@adamw@fosstodon.org avatar

adamw

@adamw@fosstodon.org

Fedora QA team lead. If Fedora isn't working, it's probably my fault! he/him/his

This profile is from a federated server and may be incomplete. Browse more on the original instance.

cassidy, (edited ) to fedora
@cassidy@blaede.family avatar

TIL Fedora is packaging a web browser app I developed for elementary OS, stopped updating over three years ago, and marked as end-of-life two years ago—yet it happily shows up in Fedora 40 if you search my name. It crashes on launch, so it doesn’t even work…

WHY??

Edit: I guess the package is being EOL'd in Fedora due to it no longer building and this thread, huzzah! My recommendation to distros: don’t package random apps and then not maintain them/communicate with upstream.

adamw,
@adamw@fosstodon.org avatar

@cassidy because it's packaged by a contributor who has not orphaned it, and it was successfully built in all mass rebuilds up to f39.

it failed the f40 mass rebuild, which starts a kind of clock which will eventually get it kicked out if nobody fixes it, but only somewhere around the f43 cycle. per https://docs.fedoraproject.org/en-US/fesco/Fails_to_build_from_source_Fails_to_install/ , this can be sped up substantially by "any interested party" by following steps 3, 4 and 5 on bug https://bugzilla.redhat.com/show_bug.cgi?id=2261079 , so you might want to do that.

adamw,
@adamw@fosstodon.org avatar

@cassidy @carlwgeorge uh, I feel like you're unfairly conflating mine and Carl's replies there. I didn't say you were attacking anyone, or anything about excuses, or fairness, or appropriateness. I just answered your question of why this was the case (by referring to the policies on retiring packages), and explained how it can be sped up, then I went ahead and did that for you. I hope that was useful.

fedora, to fedora
@fedora@fosstodon.org avatar

Fedora Linux 40 is HERE! Check out all our latest variants for desktop, server, and more.

New features include:

  • @kde Plasma 6
  • @gnome 46
  • Fedora Atomic Desktops (rebrand for Silverblue et al)
  • PyTorch / ROCm
  • And more!

Learn more and try Fedora 40 today! https://fedoramagazine.org/announcing-fedora-linux-40/

#Fedora #Linux #OpenSource #Gnome #KDE

adamw,
@adamw@fosstodon.org avatar

@fedora "You will literally be swapping to newer and newer container images when you update rather than managing independent packages and dependencies." isn't quite right. IoT was already an atomic release. But it's switched from the original ostree technology to the new bootable OCI container approach. the ultimate user experience is much the same or improved, but the implementation is actually pretty different.

mcdanlj, to fedora
@mcdanlj@social.makerforums.info avatar

I'm on #Fedora #Silverblue 40, and I'm now seeing a lot of "No video with supported format and MIME type found" errors.

I don't have mozilla-openh264 installed, so I guess that might make sense? Let's fix that...

# rpm-ostree install mozilla-openh264<br></br>...<br></br>error: Could not depsolve transaction; 1 problem detected:<br></br> Problem: package noopenh264-0.1.0~openh264_2.4.0-1.fc40.x86_64 from @System conflicts with openh264 provided by openh264-2.4.0-2.fc40.x86_64 from fedora-cisco-openh264<br></br>  - package openh264-2.4.0-2.fc40.x86_64 from fedora-cisco-openh264 obsoletes noopenh264 < 1:0 provided by noopenh264-0.1.0~openh264_2.4.0-1.fc40.x86_64 from @System<br></br>  - package mozilla-openh264-2.4.0-2.fc40.x86_64 from fedora-cisco-openh264 requires openh264(x86-64) = 2.4.0-2.fc40, but none of the providers can be installed<br></br>  - conflicting requests<br></br># rpm -qa | grep noopenh264<br></br>noopenh264-0.1.0~openh264_2.4.0-1.fc40.x86_64<br></br># rpm-ostree install --uninstall noopenh264 mozilla-openh264<br></br>error: Package/capability 'noopenh264' is not currently requested<br></br>

There's clearly something I don't understand here. I don't know how noopenh264 is installed but not requested and still causes a conflict. I must be alone in having this problem based on lack of bugs that I can find in RH's bugzilla.

I don't think I saw this problem on Silverblue 39, but I was using that for only a few days before rebasing on 40, and was previously on my "classic" Fedora 39 installation where it definitely wasn't showing up with the rpmfusion packages installed.

adamw,
@adamw@fosstodon.org avatar

@mcdanlj we're aware of this, but it's awkward - https://pagure.io/workstation-ostree-config/pull-request/508 . this kind of awkwardness is a lot of the reason the atomic desktops still aren't more prominently pushed yet, we know there are still some awkward edges to sand off.

adamw,
@adamw@fosstodon.org avatar

@mcdanlj @garrett @notting that's an interesting idea, not sure I've heard it floated before...not sure who to pass it on to, maybe mcatanzaro or @siosm ?

LouisIngenthron, to linux
@LouisIngenthron@qoto.org avatar

Okay, I'm doing it. I'm giving a fair shot for the first time in about a decade. You fediverse people convinced me.

adamw,
@adamw@fosstodon.org avatar
adamw, to fedora
@adamw@fosstodon.org avatar
bluca, to random
@bluca@fosstodon.org avatar

v256~rc1 is out! You know the drill, download it, run it, find all the bugs and report them - possibly to somebody else, I'll be at the nearest pub

https://github.com/systemd/systemd/releases/tag/v256-rc1

adamw,
@adamw@fosstodon.org avatar
adamw,
@adamw@fosstodon.org avatar

@bluca @pid_eins well, I think there are two problems, but it's the missing kmod library that breaks boot 100% of the time. the problem with dracut's hook directory being read-only is a real thing, I'm pretty sure, but it doesn't seem to prevent boot, at least on a simple VM install (I can certainly imagine it might do so in other cases). edit: I filed https://github.com/systemd/systemd/issues/32511 for the read-only hooks issue.

aral, to linux
@aral@mastodon.ar.al avatar

Just installing Fedora Silverblue and realised that if you boot into the USB disk in legacy mode, it does a legacy install. If you boot into it in UEFI, it does a UEFI install.

Wish this was made clear in the interface and you could actually choose what you wanted. I mean, I really think UEFI should be the default no matter what these days with the option to change it if you really want to, no?

#linux #fedora #installation

adamw,
@adamw@fosstodon.org avatar

@aral it's not really possible to do a UEFI native install when booted in CSM mode - it's literally not possible for the installer to write a UEFI bootloader entry, the firmware does not expose the necessary interfaces.

we could make it possible to do a BIOS native install from a UEFI boot, I guess, but it seems like a weird thing to do.

adamw, to random
@adamw@fosstodon.org avatar

today:

adamw, to fedora
@adamw@fosstodon.org avatar

today:

  • caught up with shutdown email backlog (not too bad)
  • committed some updated needles I created over the shutdown but just left lying uncommitted on the prod server, did another small test fix
  • cleaned up the discussion thread on my big Bodhi PR a bit so it's easier to review (I hope)
  • updated openQA and os-autoinst packages to latest git
  • generated a big needle cleanup commit for openQA, put it on stg for testing
  • picked at https://pagure.io/fedora-qa/fedora_openqa/issue/108 a bit
adamw, to fedora
@adamw@fosstodon.org avatar

today:

  • won the #flocktofedora #covid lottery! woohoo!
  • lots more trying to get #fedora #openqa update tests working post-F39 branching; I think I have it in a good state now, and gating is enabled for #rawhide again. All tests should pass for Rawhide updates, for F39 updates the desktop_background tests will fail as we don't have F39 backgrounds yet, I have dropped these from the gating policy for now
  • got sysadmin-main powers. somebody has made a terrible mistake somewhere
tehstu, to fedora
@tehstu@hachyderm.io avatar

Anyone happen to know if 40 adds 5 support?

adamw,
@adamw@fosstodon.org avatar
dwarf, to KDE
@dwarf@borg.social avatar

Hmm. It seems I ran into an issue with Fedora 40 and KDE 6 after today's update. Am I the only person with this issue?

When I run systemsettings, I get the following error:

$ systemsettings
systemsettings: symbol lookup error: /lib64/libKF6KCMUtils.so.6: undefined symbol: _ZN32KCategorizedSortFilterProxyModel16staticMetaObjectE

I guess it could be related to the following?:

Problem 1: package telegram-desktop-4.16.8-1.fc40.x86_64 requires libQt6Core.so.6(Qt_6.6_PRIVATE_API)(64bit), but none of the providers can be installed
  - package telegram-desktop-4.16.8-1.fc40.x86_64 requires qt6-qtbase(x86-64) = 6.6.2, but none of the providers can be installed
  - cannot install both qt6-qtbase-6.6.2-7.fc40.x86_64 and qt6-qtbase-6.7.0-3.fc40.x86_64
  - cannot install the best update candidate for package telegram-desktop-4.16.8-1.fc40.x86_64
  - cannot install the best update candidate for package qt6-qtbase-6.6.2-7.fc40.x86_64
 Problem 2: problem with installed package 
  - package telegram-desktop-4.16.8-1.fc40.x86_64 requires libQt6Gui.so.6(Qt_6.6_PRIVATE_API)(64bit), but none of the providers can be installed
  - package telegram-desktop-4.16.8-1.fc40.x86_64 requires libQt6Widgets.so.6(Qt_6.6_PRIVATE_API)(64bit), but none of the providers can be installed
  - cannot install both qt6-qtbase-gui-6.6.2-7.fc40.x86_64 and qt6-qtbase-gui-6.7.0-3.fc40.x86_64
  - cannot install the best update candidate for package qt6-qtbase-gui-6.6.2-7.fc40.x86_64

adamw,
@adamw@fosstodon.org avatar

@dwarf telegram-desktop appears to come from rpmfusion; it looks like it has not been rebuilt for qt 6.7, which is the current version in Fedora 40. this should be reported as a bug at https://bugzilla.rpmfusion.org/ if it is not already.

where did you see that error message? was it related to upgrade from f39 to f40 with telegram-desktop installed? if so it's possible that resulted in an incomplete upgrade, yes. try removing telegram-desktop and doing a system update.

adamw,
@adamw@fosstodon.org avatar

@dwarf it looks like qt 6.7 only recently went stable as part of https://bodhi.fedoraproject.org/updates/FEDORA-2024-c9a2438d21 . so I guess maybe that update got incompletely applied on your system because of telegram-desktop. same advice applies, remove telegram-desktop and run another system update. when telegram-desktop is rebuilt in rpmfusion you can reinstall it, or just use a flatpak as you said. sorry for the trouble!

adamw, to random
@adamw@fosstodon.org avatar

today: did my talk at , I think it wasn't quite as good as at devconf.cz because I was pretty tired, but went well enough! I'll post a link to a recording when one is up somewhere. lots of attending talks (some good ones on CPE team status and rpminspect), hallway track, pub quiz...now to try and get some sleep

adamw, to fedora
@adamw@fosstodon.org avatar

hey folks! I'm floating the idea of possibly dropping client support from the Fedora release criteria (so it'd still be there, but we wouldn't be required to test it every cycle and block the release if it fails). I'm looking for input.

is it a feature of Fedora you value? is it important enough to spend significant time automating (or continuing to test manually)?

thoughts here, or at https://lists.fedoraproject.org/archives/list/server@lists.fedoraproject.org/thread/VKTFTPD5SXYX7YLRU7LFVWKYKKJL5H56/ , or https://discussion.fedoraproject.org/t/fedora-server-front-page-revamp-looking-for-feedback/79037/9

thanks!

adamw, to random
@adamw@fosstodon.org avatar

today:

  • spent the entire day down a vast rabbit hole of ancient Fedora/EL release perl packaging stuff, trying to help a project which really didn't care to be helped. I should've taken a hint and quit sooner, but I am bad at stopping in the middle of things. Oh, well
  • emerged, at least, with https://src.fedoraproject.org/rpms/inxi/pull-request/1 to fix downstream problems

In case anyone was reading along at home with popcorn, here's how that test matrix turned out...

adamw, to random
@adamw@fosstodon.org avatar

today in "why boot chain security really is a thing even though some enthusiasts kinda hate it": https://arstechnica.com/security/2023/09/china-state-hackers-are-camping-out-in-cisco-routers-us-and-japan-warn/

note the quote: "Cisco also said that the hacker’s ability to install malicious firmware exists only for older company products. Newer ones are equipped with secure boot capabilities that prevent them from running unauthorized firmware, the company said."

adamw, to random
@adamw@fosstodon.org avatar

writes long email
re-reads long email to try and edit it to be shorter
adds more sentences to it instead

adamw, to fedora
@adamw@fosstodon.org avatar

today:

adamw, to fedora
@adamw@fosstodon.org avatar

today:

tulpa, to fedora
@tulpa@fosstodon.org avatar

The fact that exists and provides a valuable service just proves that Atomic makes it too hard for sysadmins to customize their installs. Because what they are doing is mostly just changing which packages get installed in the base image.

adamw,
@adamw@fosstodon.org avatar

@tulpa this is broadly the point of https://fedoraproject.org/wiki/Changes/OstreeNativeContainerStable , which we've been working through for a while now.

adamw, to fedora
@adamw@fosstodon.org avatar

: if you're having trouble reaching stuff - it's because we're being DDoSed. fun times. @nirik is on it, as always, as far as it's possible to be on it anyway...

  • All
  • Subscribed
  • Moderated
  • Favorites
  • megavids
  • kavyap
  • DreamBathrooms
  • modclub
  • normalnudes
  • ngwrru68w68
  • magazineikmin
  • Durango
  • ethstaker
  • Youngstown
  • rosin
  • slotface
  • InstantRegret
  • everett
  • JUstTest
  • thenastyranch
  • osvaldo12
  • Leos
  • cubers
  • tacticalgear
  • khanakhh
  • mdbf
  • GTA5RPClips
  • anitta
  • provamag3
  • cisconetworking
  • tester
  • lostlight
  • All magazines