What exactly is the point of rolling release? My pc (well, the cpu) is 15 years old, I dont need bleeding edge updates. Or is it for security ?
For software developers, it is better to have frequent tiny changes that can break things, than a big mess of breakage.
Do you hate distractions? Do you love steady improvements? This will affect your preference and judgement about rolling release.
The same can be true for desktop users. It also depends on how stable your software is. If you use mainly vim, dwm, and LaTeX, very few changes will break your flow.
I use a rolling release for mainly 3 reasons.
- Faster access to new (shiny) software/applications. Flatpak and the like could solve this for LTS distros.
- Security updates come faster and smoother.
- Less chance of an update breaking things. Lots of small and frequent updates, instead of rare and large update packs/stacks.
Less chance of an update breaking things. Lots of small and frequent updates, instead of rare and large update packs/stacks.
I would say a rolling distro update has a higher chance of it breaking something. Each one might bring in a new major version of something that has breaking changes in it. But that breakage is typically easier to fix and less of a problem.
Point release distros tend to bundle up all their breakages between major versions so breaks loads of things at once. And that IMO can be more of a hassle then dealing with them one at a time as they come out.
I tended to find I needed to reinstall point release distros instead of upgrading them as it was less hassle. Which is still more disruptive then fixing small issues over time as the crop up.
I would say
Is this based on experience? Or are you guessing?
I ask because my lived experience is that rolling releases break less in practice
Before I used rolling releases, I spent more time dealing with bugs in old versions than I do fixing breakages in my rolling disto.
And non-rolling “upgrades” were always fraught with peril whereas I update my rolling release without any concern at all.
Upgrades is any security or big fix as well. Those tend to be quite safe in point release distros. Upgrading to a new point release version is has all the same problems the rolling release had over the same period all bundle in one messy upgrade (which makes them a huge pain to deal with as they often compound). But between those, the patch upgrades tend to be quite smooth.
I would say the over a longer time period rolling release break in bigger way less often. But they tend to have more but smaller breakages that are easy to trivial to fix.
Good point. Yes. Small breakage means it’s easier to fix. Although, the years I’ve run my rolling release system, I’ve had it break maybe one of two times. Easily fixed. Both of those was because there was a change that needed a manual intervention, which I did not read about until after, so those were my own fault.
Where did the idea come from that rolling releases are about hardware?
Hardware support is almost entirely about the kernel.
Many distros, even non-rolling ones like Mint and Ubuntu, offer alternative kernels with support for newer hardware. These are often updated frequently. Even incredibly “stable” distros like Red Hat Enterprise Linux regularly release kernels with updated hardware support.
And you can compile the kernel yourself to whatever version you want or even use a kernel from a different distro.
Rolling releases are more about the other 80,000 packages that are not the kernel.
This is admittedly anecdotal, but my experience with point releases is that things still break, and when they do, you’re often stuck with the broken thing until a new release comes out. For this reason, among others, dist-upgrades tend to be extremely nervewracking.
With a rolling release, not only are fixes for broken things likely to release faster - if something does break, you can pin that package, and only that package, to an older version in the meantime. Then again, I’ve been using Arch almost exclusively on my desktop for about 7 years and I’ve never had to do this. I don’t doubt that things have broken for people, but as far as I’m concerned, Arch just works.
As far as security goes, I don’t think there’s much, if any, advantage. Debian, the stablest of them all, still gets security updates in a timely fashion.
Dist upgrade is only dangerous for the people who do it. Wait a few weeks before upgrading to the new release and most broken things will be fixed already. I used arch a lot, and I do like the idea of rolling releases, but at this point for the couple programs I need new features in, I just build them from source.
Rolling vs. point release is not about whether a breaking change happens or not but when.
With rolling, breaking changes could happen at any time (even when inconvenient) but are smaller and spread out.
With point release, you get a big chunk of breaking changes all at once but at predictable points in time, usually with migration windows.
I use ancient hardware (as old as 2008 iMacs) and I greatly prefer rolling releases.
Open Source software is always improving and I like to have the best available as it makes my life easier.
In my experience, things just work better. I have spent years now reading complaints online about how Wayland does not work, the bugs in certain software, and features that are missing. Almost always I wonder what versions they are running because I have none of those problems. Lots of Wayland complaints from people using systems that freeze software versions for years. They have no idea what they are missing. This is just an example of software that is rapidly evolving. There are many more.
Next is performance. Performance improvements can really be felt on old hardware. Improvements in scheduling, network, and memory handling really stand out. It is surprising how often improvements appear for even very old hardware. Old Intel GPUs get updates for example. Webcams get better support, etc.
Some kinds of software see dramatic improvements. I work with the AV1 video codec. New releases can bring 20% speed improvements that translate to saving many minutes or even hours on certain jobs. I want those on the next job I run.
I work on my computer every day and, on any given day, I may want or enjoy a feature that was just added. This has happened to me many times with software like GIMP where a job is dramatically easier (for example text improvements tag appeared in GIMP 3).
If you do software development, it is common to need or want some recently developed component. It is common for these to require support from fairly recent libraries. Doing dev on distros like Debian or RHEL was always a nightmare of the installed versions being too old.
And that brings me to stability.
On systems that update infrequently, I find myself working against the software repos. I may install third-party repos. I may build things myself. I may use Flatpak or AppImage. And all of that makes my system a house of cards that is LESS stable. Over time, stuff my distro does not maintain gets strewn everywhere. Eventually, it makes sense to just wipe it all and start fresh. From what I see online, a lot of people have this experience.
On of the biggest reasons I prefer rolling releases with large repos is because, in my experience, they result in much more stable systems in practice. And if everything comes from the repo, everything stays much more manageable and sustainable.
I use Debian Stable on servers and in containers all the time. But, to single it out, I find that actually using it as a desktop is a disaster for all of the above reasons but especially that it becomes an unstable mess of software cobbled together from dozens of sources. Rolling releases are easier to manage. This is the opposite of what some others say, I realize.
In fact, if I do have to use a “more stable” distro, I usually install an Arch Linux Distrobox and use that to get access to a larger repo of more frequently updated packages.
I have a relatively new PC and eventually I decided at Debian Stable.
Granted, I was already somewhat familiar with APT and Debian based systems, but I also was thinking to choose something different or even a rolling release distribution…
…but at the end of the day, I wanted a stable, useable, tested and functional system that I can’t easily fuck up or can restore if needed, because, well, it won’t be a first time I bork a Linux system with misconfiguring stuff or doing something straight out stupid. But this is irrelevant this case.
I ain’t that super familiar with Linux world, so I deliberately chose the safe way. My hardwares are working fine, I have the drivers that work for everything, games running amazingly well… in the past 2 years I use Linux as main OS, I had no problems not being bleeding edge. I kinda had some minor FOMO when Plasma 6 came out and I was “stuck” on 5 with Debian 12, but didn’t had to wait too much for Debian 13 that has Plasma 6 by default. Though, I reinstalled everything when 13 came out - but only because I wanted some changes on my partition table, I added a new disk and… I wasn’t quite happy how I managed some things with it so I wanted a fresh start - so wasn’t upgrading to 13, but I assume it wouldn’t be a problem either, not too long ago I upgraded my server from Debian 10 to 12, without issues. (From 10 to 11 and to 12. First I tried from 10 to 12, that was a disaster though. However, the documentation explicitly said not to do such thing, so it was on me.)
I was tinkering with my tech stuff all my life, I now really just want a stable, working OS. But it’s just personal preference, I have nothing against rolling release and I can imagine that there are scenarios where rolling release is the better choice.
It is funny. You and I landed in different places but for almost the same reasons.
I use a rolling release because I want my system to work. “Tinkering with my tech stuff” is an activity I want to do when I want and not something I want thrust upon me.
On “stable” distros, I was always working around gaps in the repo or dealing with issues that others had already fixed. And everything I did myself was something I had to maintain and, since I did not really, my systems became less and less stable and more bloated over time.
With a rolling distro, I leave everything to the package manager. When I run my software, most of the issues I read other people complaining about have already been fixed.
And updates on “stable” distros are stressful because they are fragile. On my rolling distro, I can update every day and never have to tinker with anything beyond the update command itself. On the rare occasion that something additional needs to be done, it is localized to a few packages at most and easy to understand.
Anyway, there is no right or wrong as long as it works for you.
I tend to use rolling release because I tend to update my hardware with some regularity. Also I hate opening an app or a desktop environment and when I go looking for newly announced features they aren’t there within a day or two. Considering current hardware prices the latter has been my primary motivation last couple years.
What exactly is the point of rolling release?
Newer features. At the cost of a higher risk of stuff breaking.
Or is it for security?
No, point release OSs do have security updates. It’s feature updates that they avoid.
I find that if my updates aren’t frequent enough, then I just forget to do them altogether.
I was just experimenting with arch for my first time last week and I encountered an issue with sunshine (screen streaming software) requiring CUDA for encoding with my GPU but nvidia no longer supports GTX1080 for CUDA, which means I would have to resort to software encoding. Or I can stay on an older version of CUDA but then it will eventually become or maybe already has become incompatible with the rest of the rolling packages. Are there solid workarounds for these types of scenarios?
You’ll need to update to a point release sooner or later.
Are you the kind of person who lives to peel off the band-aid or pull it off in one go?
I prefer to peel mine. I’ve learned from pulling stitches by ripping it off.
On a more serious note: btrfs and timeshift are 👌. If there ever is a botched package, I’ll just roll back to this morning and keep working. It’ll probably be fixed by tomorrow.
For me its security patches. I frequently lock app versions manually.
I do have an old laptop that uses a fixed release, because it sees infrequent use.
One needs to adjust whats needed per usecase. For me that means daily drivers get semi-rolling or rolling. Where stability is neede/older systems, fixed releases.
They are cool cos you get to say “btw I use <insert-distro-name>”.
Also, one big advantage is the end of big disruptive updates - e.g. the one from Win10 to Win11.
You don’t have to live on the edge either. Arch for example has an LTS version.
Interesting. I did not know Arch had an LTS version.
It has an LTS kernel. Not a separate version. This does not make everything LTS. This is very different from LTS distros.
Ah. Not as interesting then
I follow linux news I like hearing about new features and have them come to my system quickly. I’d hate to have to wait years for anything new. I’m not worried about stability, i run nothing critical and if I have an issue I’ll just fix it.
It is not just about your pc hardware. I much prefer running the latest software on it as I regularly use features from tools added in the last version of something. I would hate to have to wait 6 months to a year to be able to use new features that make my life easier. That might not be every bit of software I use but enough core things that I would notice.
For me its especially the DE
Agreed. That is why I think Linux Mint Debian Edition (LMDE) is a better desktop than Debian Stable. The DE and everyday GUI tools stay fresh even if most of the distro is a time capsule.
