Are you fucking kidding? I just switched to bazzite from Mint after dropping windows.
post
All the issues bazzite has with the Steam flatpak are solved by this AppImage
Bazzite shutting down would be tragic ( for me ). I've been quite happy with it. Slowly convincing people to switch. Bazzite was also a relatively low step compared to fixing them a standard distro as all the stuff they want/need is already readily available.
Telling people I decided to promote another distro because the old one stopped is only going to make me have to restart my efforts with an additional hurdle ( how do you know this one won't stop ).
I tried a few distros. But I didn't like garuda very much, nobara was okay'ish. Bazzite really clicked for some reason, despite needing a bit of getting used to.
I'd hate to go distro hunting again. Or try and update cpu schedulers and not brick the whole system.
Kind of amazing that Steam is a 32 bit-only program on Linux, and everyone is making the Fedora project out to be the bad guy here somehow.
Wouldn't the alternative be losing a LOT of old games on Steam?
There should be nothing wrong with a 64 bit program launching a 32 bit program. If anything, it sounds like Valve is using that as some sort of "comparability insurance"; if you can run steam, you can run almost all games on steam.
MacOS stopped supporting 32-bit with Catalina and, according to appleinsider.com, older games won't work on Catalina at all.
Steam is just one if the things that would break. It's more than just gamers pushing back against this.
I wish people would read the discussion on the Fedora forum...
Can you post the discussion you mean?
Can someone Eli5 egg 32 bit architectuur is still a thing? Why should Steam be still in 32 ?
Old games and software that will never be updated were built to run on 32 bit libraries.
Yeah, but I don't get that I'm ringing 64 bit since win7 and never had trouble running old games.
So why still run an is that is dependent on 32 bit?
Win64 contains a full Win32 system.
The 32 bit libraries are still available for you to run that old software. If it was removed entirely those old games would not run
Followup question, if you don't mind! What still needs to be maintained on the Win32 system on behalf of the Fedora maintainers? If everyone has moved on from 32bit, and the old stuff doesn't change, where is the maintenance requirement? Could we not find a "final" version and leave it static, but still available in the package manager?
Is it that packaging requirements change for different systems to keep up with hardware drivers/new package managers/kernel removing deprecated features/security vulnerability patches?
If everyone has moved on from 32bit, and the old stuff doesn’t change, where is the maintenance requirement?
The problem is that it's not old unchanging code, people want the latest supported version so they can still run their 32-bit binaries with the latest supporting libraries.
And if the upstream developers don't consider 32-bit support important, then it falls on the distro maintainers to patch the code to keep it running in these situations.
That's essentially what has happened in some distros like Arch Linux where you have to explicitly install the 32bit versions alongside the 64bit versions.
Also, Steam provides some 32bit libraries exactly for this reason.
For one, its just a burden for maintainers to continue to make sure stuff still works with the presence of those libraries. Unmaintained software is increasingly subject to vulnerabilities, adding an additional burden for maintainers that cover off on the security aspect of packages. It also can hold back further development in other areas simply because they need to be able to install those older 32bit libraries.
The solution to a lot of these problems is containerisation, whereby you're effectively able to shift these issues off into their own area that has no affect on the rest of the operating system, but can be safely accessed when needed.
Steamos is based on arch and also immutable. Can't they do something like that. Never liked bazzite cause it was fedora based. I feel like any gaming distro should be arch based and making it immutable would fix the arch breaking fear.
Why is everyone downvoting? It's not really a stupid question when they apparently already tried it.
Apparently they looked at it and had issues and bugs with arch which could not be fixed:
https://www.reddit.com/r/linux_gaming/comments/1lj9oqw/comment/mzkxkz2
Building a distro is hard. Building a distro on existing architecture like uBlue is easier, but still hard, especially if you want to do it right. I believe uBlue was the right choice. This is just someone over in upstream doing something stupid.
You do realise that dropping 32bit libraries would mean all Fedora flavours get effected, right?
I’ll just move to a distro that isn’t affected.
Just use Steam, Lutris, Heroic, and whatever as Flatpak. Flatpak solved the problem years ago. 32bit libraries there are just runtimes working independently of the host OS packaging.
It's fine to relegate x86-32 to a sandboxed environment.
Build Bazzite on the latest Ubuntu stable with snaps removed and 32-bit libraries installed alongside Steam .deb
I don't see it happening, ubuntu tried and backed out last second years ago. Project mangers just get fussy about putting in resources they don't directly see the use of. It's like that new manger you get that thinks they should remove all these processes as they don't see the use with them. Then slowly as they learn the hard way add all of them back.
It's like that new manger you get that thinks they should remove all these processes as they don't see the use with them. Then slowly as they learn the hard way add all of them back.
And this is the reason “why?” is such a powerful question. If you’re coming into a role and want to reduce overhead, the first question you should ask is why things are done the way they are. In many cases, you’ll discover real tangible benefits to the processes that you would have otherwise missed.
In some cases, the answer will be “I don’t know, that’s just how we’ve always done it.” And those are the ones that are worth scrutinizing further, because that’s one of the most braindead sentences in the English language. If nobody can explain why a process is in place, it’s probably worth reexamining.
“I don’t know, that’s just how we’ve always done it.”
In my experience there often is a reason for it, it's just that the person who knew the reason for it left the company and it was never properly documented, so now it's just cargo-cult-policy without any understanding behind it. So you're right, there's no way to figure out why it's done that way or if it should continue to be done that way without thorough reexamination and a pretty analytical approach, and when those reasons are old they can certainly turn out to be badly outdated, but I'd also caution against just blanket assuming that it probably isn't necessary simply because it's "the way we've always done it" and no one seems to know why. The erosion of institutional knowledge is relentless, but that doesn't mean it was never known or never for a good reason either. It's not braindead to follow a policy you don't understand the reason for, it might be lazy and it might be putting too much trust in the people who made the policies, but it's not always wrong. Sometimes the policies are written in blood, and you not knowing that doesn't mean it's not a good policy.
Which is why I said those procedures are worth reexamining, not just outright discarding. I tend to be against the DOGE chainsaw “just start hacking things off until it stops working” method. But asking “why” is a good litmus test for whether or not a procedure is worth spending time on.
It’s entirely possible that the “always done it this way” method is in use because that method works. The method was built using institutional knowledge that has since been lost. And those are worth examining for the exact inverse reason; To be able to reverse-engineer the institutional knowledge, and fucking write it down for the next person.
Just a heads up, you can vote one what you think about Fedoras proposed change. They may not listen to it but it doesn't take much effort to let them know what you think about it. https://discussion.fedoraproject.org/t/f44-change-proposal-drop-i686-support-system-wide/156324
A lot of apps and most games out there are 32 bit and need 32 bit support, trying to pull an Apple is naturally going to be unpopular and unproductive. This is one of those things about Linux and Linux development which is inherently problematic, which is the seemingly severe aversion to legacy support, despite the fact that Linux is often praised for running on older, less capable hardware, as well as taking better advantage of current gen hardware.
I don't even know why this is even being discussed, it's an awful idea and should never be implemented.
That would definitely suck.
I just recently went over to Bazzite as a former long time Windows user, and I've been enjoying it more than I thought I would. It'd be a shame if I have to go and look for another OS that would work well for me
If it comes to that, rebase to Silverblue and use the Flatpak version of Steam.
Nobara is likely the best candidate for you. https://nobaraproject.org/
Potentially the same problem since Nobara is also Fedora based.
What's even the argument for dropping 32-bit support? Just developer experience?
I'll be gutted, bazzite feels like my forever distro and I really don't want to give up atomic either!
Didn't Ubuntu propose the same a while ago and had to step back because of all the backslash?
Is the change coming from Red Had this time? They're enterprise only so it's possible they don't care about home users whom are the ones still in need of 32bit libraries, I think big enterprises would use Windows virtual machines for that.
I'm not personally impacted since I use Linux MX on my gaming desktop (Debian based) and Debian stood up during the Ubuntu debacle to state they have no intention whatsoever to remove 32bit libraries in the foreseeable future, but it's certainly a blow for a lot of people, I hope Fedora change their mind about it.
Not that this is the final word on the matter, but it is perhaps hopeful that a plurality of voters are strongly opposed at present.

Can someone enlighten me on how such communities plan to deal with the 2038 problem? Is it even relevant to them?
Came here to amplify this crux: Communities need reprogram older software to be Unix year 292,277,026,596 compliant.
Time to reinvest.
Not really. 32-bit libraries are mostly relevant for people who want to play old games, which mostly means old Windows games. Windows software doesn't tend to use Unix timestamps a lot.
all 4 comments