=== guiverc2 is now known as guiverc [15:22] Basically just been head-down on this today: https://git.lubuntu.me/Lubuntu/installer-prompt/commit/7ddbe923e2c6fbf6c9803e6c9f6de565009ac004 [15:22] -ubottu:#lubuntu-devel- Commit 7ddbe92 in Lubuntu/installer-prompt "Add WiFi selection feature, just the basics. Needs lots of cleanup and UX love" [15:23] The absolute basics work: on my laptop, when selecting a WiFi network that's already saved, it autoconnects with that existing connection. If it doesn't already exist, it creates one. [15:24] Now, the popup box isn't guaranteed to stay where it is, the next is white so you can't see it, and if you enter an incorrect password there's no way to change it. :P (much like nm-tray, unfortunately, besides the GTK editor we have there) [15:24] *the text [16:43] Kay, so this will most likely need to go on the backburner for the next few days/week. [16:44] arraybolt3: It would be up your alley to finish up, otherwise I can take care of it myself when I'm back [16:44] This is my local diff: https://dpaste.com/HMBWHF42L [16:44] Otherwise it's all pushed. [16:44] That currently won't compile because there's a mismatch in the connections - just trying to figure out that dialog. [16:45] (ie what exact state changes are needed?) [16:45] Anyway, I'll be leaving right now to make a site trip to Madison, so I'll be a Talk to Texan for a bit :P [16:45] Yeehaw 🀠 [16:46] Essentially, probably just needs another full time day or two to implement the l10n backend (as discussed) and finish up this quirky mess. [16:46] Either way, it now works fine if all you need to do is run English and you type your WiFi password first try/are trying on a VM. :P [16:47] Anyway, o/ [17:44] [matrix] Nice, I'll see if I can take a crack at it. [17:45] [matrix] (Usually signal connect issues means you need a QOverload in my experience.) [17:55] [telegram] Essentially, the last thing I need to clear up is whether network manager can actually emit signals when the password is failed and not just disconnect, which gives a generic signal that can't really be interpreted [17:56] [telegram] We might also consider using q u texts for insuring that multi monitor situations work well comma I had a testing instance that had that at some point and it worked pretty well for insuring consistent states [17:56] [telegram] Mutex [17:57] Mutex for multimonitor support? [17:57] I don't get it :P [17:58] [telegram] On my 2 monitors set up. There are technically two instances of the installer prompts, and when both of them are connected at the same time when it's trying to read a manipulate elements, it can cause a race condition where only one of the monitors is updated with the correct information [17:59] [telegram] So, as an example, when network manager emits a connection changed events, it will only update the status on one of the monitors, not both and the lock solves that [17:59] [telegram] How? I have no idea? Because logically that does not make sense, there are 2 separate instances. You would assume there would be two separate locks [18:00] [telegram] Anyway I digress [18:00] [matrix] Ahhh. [18:00] [matrix] That's really weird, I'll try and debug that. Maybe it's somehow starting two windows of the same instance...? [18:03] [telegram] The state transition at the bottom is fully implemented. So feel free to debug that sure [18:03] [telegram] You could likely simulate it in the main file by starting two non full screen instances and seeing what it does [18:04] [telegram] Otherwise, if you have a multi monitor set up or can reproduce one in a virtual machine. Go for that as well [18:04] [telegram] I've been iterating right on my local machine. And I will admit that I have not tested a wire connection yet either. So that is something to look into as well comma I don't want to bog you down too far with it since it is something I can handle in the next week but if you can help me fix some stuff up and have the time, I certainly would appreciate it [18:08] [telegram] Just ran a quick full install of todays daily and selected "Full installation" and none of the 3rd party apps were installed ..is this a known bug? [18:13] [telegram] Yes that's some user interface work I need to do to hook everything up (re @Leokolb: Just ran a quick full install of todays daily and selected "Full installation" and none of the 3rd party apps were installed ..is this a known bug?) [18:13] [telegram] Essentially, when you select full installation, it should check all of those boxes automatically and apparently does not do that comma but if you check the box as manually it does work [18:14] [telegram] Anyway I appreciate you bringing this up [18:15] [telegram] Also essentially at this point. The regular install and the full install options should the exact same thing except the only difference should be those checkboxes that I just mentioned [18:16] [telegram] ok was about to file bug ..should I - your call? [18:17] [telegram] I think a bug report just a track. It would be a good idea, but it's a known bug. So feel free to assign it to me and set it to confirm [18:18] [telegram] Will do and it is on the way! (re @tsimonq2: I think a bug report just a track. It would be a good idea, but it's a known bug. So feel free to assign it to me and set it to confirm) [18:18] [telegram] Thank you very much! I appreciate it [18:25] [telegram] For the record ..https://bugs.launchpad.net/ubuntu/+source/calamares/+bug/2045580 [18:25] -ubottu:#lubuntu-devel- Launchpad bug 2045580 in calamares (Ubuntu) "Selecting 'Full installation' fails to install 3rd party apps" [Undecided, New] [19:34] [telegram] remind me what package manager we use? [19:34] [telegram] is it muon? [19:34] [matrix] apt [19:34] [matrix] :P [19:34] [telegram] because that's a candidate for removal per Release Team since it was purged in Debian [19:34] [matrix] but yes, muon is what we used to use for graphical package management [19:34] [telegram] *slaps arraybolt3 into the abyss* [19:34] [telegram] are there any potential alternatives we've looked at? [19:34] [telegram] We have muon and Discover [19:35] [telegram] @kc2bez is Discover a suitable dropin for Muon [19:35] [telegram] i have to ask because see my ping in #ubuntu-release on IRC [19:35] [telegram] or was it #ubuntu-devel I forget but i pinged you in one of those places after Vorlon poked me [19:35] [matrix] Discover is more "app store"-y, Muon was more of a package management utility. [19:35] [telegram] also see tsimonq2's response [19:35] [matrix] Synaptic would be the best Muon alternative, but it's GTK. [19:35] [telegram] copied from #ubuntu-release: [19:35] [telegram] > tsimonq2, teward: what are your thoughts on https://code.launchpad.net/~vorlon/ubuntu-seeds/+git/lubuntu/+merge/449128 ? (which now needs retargeted for noble) [19:35] [telegram] Gah, not on IRC right at this moment [19:36] [telegram] @kc2bez thta's fine it's why i copied the message here from vorlon [19:36] [telegram] and apologies if i multiping you vorlon ;) [19:36] [telegram] refer to http://bugs.debian.org/1029171 also [19:36] -ubottu:#lubuntu-devel- Debian bug 1029171 in ftp.debian.org "RM: muon -- ROM; Abandoned upstream" [Normal, Open] [19:37] Sad, Muon was superior to Synaptic IMO. [19:37] [telegram] given that the bug states "It’s been superseded by Plasma Discover, using packagekit as a backend, which is actively maintained." i think we have to look at that [19:37] [telegram] Discover is sort of heavy but in theory it provides the necessary functionality [19:37] [telegram] and either use Discover as our sole option and make a note someone can install/use Synaptic if they must, or otherwise [19:37] [telegram] tsimonq2 gave a tentative ack without poking here, asking for RikMills opinion first [19:37] [telegram] but as Lubuntu is not Kubuntu, etc. it's a discussion that should also be here too @tsimonq2 [19:37] Discover really is just not a Muon alternative. It's a whole different app with a similar use case. I think it's probably best to drop Muon and only keep Discover, but still, deep sigh. [19:38] [telegram] because Lubuntu could independently decide things from RIkMIlls / KDE / Kubuntu [19:38] (I just use apt and apt-cache for the same features I used to use Muon for, but Muon was just so cool and useful.) [19:38] [telegram] We shall survive without muon given the demise [19:43] [telegram] @teward001 I say go ahead and give that an ack [19:58] [telegram] stop making assumptions. like the fourth time in the last month. I love you but come on. (re @teward001: because Lubuntu could independently decide things from RIkMIlls / KDE / Kubuntu) [19:59] [telegram] I wanted his advice on whether it should be totally killed, since it is a KDE project [19:59] [telegram] It doesn't mean he's making a decision for us, I'm just seeking input from all involved/affected parties per CoC [19:59] [telegram] also you may note it's still seeded in Kubuntu too [19:59] [telegram] so at the very least they need to be aware [20:02] [telegram] it was my assumption already that vorlon is poking individual flavor teams as necessary (re @tsimonq2: stop making assumptions. like the fourth time in the last month. I love you but come on.) [20:12] [telegram] Incorrect assumption. The package has already been removed from Debian so vorlon is trying to get it removed from the archive altogether to get the new muon from Debian in, which is a totally different component more related to Python than a desktop app. He initially pinged *all flavor leads* to get it unseeded, but it went unheeded from Lubuntu and Kubuntu. I was the only one that answered, and this was back in August. (re @tew [20:22] tsimonq2: I'm not sure if you are disagreeing with the removal of Muon or not? [20:24] Here's the thing: there are applications for which the new muon is a dependency, so there's stuff being blocked from being in the archive. [20:24] I don't have specifics, so don't @ me [20:25] Eickmeyer: Sure, I just have the ability to ack vorlon's final MP but we don't have council quorum yet. [20:26] arraybolt3: I totally get that. Just throwing in my advisory 2c. [20:27] (There's a reason I'm listed as a Lubuntu Contributor in the release announcements. :) ) [20:28] +1 :) [20:28] And thank you! [20:28] Glad to help. It's because I care. Sometimes a little too much. :) [20:30] BTW, ALL: Edubuntu is 🀏 THIS close to having Pi images, so when you're ready I can help with that. [20:30] WOOHOO [20:31] I have a Pi just waiting to be thrown into action there. [20:31] There, however, is a pretty nasty bug with ubuntu-image which is a blocker. Bug 2045336 [20:31] -ubottu:#lubuntu-devel- Bug 2045336 in Ubuntu Image "[3.1+snap3] Attempts to write contents into a file which is not a directory with trailing slash" [Undecided, New] https://launchpad.net/bugs/2045336 [20:39] arraybolt3: RE: #ubuntu-release: Are you, kc2bez, and teward all not council and +1 on removing muon? Isn't that quorum? [20:42] I am no longer on the Council. [20:43] I was on it for a few weeks and then had to take a six-month hiatus, during which I voluntarily resigned. wxl was elected in my place. [20:44] So I am +1 on it, but my vote doesn't really count. [20:48] Ah, gotcha. I missed that memo. [20:54] myself, kc2bez, and simon are though [20:55] so if we all agree (and it's clear also from RikMills' statements in -release that muon is dead), then we can just merge it [20:55] and ack +1 it [20:56] [matrix] I thought Simon was -1? [20:57] [matrix] But yeah, if it's +1 from all three, I can go ack it. [20:57] [telegram] if i was reading his comments to vorlon in -release right he gave it a provisional ack based on what RIkMills' opinion was [20:58] [telegram] which I believe because RikMills says it's not release-supportable would make that a +1 [20:58] [matrix] ah ok, I see it now. [21:01] I thought we were waiting on RikMills for some reason :P I need to learn to read backlog. [21:02] * arraybolt3 continues to fight with an unbelievably rebellious set of virtualization programs that can't figure out how to attach more than one monitor to a VM without complex antics [21:05] Can I just say I hate all things VBox, especially when Guest Additions comes into play. [21:24] tsimonq2: OK, I see why the lock works - there is indeed only one lubuntu-installer-prompt instance, it's just spawning two windows. [21:24] Also it looks broken when the two monitors are different sizes. [21:25] So. Prepping dev environment (in VBox... grumble) [21:41] OK, fixed layout issues, now multiple monitors with different sizes look right :D [22:00] ok, I got it to compile. [22:07] guiverc, LeoK: Next time one of you does a Noble install, please open QTerminal in the installed system, install all updates, then run "sudo apt update", and then run some command that doesn't exist (`b` works). [22:07] er, not quite... [22:07] do a Noble *minimal* install, open QTerminal, install all updates, run sudo apt update, then run a command that doesn't exist. [22:08] Over here I'm getting "WARNING:root:could not open file '/etc/apt/sources.list.d/ubuntu.sources': Unable to parse section data" [22:09] tsimonq2: What exactly did that local diff you shared with me do? I got it to compile, but I'm not quite sure what to use as the commit message :P [22:12] meh, I think I figured it out closely enough [22:14] oh grief [22:14] * arraybolt3 thinks I'm going to shrink my commit message [22:14] https://i.imgur.com/gS6hqDs.png [22:15] * Eickmeyer revels in disgust [22:15] whew, alright, better [22:15] πŸ˜‚ [22:15] https://i.imgur.com/kd9hkrm.png [22:25] arraybolt3: Still looks like it's more than 50 characters, but I'm not about to count. [22:39] [telegram] My commit messages get superlong sometimes heh [22:39] [telegram] Anyway, you got it to build? [22:39] I did :) [22:40] Ended up being a signature mismatch problem possibly combined with QSharedPointer weirdness. [22:41] [telegram] Oh nice. Did you happen to get the WiFi "wrong password" box to show? [22:41] I didn't test it on a machine with WiFi yet. [22:41] I have a machine that is suitable for that but ERR:WILL_TAKE_ME_A_BIT [22:41] [telegram] Makes sense. That's the big hangup tbh [22:41] [telegram] Anyway, will be occupied with dayjob over the coming days probably [22:42] [telegram] So do that or MOTU stuff or whatever you want :) [22:48] May as well start by flashing my SD card [22:49] also I got a borderline mad scientist idea for making a chkdsk for Linux command that uses WinPE in a VM for fixing NTFS partitions :P so I've been researching that a bit [22:50] That doesn't seem openly licensed at all :P [22:50] There's a fsck for that. [22:50] tsimonq2: no it isn't :P but it might be a better alternative than booting into Windows to avoid losing data to ntfsfix. [22:51] * tsimonq2 thunks the mad scientist upside the head [22:51] anyway, totally off-topic for this room [22:51] Flashing a Noble ISO now so I can test the new code and try to get it working a bit better [22:52] If we were pedantic about room topic here, everyone would've been ejected a long time ago. [22:52] valid :P [23:02] arraybolt3, ack & I've seen the ""WARNING:root:could not open file '/etc/apt/sources.list.d/ubuntu.sources': Unable to parse section data"" myself on a box here; i think when I ran a command from history using `rcp` [23:07] Thanks, good to know! [23:07] I'm guessing it's a minimal install bug. [23:21] Turns out you can have a VirtualBox disk image mounted on your host system *and* booted simultaneously. #BadIdea #oops #SomehowMyVmSurvived [23:33] tsimonq2: So remember the old lxqt-powermanagement issues we had before, where systems without a battery would get a notification? We now have the reverse problem - upon booting a system with a battery a notification is made that the battery is present and charging. [23:33] I'm seeing this in VirtualBox, which allows the VM to see the host's battery status. [23:57] Alright, so the incorrect password warning does **NOT** appear. Why? Because NetworkManager::activateConnection immediately returns and the QDBusPendingReply immediately finishes very quickly, even though the connection has only entered the "connected" state. [23:58] So when the "not connected" state is re-entered, that isn't caught. The connection doesn't fail. [23:59] But it does fail. :P