] Sent the message (sans the ###############DRAFT bit). [19:57] The entire stack of everything [19:59] I should probably finish up on that sync. [19:59] Gogogo :) [19:59] * arraybolt3[m] wishes my desktop could beam its IP address to my Chromebook without me having to get up [19:59] Do you remember what your measure of success is? [19:59] (I like to SSH into my desktop and do everything from where I'm at.) [20:00] arraybolt3[m]: Static IP? Convoluted DDNS? :P [20:00] tsimonq2: Yep, the diff. Right now it's clean except for one changelog entry that i added to the backport, but I made the last couple of commits at the wrong spot. [20:00] tsimonq2: Just a .local DNS would work - my Pi can pull it off somehow. [20:00] Okay, so at least you know where you're at [20:00] That's a good start [20:01] arraybolt3[m]: Yeah some kind of reverse tunneling? [20:01] Yeah but then I'd need my Chromebook to tell my desktop about its IP. [20:02] SSH from the desktop to the Pi using the reverse tunnel options, which will open a port on the Pi you can SSH to [20:02] tsimonq2: Oh good grief. I think I should just set my hostname and see if that does it. [20:03] arraybolt3[m]: It's a one line command, not telling you what to do but def consider it :) [20:03] What's the one line command? The Pi thingy or the hostname? [20:05] arraybolt3[m]: The Pi thing [20:06] If you knew my WiFi speed and network setup, you'd know why I think that's a bad idea for my setup. [20:06] https://www.howtogeek.com/428413/what-is-reverse-ssh-tunneling-and-how-to-use-it/ [20:06] (I have a hotspot, not a router, and it gives me no Ethernet ports, and the WiFi is snail slow - I have to use compression on SSH just to make it bearable.) [20:06] Meh, unless we're talking mobile internet speeds on your LAN I think you may be fine [20:07] (My desktop is at least able to tether to it via USB.) [20:07] Yeah, mobile internet speeds on my LAN. [20:07] Ah. [20:07] In fact, slower than that. [20:07] * tsimonq2 kicks arraybolt3 and tells him to upgrade :P [20:08] Good luck with that - it was either this, dial-up, or sattellexpensive. [20:08] [telegram] ah i can't fix your will to live simon 😔 [20:08] [telegram] :P * [20:08] I've definitely been in the spot where I don't have the money so I won't dig too personal but I spent $500 on a small little Protectli Vault 2 port and a Unifi AP which works wonders [20:08] arraybolt3[m]: Fair enough :) [20:09] @teward001: No but you can fix Gitea though :P [20:09] Also we need a CI [20:09] :P [20:09] And a Matrix instance to migrate our channels over to :P [20:09] At least it's better than the 150 GB a month I had before. [20:10] I guess I could always get an extra router with a USB port for a WAN port and be in business, but this works well enough. I'm one of those "why spend money on convenience when you can just code it" guys. [20:11] [telegram] simon: one thing at a damn time bro [20:12] [telegram] other than SSH which is not working on gittea yet, what's broke [20:12] [telegram] (you aren't council so you don't have admin on the instance) [20:12] [telegram] re: CI, i'm still working on Drone, been busy [20:12] [telegram] re: matrix, again, busy [20:12] [telegram] if you want to nag me on these things, I'll happily exercise the right to boot you in the face for violating the Ubuntu CoC of being nice to others. [20:13] Sorry, just was interested, not trying to nag. [20:13] [telegram] you aren't Simon are you [20:13] [telegram] arraybolt3: you aren't Simon 😔 [20:13] No, but I'm the one who asked for GiTea and Drone updates. [20:14] [telegram] your message went missing in the abyss [20:14] [telegram] Simon on the other hand is annoying [20:14] [telegram] you simply asked if there were any updates. Simon's saying "so we need X, Y, Z" which is different than asking if there's any updates [20:14] yes I'm simply giving you crap teward :) [20:15] well don't give me crap this week [20:15] i've been tasked with rewriting this monolithic piece of shit from another section of my FT job that is old, obsolete, monolithic, and needs nuked [20:23] aHA! I just figured out that the hostname stuff is working just fine on my desktop, I just had the hostname wrong. No more IP address hunting! [20:35] Simon Quigley (Developer): https://github.com/lubuntu-team/libqtxdg-packaging/pull/1 (Since I backed up, I had to force-push my changes, so double-check that nothing broke.) [20:35] Pull 1 in lubuntu-team/libqtxdg-packaging "Backports/jammy" [Open] [20:36] Also check the changelog, I may have botched a couple of lines there. [20:42] (Also, the diff is coming back virtually spotless and the build is Lintian clean except for a thing about the "no-manual-page" override for a lintian-overrides file - the syntax for that file is a bit different between Ubuntu and Debian.) [20:43] (Should that slight syntax difference be an Ubuntu delta? Or do we just live with the Lintian gripes and do a full sync? Or better yet, do we fix Ubuntu's Lintian?) [20:58] Nah it shouldn't be a delta [21:00] Simon Quigley (Developer): And these extra lines in the changelog marked "experimental" are OK? [21:00] * Imported copyright file from upstream. [21:00] * Imported control file from upstream. [21:00] > <@arraybolt3:matrix.org> Simon Quigley (Developer): And these extra lines in the changelog marked "experimental" are OK? [21:00] > * Imported copyright file from upstream. [21:00] > * Imported control file from upstream. [21:00] It looks like something went wrong there. [21:01] That's what I thought. I'll remove those and do another add-commit-push. [21:02] OK, fixed it. [21:03] This is the entire diff between the Debian and Ubuntu debian directories: https://pastebin.com/Kusw5c5x [21:04] One changelog entry, Debian has a gbp.conf file, Ubuntu has a .gitignore file, that's it. Builds Lintian clean (except for stuff that can obviously be ignored). [21:06] Cool, but is that changelog entry in ubuntu/kinetic or backports/jammy? [21:06] Backports/jammy. [21:06] ubuntu/kinetic should be an even clean diff [21:06] arraybolt3[m]: Cool [21:06] The ubuntu/kinetic entry should be even cleaner. [21:06] Oh shoot, I need to push an extra commit in Kinetic. [21:06] (Those two messy changelog lines are still there!) [21:06] * arraybolt3[m] shakes fist at Git [21:07] They aren't here: https://salsa.debian.org/lxqt-team/libqtxdg/-/blob/debian/experimental/debian/changelog That is why I said something went wrong. [21:08] Dan Simmons: Exactly. I had added them during the last little bits of the sync, only in the Ubuntu side. [21:09] Alright, the Kinetic branch's diff is now spotless (except for the gbp.conf and .gitignore files). [22:42] Simon Quigley (Developer): https://github.com/lubuntu-team/pavucontrol-qt-packaging/pull/1 [22:42] Pull 1 in lubuntu-team/pavucontrol-qt-packaging "Backports/jammy" [Open]