| enigma9o7 | I'm still confused about this, so I thought perhaps it was due to phased-updates being introduced since kinetic is in beta now, and `apt-mark showhold` returns nothing. So I followed the procedure to disable phased updates outlined here https://help.ubuntu.com/community/PhasedUpdates but the same thing, libc++1 is still held back, and I dunno why. | 00:17 |
|---|---|---|
| enigma9o7 | [and it still tries to remove it and disord if i run dist-upgrade which i dont understand] | 00:18 |
| rbasak | Does discord depend on an older version of that package? | 00:25 |
| rbasak | Ubuntu doesn't ship Discord, so it's an issue for wherever you're getting your discord package from really. | 00:26 |
| enigma9o7 | I don't think so. It appears I can manually update that package, which brings in a couple of its dependencies too. https://termbin.com/pagr | 00:27 |
| enigma9o7 | (to the first question re older version of package) | 00:27 |
| enigma9o7 | As to second one, I thought that at first, it was discords fault, wanting older verison. But now I'm not so sure. If I just try 'sudo apt update' I get the one package held back, nothing about discord. | 00:28 |
| enigma9o7 | And I dunno why thats held back, cuz if it was held back for phased updates, I think I've disabled that. | 00:28 |
| rbasak | If you use "apt policy" to get exact package versions of the relevant packages, then you can say "apt install foo=1 bar=2" etc to specify exactly what you think should work, and then apt will tell you exactly why that's not possible. | 00:29 |
| rbasak | There was a bug in apt's solver BTW. Did you update apt? | 00:30 |
| enigma9o7 | https://termbin.com/nbiw | 00:31 |
| enigma9o7 | I didnt update apt seperately, but other than this one held back package, everything is up to date. | 00:31 |
| enigma9o7 | And specifying the exact version does the same t hing as in the eralier paste, it's happy to install it along with its dependencies. | 00:32 |
| rbasak | I mean specify the versions of packages *at the same time* | 00:34 |
| rbasak | Then apt will tell you why they conflict. | 00:34 |
| rbasak | Like ask it to do what you think dist-upgrade should allow you to do, and it'll tell you why a pacakge has to be removed. | 00:35 |
| enigma9o7 | The only reason I tried dist-upgrade is because libc++1 was held back during upgrade. | 00:36 |
| enigma9o7 | So asking it to install libc++1 is what I expect it to do. | 00:36 |
| enigma9o7 | And it does already try to install the new version whether I specify it or not. | 00:36 |
| rbasak | You also expect it to not remove discord. | 00:36 |
| enigma9o7 | Ah right, of course I don't expect that. | 00:36 |
| rbasak | So say "apt install libc++1=X discord=Y" | 00:36 |
| enigma9o7 | Sure, but it's the same, but ok. | 00:37 |
| rbasak | BTW, phased updates do not apply to kinetic/devel | 00:38 |
| enigma9o7 | They didnt before, but I thought perhaps n ow we're beta thats what's happening. | 00:38 |
| enigma9o7 | https://termbin.com/3zdr | 00:39 |
| enigma9o7 | (which was exactly the same as when I just did apt install libc++1 two pastes ago) | 00:39 |
| enigma9o7 | So if definintely no phased updates, and no conflict with anything, and apt-mark doesnt think its held, then why is it being held back? | 00:40 |
| rbasak | You said you were doing dist-upgrade before, but your pastebin says you used upgrade. | 00:40 |
| rbasak | "apt upgrade" will not remove packages | 00:40 |
| rbasak | (if that is required to bump a package version) | 00:40 |
| enigma9o7 | But that doesnt seem to be required. | 00:41 |
| rbasak | That's why "apt upgrade" is holding packages back - because upgrading them would require other packages to be removed. | 00:41 |
| rbasak | Your "apt install" line in your latest paste does propose to remove packages to achieve what you expect. | 00:41 |
| enigma9o7 | Ah ok. and maybe one of those is a discord ddpendency! | 00:41 |
| enigma9o7 | finally things making sense! | 00:42 |
| enigma9o7 | lemee check | 00:42 |
| enigma9o7 | thanks btw | 00:42 |
| rbasak | Try "sudo apt install libc++1=1:15.0-55.1ubuntu1 discord=0.0.18 libc++-1-14" for example | 00:42 |
| enigma9o7 | wait no what i said is wrong | 00:43 |
| enigma9o7 | cuz if it was a discord dependency then it'd want to remove it when i do the manual install | 00:43 |
| enigma9o7 | This makes no sense at all argh. And why would I want both of those installed at once? | 00:43 |
| rbasak | That'll make apt tell you why they cannot be installed at once. | 00:44 |
| rbasak | Probably they conflict with each other. | 00:44 |
| enigma9o7 | but I dont want both installed at once | 00:44 |
| rbasak | And discord is involved somehow. | 00:44 |
| rbasak | Anyway, I need to go. | 00:44 |
| rbasak | I think it's pretty clear that discord's requirements are involved here. | 00:44 |
| rbasak | I think you're in for a world of pain if you're trying to run the development release but also with third party apt repositories. | 00:45 |
| rbasak | You're inevitably going to bump into more of these cases. | 00:45 |
| rbasak | This sort of dependency hell is why snaps exist. | 00:45 |
| enigma9o7 | Discord doesn't offer an apt repository, they make you redownload every new version manually. But ok, right now it's easy to work around, I just want ot understand why it saying it's held back. | 00:46 |
| enigma9o7 | And I think at least you've told me, it will hold back updating things that will remove other packages. | 00:47 |
| enigma9o7 | But, the thing I can't figure out is why it wants to remove discord if I do the dist-upgrade. | 00:47 |
| rbasak | But doing that is necessary to follow the development release. | 00:47 |
| enigma9o7 | https://termbin.com/mp94 manually worked fine | 00:49 |
| enigma9o7 | I really don't understand. | 00:49 |
| rbasak | discord is scored by apt policy only at 100 because it's not available in an apt repository | 00:50 |
| rbasak | Probably it's being deprioritized because apt assumes it comes from a single apt archive but no longer ships. | 00:50 |
| rbasak | This whole third party apt (or deb) is a huge hack. Even more for the development release. | 00:50 |
| rbasak | You're just bumping into the chaos that ensues. | 00:50 |
| === arraybolt3 is now known as star_n_question_ | ||
| === star_n_question_ is now known as arraybolt3 | ||
| fizzled | I did something daft with 22.10 desktop, If I recall I installed what I thought was the latest AMD 4500U driver. How do I fix this message: W: Possible missing firmware /lib/firmware/amdgpu/vcn_4_0_4.bin for module amdgp | 09:20 |
| fizzled | u | 09:20 |
| arraybolt3[m] | alkisg: OK so there's a regression in LXQt that appears to have been caused by Kinetic updating from Qt 5.15.5 to 5.15.6. I'm trying to see if it goes away with 5.15.5, so I can git bisect. So I have the Qt code from Git, I've set it to 5.15.5, compiled, then installed. Now I just want to make it available system-wide so that I can see if it fixes the regression. | 17:02 |
| arraybolt3[m] | Currently in /etc/X11/Xsession.d/50local, I have "export HELLO=there", "export LD_LIBRARY_PATH=/usr/local/Qt-5.15.5/lib", and "export PATH=/usr/local/Qt-5.15.5/bin:$PATH". Both $PATH and $HELLO are now set and I can view them with `echo` in QTerminal, however LD_LIBRARY_PATH continues to be blank. | 17:04 |
| howarth | Is anyone else seeing iwlwifi throwing 'CSR_RESET = 0x10' and failing to load often when booting? | 17:29 |
| howarth | https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1975394 looks very similar although that bug report doesn't mention the kernel version used | 17:30 |
| -ubottu:#ubuntu-next- Launchpad bug 1975394 in linux (Ubuntu) "iwlwifi sometimes failes to initialze on Z690" [Undecided, Confirmed] | 17:30 | |
| lotuspsychje | howarth: that bug filer mentiones kernel 5.17, kinetic is on 5.19 | 17:33 |
| howarth | Basically half the time I boot under kinetic, I get that failed startup of the intel wireless driver | 17:36 |
| howarth | Never had that under 22.04 LTS | 17:36 |
| howarth | 05:00.0 Network controller: Intel Corporation Wi-Fi 6 AX210/AX211/AX411 160MHz (rev 1a) | 17:37 |
| lotuspsychje | yeah we had a lot of bugs on that chipset before | 17:38 |
| lotuspsychje | file a new bug perhaps howarth | 17:39 |
| howarth | should that go under linux-generic or linux-firmware? | 17:40 |
| arraybolt3 | Probably linux-generic. | 17:40 |
| arraybolt3 | (I don't think Ubuntu can do anything about the firmware code and if the device sometimes works the firmware is probably there already.) | 17:41 |
| lotuspsychje | howarth: ubuntu-bug linux | 17:45 |
| howarth | https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1992194 | 17:54 |
| -ubottu:#ubuntu-next- Launchpad bug 1992194 in linux (Ubuntu) "random CSR_RESET errors in iwlwifi appear under kinetic" [Undecided, New] | 17:54 | |
| howarth | done | 17:54 |
| lotuspsychje | tnx looking good howarth | 17:55 |
| howarth | My guess is that some obscure new wifi feature got enabled between the kernel releases | 17:56 |
| howarth | good luck figuring out which one to disable | 17:56 |
| arraybolt3[m] | howarth: If it's a regression in the code, git bisect should turn it up easily enough. | 18:10 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!