Are available packages being maintained at all?

[wlatendresse]

If you come across a non-working package on a current version of a distribution, create a bug report and find that the version available in the distribution is already 2 years old, and newer versions exist, then the question arises as to what is actually going wrong with this approach. Because, among other things, this is exactly the task of the distributions, otherwise you can completely save yourself the creation of a new distribution, because there are enough distributions.

So how can it be that on the one hand there are endless distributions, but on the other hand you have to realize that many things are unfortunately often too outdated and unmaintained. Is the number of distributions too large? Or are the morsels that you intend to chew as a creator of a distribution too big?

Even if the package in question is fortunately an unimportant package (HyperRogue), the question still arises as to which other packages, which may even be security-relevant, are not up to date.

I would like to read some opinions on this, because after all, "Linux Mint" is represented in the top places according to DistroWatch. If I notice something like that, then I honestly don't know why.

[rcalixte]

Even if the package in question is fortunately an unimportant package (HyperRogue), the question still arises as to which other packages, which may even be security-relevant, are not up to date.

I would like to read some opinions on this, because after all, "Linux Mint" is represented in the top places according to DistroWatch. If I notice something like that, then I honestly don't know why.

Different distributions have different approaches to software management. Some are called "rolling releases" which mean they tend to ship newer software faster. The trade-off is that it is possible for bugs to creep in and degrade the software.

On the other end, there are "stable" distributions though they can be considered closer to "static" in practice. These distributions focus on identifying a moment in time where software is known to be stable (even if that means known but non-critical bugs). The trade-off is that new features and/or bug fixes are not distributed except for the regular interval of releases. The benefit though is longer-term stability as less of the system changes over a longer period of time. Contrary to your statement, there are indeed security updates and you can typically find this information in the core repositories for a given distribution. For Mint, those security updates would come from Debian and/or Ubuntu unless the software is maintained and provided directly by Mint.

Debian and Ubuntu, which Mint largely uses as a base, both release approximately every other year with releases that have a longer shelf life, including security updates. The other end of that spectrum is Arch which is constantly shipping the latest known versions of maintained software. In the middle, there is Fedora, which has a new release about every 6 months and only supports the latest two releases, meaning up to only one year of updates are ensured. These comprise the majority of the Linux ecosystem.

To speak directly to your experience, this is unfortunately a common pattern once the long-term stable releases reach the back end of that cycle. The current stable version of Mint is based on Ubuntu 24.04 LTS which, as the name suggests, was released in April 2024. The next release of Ubuntu's LTS is slated for release in 26.04. Nothing has been officially announced regarding how it impacts Mint.

[wlatendresse]

@rcalixte
The explanations of the first three paragraphs are welcome in principle. However, the facts are very clear to me, as I have been moving in this knowledge area for decades myself.

My statements were rather aimed at the fact that nowadays people have theoretically committed themselves to "continuous delivery and integration" (which I also consider questionable to some extent, but that's another topic), but that on the other hand I come across such things again and again (and this fact is still relatively harmless, since it's just a non-functioning game), which belong to the scope of the tasks of distributions and yet are factually undone.

The real point is: What is NOT done by the distribution manufacturer is automatically imposed on the users of the distributions, which makes it a multiplier in the sense of unnecessary waste of time.

In other words, this wasted time could probably be spent more productively, x-fold. This is a great pity, because the open-source community is unfortunately at a disadvantage compared to the "traditional operating systems" anyway. Unfortunately, such problems and carelessness only intensify this. Or why hasn't Linux long since taken over the desktop? (This is, of course, a purely rhetorical question.)

In short, a non-existent package is better than an outdated package. This way, a user knows immediately where he stands and that he has to take care of it himself anyway.

The impression that there is a solution when there is no solution at all is not being created (a problem of a large part of the IT industry, unfortunately!).

Another example of bugs that simply should not exist: Magic Keyboard Layout Switching

[Jopp-gh]

Open source isn't comparable with commercial -products. I'm very positively impressed how open source managed to build what we have today: a working, alternative system with all kinds of apps for every profession.

I get your problem and think the solution is in your yourself; open source means also:do it yourself and enrich this community.
Of course, not everyone is able to code, but sending bug reports is another way, or to donate.

If you need a quick solution, try to check for alternatives: flatpak and appimages get the latest updates.
Otherwise check GitHub for projects to fill the gap. People help each other in the open source community, be it on discord or mastodon, fediverse, etc.

Open source brings people together (sometimes with a lot of friction but we can always learn something in the process if we keep optimistic and open minded).
The "time waste" you see is our learning experience and differs with the philosophy commercial products accustomed us to think.

I see commercial products more like "fast food", driving consume of things we believe to need.
Also, some distro is cutting edge like arch, but you can find a lot of good info on distrowatch.
You picked with Linux Mint the most conservative distro, sorry. So, it's not really a good idea to complain about Mint's update cycles.

Currently there's a big influx of windows users migrating to Mint, people need to realize Linux is made of donators and volunteers.

Finally, commercial and open source is like comparing apples with oranges, that's my opinion.

[Fantu]

Simply, if someone wants all software keep updated to the latest upstream versions, they should use "rolling release" distros, such as Debian testing or, even better, Arch.
Obviously, bugs are more frequent, and I personally don't recommend it for distros aimed primarily at novice users.

[wlatendresse]

@Jopp-gh
I hope your comment is just a joke, because

1. I have filed a bugreport for the application specified above.
2. I have reported a complaint towords the linux distributor (the exact reason is mentioned above).
3. It does not help at all to change the situation and obviously simply ignores my other texts and comments.

Your whole comment sounds like auto-generated AI bot gibberish-comment whose sole purpose has been to write exactly that comment. This is exactly what makes today's Internet a big pile of fly food.

[Jopp-gh]

Read what Fantu writes and grasp what open source is. Open source is not what you think it is.

Also, you judge people so easily, bc you seem to use AI very very much. I've used it once to test it, for fun. And that's it.

If you can't stand other opinions in a discussion (you don't discuss, in fact, you demand) then let it be to post on social networks. Social means: discussing, understanding, be fair. You're not.

You think to say a lot of smart things but in truth your are new here otherwise you wouldn't ask such questions. Step down from your cloud and start be humble to learn something.

[wlatendresse]

@Fantu

If years old packages are included in distributions, and the user has to update them himself anyway, why are they included there at all? To cause confusion at first?

It simply makes no sense!

And I really don't need any explanation about the differences between individual types of distributions, I'm certainly veteran enough for that.

My main concern was to point out and put up for discussion that and possibly whether some practices may seem quite questionable.

[Fantu]

firstly most of the Mint packages come from Debian, even if old and not kept updated to the latest upstream version they can remain if they do not have unresolved RC (release critical) bugs at the time of the release of a new version of the operating system.
NOTE: This refers to the version at the time of the pre-release "freeze." From that point on, no new upstream versions can be included, only bugfixes.

If they haven't been updated in time (with any new upstream versions at the time of the freeze), there could be several reasons:

• if the new upstream version is newly released and not enough testing or too many bugs, rather than risk not being able to fix the relevant bugs before release,
• if there are blocking dependencies, new packages not yet included or not included in time, not updated enough but which cannot be updated or would break other packages that depend on them
• if the maintainer doesn't have enough time (mostly we are volunteers who maintain the packages in our spare time)
• if the maintainer is no longer active, possibly after a certain amount of time, you can contact the MIA team.
• if the package is orphaned, then there is no permanent maintainer but any changes are made by occasional contributors; in most cases, only RC bugs are fixed.

[wlatendresse]

@Fantu

For me, this essentially boils down to three points:

• If the majority of packages are already based on "Debian Stable," why not simply use "Debian Stable" instead? (Which I'll probably do again soon.) I see no advantages over "Debian Stable," apart from the custom desktop environment. Yet, "Linux Mint" simultaneously proclaims itself to be a "distribution for beginners." I see a clear contradiction here. Because a distribution for beginners would keep the packages as up-to-date as possible, provided that a more recent package is not "more buggy" than the existing one, especially since the aforementioned package obviously has significant problems in its current version, and it would probably be good to have a more recent version
  • because of the associated possibility of easily checking and reporting to the developer whether problems still exist in the latest version, and
  • so that a beginner (i.e., Mint's target audience!) doesn't have to worry about checking whether a package is up-to-date, let alone has to compile a more recent version himself.
• If all the points you mentioned are crucial for keeping packages up to date, then there seems to be a lack of communication channels and automated processes that consider such possibilities and react accordingly, perhaps semi-automatically. Because (once again, the question!) if the user (of whom there is hopefully more than one, which automatically multiplies the associated effort and perhaps calls into question the purpose of a distribution) has to take care of keeping a package up to date himself, then EXACTLY WHAT is the point of including it in the distribution at all?
• The status of the package in question has not changed in "Debian Sid" either (which admittedly doesn't necessarily belong here, but perhaps illustrates one of the problems why the current tree-like interweaving of distribution dependencies is, in my opinion, unfortunately quite a problem, not only because of potentially outdated package versions).

[Fantu]

I'm not going to write pages about how things are or could be regarding package management or distribution.
I just want to write about an essential point that few people understand: open source software is taken too much for granted, almost everyone demands a lot while giving nothing or very little in return. And this is the real main problem!

There are plenty of people who criticize and make demands, but most of them contribute little or nothing.
Before saying that everything should be kept up to date and quickly, you need to have the resources to be able to do so technically, while also maintaining excellent quality.
Open source software is exploited too much without receiving anything in return, or too little, even from companies that invest little or nothing in contributions to the open source software they use. If they used something similar but proprietary for the sole purpose of licensing (and therefore no support), they would have to spend from thousands of euros per year to millions or more, whereas the contribution they make for the software they use is zero or almost zero, but a few thousand euros for fairly large items. Then there are a few exceptions that invest ‘a lot’.
If there were enough resources, not only money, but above all people and time dedicated to it, then we could try to keep all the software up to date, tested, and good enough.

In the current situation, Debian seems very good to me despite its limited resources; in fact, I would say it's excellent. Of course, there are software programs that are better maintained, updated, and tested than others, but this depends almost entirely on how much investment they receive and/or how many other people volunteer their time and how much time they devote to it.

Mint also seems to be doing well, despite its even more limited resources.
The decision to base it on LTS versions of Ubuntu and stable versions of Debian is excellent, given the limited resources and the main target user group.
Basing it on a rolling release would waste too much time and lower the quality, resulting in a significant deterioration, mainly for inexperienced users.
Even the recent decision (https://blog.linuxmint.com/?p=4991) to move toward longer releases and more time spent developing and testing new versions seems like a great improvement to me. Doing the opposite—more frequent releases and more up-to-date software—would certainly make things worse.