Nitrux 2.5: The latest update to a radical Linux

Nitrux OS is one of the most stylish and innovative distros we've seen so far: systemd-free, based on AppImages, and with a very unusual desktop. Nitrux OS version 2.5 is the latest version of this innovative distro, led by Uri Herrera from Coacalco on the outskirts of Mexico City. It's nominally based on Debian and KDE, but …

[drankinatty]

Wish Nitrux well, but appimage? No thanks

I'm always glad to see new and innovative Linux distributions, but the choice to deliver apps in appimage will alienate many users. While the appimage concept is nice, it's just another attempt to make Linux apps universally installable on multiple distros -- but comes with baggage. See the appimage "Best Practices" for a list of some of the bigger bags Nos. 1-4. (https://docs.appimage.org/reference/best-practices.html)

If you tweak Linux to your liking building and installing various applications on your must-have list, or grabbing appimage versions of your app that use different library versions from the system, then No. 2 kicks in "The AppImage needs to include all libraries and other dependencies that are not part of all of the base systems that the AppImage is intended to run on." Now you have multiple versions of the same library installed, and quite a number of them (the old 50M Qt "Hello World" example).

If newer UEFI firmware is required, that is another stumbling block, but I have to admit, the dark-theme screenshots look pretty good.

[Robert Grant]

Re: Wish Nitrux well, but appimage? No thanks

What's the matter with appimage? I'm only a light Debian user and don't understand such things. Is it that all the dependencies are bundled with each thing you want to install?

[Liam Proven]

Re: Wish Nitrux well, but appimage? No thanks

[Author here]

As some commenters like to point out, there are a *lot* of different distros out there. Packaging for all of them is a complicated process, doable for larger projects targeting the main distros, but problematic for small projects. That's why cross-distro packaging projects exist, and like them or not, they will probably continue to grow.

For instance, you may recall my article on running WordPerfect on modern Linux.

https://www.theregister.com/2022/07/20/wordperfect_for_unix_for_linux/

For a decade or more it was _very_ difficult to install the old freeware Wordperfect 8.x onto any modern distro. It requires ancient versions of dozens of libraries. I never succeeded in getting it totally working.

Now, there is a script to do it for you, but if that is not maintained, it will become hard once again.

This is also a problem with the official Abode PDF reader.

Cross-distro package formats solve that, and as such, they are useful. Yes, natively-packaged apps have advantages, but try installing an app that is no longer supported by your distro -- such as DOSemu 1, say -- and you will quickly run into major problems.

Currently, there are 3 of them: Snap (needs supporting tools and systemd), Flatpak (GUI only, needs supporting tools), and AppImage (does not need systemd or any supporting tools.)

As such, it's the one I personally prefer.

For a small distro project with just a few people, picking an off-the-shelf OS helps, but still leaves them the problem of packaging their desktop and their own apps.

AppImage simplifies this.

So, maybe it is not ideal, but it solves some real-world problems and enables lone developers and small teams to do things that otherwise might be so complex and difficult as to prevent the projects from ever seeing a public release.

As such, I am in favour of it.

Yes, there are issues around AppImage and software updates, but if there's a single known source of the appimage files, and some infrastructure to check for new versions -- which Nitrux does seem to have, by the way -- then this is a solvable problem.

And for comparison, this is very similar to the way that apps are packaged for Apple macOS. It seems to work for them, doesn't it?

[Robert Grant]

Re: Wish Nitrux well, but appimage? No thanks

You know you've made it when your abode has an official PDF reader.

[Ozan]

Re: Wish Nitrux well, but appimage? No thanks

I feel that we collectively forget about static linking.

[CrazyOldCatMan]

Re: Wish Nitrux well, but appimage? No thanks

Now you have multiple versions of the same library installed

Which is not the same problem it would have been twenty years ago when drive space was rare and expensive - nowadays when terabyte drives are the norm (in newer machines anyway) who cares?

(Yes - I would have cared when I was installing Slackware SLS on my mighty 80MB HD attached to my SB16 card - but on any of my current machines - even the old AMD K6 - I really wouldn't care).

And it does have the advantage of being able to have two incompatible versions of the same library on the machine, separated from each other..

As to UEFI - the least I have to do with it the better.. (Dell T40 servers *won't* legacy boot so instead of being my shiny new firewall (Sophos UTM doesn't do UEFI *at all*) it's now my shiny new TrueNAS server..)

[Will Godfrey]

Ummm

Devuan testing has had kernel 6.0 for a while now.

uname -a

Linux devuan 6.0.0-2-rt-amd64 #1 SMP PREEMPT_RT Debian 6.0.5-1 (2022-10-28) x86_64 GNU/Linux

[Liam Proven]

Re: Ummm

[Author here]

I am afraid that for this particular writer, life is too short for "testing" versions. Final release, and maybe the occasional beta, and that's all.

[Anonymous Coward]

This distro...

...sounds like a fucking nightmare to me.

[oiseau]

So much waste ...

The big issue/problem with Linux:

---

The Linux ecosystem is home to literally tens of thousands of highly qualified coders/programmers, a rich pool of talent which could be harnessed towards the same goal and put Linux firmly on the desktop.

But ...

If they all want to be able to shout "Look Ma!!! I rolled my own [fill in here]! and be a primadonna for all of 15', that is certainly not going to happen.

Ever.

---

In the name of the Holy Option®, any/every Tom, Dick or Harry wants to be the next Linus Torvalds (or, the gods forbid, the next L. Poettering) and have his own home rolled/forked distribution or version of [fill in here].

Had a look at how many Linux distributions there are today? It's mind numbing.

Check and see how many dead/abandoned forked distributions fill the Great Distro Cemetery®.

Dead/abandoned istributions that, in many cases, were to all intents and purposes practically identical to the one buried next to it or to the one it was forked from.

Save some innovative theme (?) or dubious concept like being appimage based, just enough to say it is different.

A long time example: just how many bloody packaging systems does Linux need?

I don't have the data but I am confident that if I say that none of the aspiring primadonnas I cite above ever thought about contributing to the Linux effort by pitching in as a maintainer/developer or contributor to some part of the OS.

No ...

They could not picture themselves as part of a team, of a larger purpose: they had to roll their own.

And here we are today: slowly being corroded from the inside by MS with no one having much of a clue.

So much waste ...

Have a good week-end.

O.

[Liam Proven]

Re: So much waste ...

[Author here]

Yes and no.

In this case, mostly no.

I have written about desktop diversity recently. I think it's massive waste of human effort that although there are about 20 desktops in some kind of active current development for FOSS xNix, and about 18 of them are just partial, incomplete re-implementations of Windows 95 – and about 75% of those use C and Gtk.

That's not creativity, that is a near-monoculture.

I like Xfce as a desktop, for example, but umpteen different Debian or Ubuntu derivatives with Xfce on top is not innovation or variety.

Most of them have one or two things they do well. I try to highlight that, but mostly I find this projects of little interest.

Nitrux does something different, and it does it using existing tech from other projects: the core OS from Debian, the core desktop infrastructure from KDE, the app format from AppImage.

This, for me, is an example of cleverly using existing off-the-shelf tech to do something genuinely different.

[schultzter]

Experiments, not failures

Rather than refer to all these genuine efforts as failures or wastes of time consider them successful experiments that taught us something.

It's easier to try something new when you can boot into it off a USB key rather than have to install it into your current setup.

[Liam Proven]

Re: So much waste ...

> I find this projects

I find *these* projects. D'oh.

[Adair]

Re: So much waste ...

How about seeing many of these projects not as 'waste', but as 'learning experiences'?

In fact it matters not whether there are three Linux distros or three hundred thousand. The whole point is that it is an open landscape—people can do what they want, no one is under any obligation to use any or all of what is on offer.

Some of us do rather like to have everything 'nice and tidy', according to our own lights, which obviously excludes everybody else who doesn't share our particular view of what constitutes 'nice and tidy'.

Then there are those who a quite happy living in the jungle, with all it's multitude of niches and attempts to exploit what the environment offers.

[weladenwow]

Re: So much waste ...

Ah but you don't have to worry about all that.

Intall the version / distro that suits your circumstances, and off you go. Sumple as. No?

[Sartori]

Great articles

Thanks for these, I have really enjoyed reading them. I dip in and out of Linux every now and then (though haven't for a while now) and these articles have definitely given me an urge to get back into it. Now I just need to figure out which distro to use, might be nice to see what the state of Linux gaming is like since I fairly recently got a 3060ti card.

[coredump]

Re: Great articles

Likewise. Cheers to Liam for putting in the time and effort to dig in a bit to OSes that I might never have heard of, let alone tried.

I don't always agree with the analysis or conclusions, and I won't try them all as Liam has done, but I've definitely found the write-ups interesting and informative, and went on to look into a couple releases (e.g. MX) after Liam's review.

Not sure if MX will become a daily driver for me vs. today's typical Debian, but it's good knowing it's out there as a viable option. Likely that wouldn't have happened for me without Liam's write-up.

Well done!

[Steve Graham]

It's always interesting to hear about different approaches, but this project doesn't seem to have the critical mass that would give me the confidence to put my eggs in its basket.

(I'm having a scambled metaphors day.)