Would Trisquel benefit from being based on Manjaro in the future?
I do understand that there is Parabola, but as it stands right now with Trisquel, it is based on the LTS releases of Ubuntu and currently we are behind an LTS release. Would Trisquel benefit going forward to be based on something like Manjaro with the semi-rolling release structure? It could benefit from the Trisquel team cleaning up packages (like Ubuntu) and have their libre version of the AUR. Having more up to date packages could be a benefit and would cut down on having to prep an OS for a specific release cycle.
Just something to consider. :-)
Trisquel follow LTS releases for stability.
https://wiki.parabola.nu/Rolling_vs._LTS
Rolling vs LTS is a debate but do you really need the "latest" version of "bash" or "coreutils"? Documentation are the same more or less between releases. Packages in Trisquel are not obsolete, you have security fix and a ChangeLog that tell what about it.
Speaking about packages, the link is broken. http://packages.trisquel.info/
"Stable" in software sense doesn't always mean "stability", it is more in line with locking a version down. For server use, I can see using that, but for every day applications, it is nice to have newer software. The latest Trisquel is based on software from 2018 and the most recent LTS of Ubuntu is already a year old and Triquel 10 isn't close to being out yet.
Agreed. I tell you a truth about Ubuntu LTS.
Q: Why could Ubuntu LTS provide 5 years of support?
A: Because Debian provides 6 years of support, and Ubuntu LTS is released one year after corresponding Debian release (6-1=5).
"Stable" means more of "immobility" and less of "stability". Bugs appearing in (semi)rolling released distributions get fixed very quickly, whereas bugs in an LTS distribution usually never get fixed.
Actually, Ubuntu LTS itself was never advertised for average users. It may benefit certain enterprises and organizations. But it would be much better if such organizations simply switch to Debian stable.
Manjaro and other Arch-based distributions have major freedom issues. Their main repositories don't reject non-free software. One workaround is to create a package that is conflict with any known non-free package, like what Parabola and Hyperbola do.
Anyway, re-basing on Debian sid, Arch, or even Fedora rawhide at least partially solves the glaring obsolescence problem, which explains why I never recommended others to use Trisquel, because I believe we should explore the future rather than dig the past.
One obvious trend is that users are increasingly appreciating freshness rather than so-called stability. Top three distributions on Distrowatch.com are all Arch based.
If Parabola is based on Arch and Manjaro filters through Arch, what's the effort involved in using Parabola's repositories for a version of Trisquel and then following Manjaro's package releases? I know there's the AUR involved too which has non-free software, but as mentioned earlier, a libre Aur would be nice :-)
I don't see that Manjaro filters through Arch. In fact, Manjaro has no short of non-free software. I don't know what would happen if I simply install Arch and then install use the "filter" provided by Parabola (i.e., a specific package that is conflict with all known non-free packages). AUR is another problem.
CORRECTION to my previous post: All Distrowatch.com top three (MX, Manjaro, Endeavour) can be rolling or semi rolling. MX is Debian stable based, but users can edit the apt configuration file to make it testing or unstable.
> I believe we should explore the future rather than dig the past.
Is that why you have been repeatedly preaching the socialist revolution in this very forum?
As you once mentioned, we actual Trisquel users are all archeologists: "We dig, dig, dig, dig, dig, from early on to night...to dig, dig, dig, dig, dig, is what we like to do!" "In the past...in the past...Where a million packages...shine!"
I'm not sure if Manjaro is the best choice as a base, but I also find the packages in Trisquel outdated for my needs. I've been using Debian sid (main only) for about a year now and I find it pretty stable, though very occasionally there is some minor breakage. If I used Trisquel, I would probably end up installing a lot more software from random PPAs, compiling from source, or Flatpak.
I have a growing folder of App Image files on my Trisquel system as well as a few PPAs.
I am a translator!
I did not feel the need for any newer version than those in Trisquel repositories.
However, as I was asking for support about Jami and Gajim, I was told to use newer versions, they would not provide support for the versions in Trisquel repositories. So I got Gajim from Guix and Jami from the PPA.
Other than that, I use the Protonmail bridge Appimage, I compiled Barrier (not in Trisquel repository) and I have a PPA to mount the content of a Google Drive.
I tried a few things from Guix and Flatpak but concluded that they either did not work well enough or were not really useful.
I saw similar ideas in the past in this forum. I think this idea is
better and more appropriate to be proposed to Parabola and Hyperbola
distros rather than Trisquel.
With respects,
Malsasa
Refusal to criticism does no good. We just wanted to make Trisquel better. Ubuntu LTS is one of the worst possible bases.
Otherwise, Trisquel will eventually become laughing stock. Actually, Trisquel is already laughing stock in our community.
Once I tried to make a presentation using a 100% free/libre computer (coreboot, no ME) running Trisquel 8. But my Atheros wireless NIC was recognized as wired NIC, and I had nowhere to enter SSID and password. The participants of my presentation said that Trisquel was the stupidest distribution. And thanks to the "stability", this problem never got fixed until EOL.
My PPAs and App Images are for programs that aren't even in the repo let alone too old.
I have always been very happy with Trisquel (7, 8 and 9), and I also understand that different needs call for different choices. Until now, I have found no convincing argument to support the idea that Trisquel would be covering its niche better by rebasing, but I share some of the recurring worries about some of the possible directions taken by Ubuntu in the future, and how badly they could affect a free GNU/Linux distribution based on Ubuntu.
That said, being based on the previous LTS also gives the Trisquel maintainers more time to see what is coming in their way. I trust them to make the right choices, because I see no reason not to. I do not believe they are going to take their cues from random people who clearly stated that they are, in fact, no Trisquel users.
EDIT: Deleted part of a sentence that might have come across as disrespectful for parrots. I like all parrots, especially the birds.
Maybe it would be a good thing to start new Free GNU/Linux Distros. It should not to be difficult to fork and purge a distro based on Arch/Parabola.
It is not endorsed by the fsf but if you like very old software ;-) and people do
And why not a Linux From Scratch with only Free Softwares. https://www.linuxfromscratch.org/ or Gentoo?
I like Trisquel a lot because the package manager and packages are very very good.
Manjaro is a mess bloated with cr-apps that is barely usable for both enthusiasts and tech illiterates, it's far worse than Ubuntu, it would be more convenient to make it based off Gentoo or LFS and those would require time and effort compiling stuff from scratch but even then it would still require less time than removing GiBs of useless cr-apps and nonfree malware that all serve a same purpose and confuse/mistreat users.
Insightful video about "usable" corporate Linux-based distributions: https://yewtu.be/watch?v=mL9ztTzrY6Y
I do agree a rolling release should be available though, being able to pick between LTS and rolling is usually welcome on any distribution, but i don't think Manjaro would be a good choice for a rolling release. Also considering Parabola is a thing, i doubt it will ever happen, as long as other libre rolling release distributions exist there's no point on making other releases, the end goal is having free software the way the user wants, not wearing a distro flag of fanatism, let that behaviour for zoomers and children arguing on Reddit about first world nonsense like which nonfree videogame sucks and which doesn't.
Indeed, Manjaro is very bad. I just gave it a try, and immediately trashed it.
Briefly, it doesn't use Arch's repositories. Its own repositories are divided into three sections: stable, testing, and unstable. It tries to copy Debian's structure, but this results in a terrible mess.
I tried to switch to "unstable". The updating itself was smooth. But the system couldn't be properly logged out. When I forced a reset much later, the system was no longer bootable.
Actually, most derivatives of Arch don't solve the freedom issue of Arch: that its main repository doesn't reject non-free software. By contrast, many of them include additional non-free software to it.
I've never tested Manjaro for more than a couple hours. Couldn't find any redeeming value in it. It's basically the Ubuntu of Arch respins as far as I can tell, which is not a good thing. Very bloated and much too slow for my tastes.
If I wanted to run Arch, I would just run Parabola again. It might break fairly often, but at least it's light and fast when it does run.
The other Arch distro I really liked was Artix, which swaps out systemd for openrc, runit or s6. It was really fast and elegant. However, it would not run on the Linux-libre kernel from the AUR, so I stopped using it. Now that I've taught myself to compile the Linux-libre kernel "the Arch way" this past week for my experiments in libre-tizing Void Linux, I could go back and give Artix another try. Could be fun.
I tested several Arch-based distributions (Endeavour, Manjaro, Garuda, Artix, Arco, Archman, etc.), I agree that Artix is relatively better.
Of course, they all fail to solve the freedom issue inherited from Arch.
Yet another reason i just discovered to stay away from Manjaro in any regard, they apparently switched the default browser from one that could be libre such as Firefox, Midori or Ungoogled Chromium, to Vivaldi, a nonfree browser that makes big claims about privacy but is unable to prove them, because it's nonfree. We all know that vanilla Firefox is not very privacy friendly, but at least people can easily audit, and legally get rid of anti-privacy features with it, with shitty browsers like Vivaldi this is not possible.
https://vivaldi.com/blog/vivaldi-is-the-default-browser-on-manjaro-linux/
> We all know that vanilla Firefox is not very privacy friendly, but at least people can easily audit, and legally get rid of anti-privacy features with it
The "easily" part is debatable. Not easy to audit, and not easy to correct all of Firefox's privacy issues. And after you teach yourself to configure it for greater privacy, you'll find that Mozilla often breaks its configuration options in newer versions, making some of your privacy hacks no longer available.
Fortunately, abrowser is compiled with some of Firefox's privacy "features" left out. Rubin does a very good job with abrowser, and keeps it up to date.
There's no shortage of non-free software in Manjaro's repository.