Born from controversy, in the fiery hearts of rebellious Debian users and developers, who could no longer stand for systemd's takeover of the Linux world, Devuan came to be. In other words, it sounds a lot like a fad. So what is it really like?
The Debian derivatives series takes a closer look at some of the best known, and some less-known Debian derivatives, to find out what they have to offer, how they try to be more, better, or simply different from one of the most popular base distribution in the history of Linux.
To systemd or not to systemd...
The controversy itself had started as a flame-war, just around the time when systemd was beginning to gain serious traction. A lot of mud was hurled in both directions, with accusations, false narratives, and all the trimmings your usual internet comment war involves, often fuelled by the blaze of youthful knowing-all-better. Debian was no exception, an exodus soon followed.
Without taking sides, it's fair to say that systemd has both advantages and disadvantages. Beyond the tumult of the many spoken/written/screamed/whispered words that came from either side, there are legitimate points raised both pro and contra. Naturally one can only stay out of a debate gone too far by not even going to discuss these, but that is not the point either. Devuan developers thought they can do better and, unlike a lot of vocal systemd critics, they actually did something about it.
Let's see how they fared.
Init freedom is the word used on the tastefully designed website, which pretty much sums up what Devuan is trying to achieve. If it sounds like a lofty goal, it's because it is. They are here to save you from the evils of systemd! And how are they going to do it? By pushing sysvinit back down our throats. Although, and this is good news, alternative init systems are also promised in the long-term. How long, is unfortunately not known yet.
To substantiate their claims that everybody at Debian is wrong in their decision to adopt systemd, they link to a suckless.org page that does not even try to appear to be objective. While of course the Debian's systemd debate article cannot be accused of true objectivity either, it is at least somewhat better rounded, and less loaded in its language. Also, did you notice the little word "debate" in the title? That implies a common decision.
In short: Devuan's promise is to heal the pride of people who were voted down in a regular debate. Shaky foundations to say the least. Basically, Devuan is a distro based around an init system, or at the moment it still is. All other distros seem to do things the other way around. But does it work?
Of course, it does. It's based on Debian Jessie, which was rock solid even on its day of release, and now several years and point releases later, anything based on it will "suffer" the same fate. This also means that Devuan's currently only stable release is considerably behind Debian stable. While theoretically it might just make it something like "super-stable", in reality, it rather makes things just dated. Like Linux kernel 3.16 dated. You know how the usual criticism of Debian stable is being dated? Devuan manages to add to that. (Did I mention it was dated?)
Apart from the fact that your grandfather might feel young while using it, Devuan works like a charm.
The grub screen looks nice, and the Jessie installer invokes a subtle sense of nostalgia. Installation is just like that of Debian, minus the fact that there is no GNOME, Cinnamon or KDE available. You know, freedom comes with a compromise of not being free to use what you like, because they might be too tightly integrated.
Not that GNOME's decision to force its users to use systemd would deserve a cookie, and for that GNOME will not be sorely missed. In fact, GNOME's design decisions might have done more to keep the flame-war alive than sydtemd itself, but that it a different discussion.
Devuan aims to be slim and simple. While that's reflected by the Display Manager of choice (slim), the installed system does not feel "minimalistic" at all. In fact, XFCE comes with the full package, with all bells and whistles installed, although the additional software selection does, in fact, adhere to a much simpler baseline.
Apart from that, everything worked as it would have for Jessie. That was, right until restarting the Virtual Machine in which Devuan was installed, at which point an EFI propmt was all that seemed reachable. Surely this glitch could be fixed with a little tweaking, more time, and especially more willingness to look into the problem, but a distribution that calls itself "stable" might consider it a premium to allow such errors to creep into the release.
So what else does it do?
Surprisingly, quite a lot. You would expect the infrastructure of a distro that was born out of hurt pride to remain niche, but that is not quite the case with Devuan. Even the first impression upon visiting the website is that of professionalism, and the distro does not disappoint.
There are serious-looking mirror and developer lists, and as the footer of the site reveals Dyne.org's hand to be behind all this. In other words, Devuan is receiving genuine patronage.
They already have a number of derivative distros, although none of them is really worth mentioning as of yet. They do make this easy, a quite useful distro-kit allows everybody and their grandmothers to simply re-brand Devaun and call it a distro. Whether this further fragmentation of an already fragmented landscape is good, or even necessary, is up to everyone to decide, but surely, choice is better than the lack of it.
So why no love?
It's not exactly that simple, is it? While Devuan continues to rely on the hard work of the massive Debian developer community, and will always require loads of work to make new releases after any new Debian release. So basically Devuan will always be rather outdated. Even though the original plan was to "be ready by the time Jessie is ready", Devuan 1.0 (the first stable version) was released nearly two years(!!) after Debian 8.0. At the time of writing, Debian Stretch is at its third point release, while Devuan's next iteration is still nowhere to be seen. Is this sustainable at all? Time will tell. Does it make sense? Weeelll... it might, for some.
Basically, the initial idea of Devuan was to reap all the benefits of Debian's development process, including its democratic decision-making, then badmouth a democratic decision, calling it names, and pretend to know better. If this all sounds a lot like an utterly ungrateful, self-important brat treating his fair, albeit probably strict parents like crap... well...
The "Veteran Unix Admins" group, who first announced Devuan was never before heard of, but the fact they call themselves "veteran" indicates that the average age might not be all that high. Couple that with calling their move an "exodus", and generally using such pompous language in their initial announcements... and you'll see. Food for thought: Calling one's flounce an "exodus" alone speaks volumes of one's sense of self-worth, regardless of how many similar minded followers they might gather in the process, or the hype they manage to generate.
Fortunately though, there were signs of reasonable thought even in the original announcement:
The Devuan distribution will make an effort to improve its relationship with both upstream and downstream and, particularly in its gestational phase, will do its best to accommodate needs of those downstream distributions willing to adopt it as base.
and as of now, the official website states that
As of today the core of the vision remains the same: Devuan should be a minimal, stable system which honors Debian history and embraces innovation while maintaining backwards compatibility and interoperability.
Which at least sounds like some maturity is making its way into the minds of the project's maintainers, possibly as a result of some real "veterans" having joined them. This, and the fact that they do intend to explore other options for an init system might be a sign of creeping maturity.
Offering alternative init systems would actually make sense if they want to keep talking about "init freedom" because as of today, they are (quite ironically) just forcing us to use their own most favoured init. So if the term "init freedom" sounds like another exaggeration, that because it is. Init freedom is not a thing, not outside of Devuan developers' vocabulary anyway. The wording implies it has something to do with real ideas such as "software freedom", but that is not the case. It is, for a lack of a better word, basically just bragging.
Should you switch then?
It depends, of course. Do you absolutely need sysvinit and a seriously dated system? Like does your life depend on it? Because being on the losing side of a flame-war (see note before you flounce) might not be worth the compromise of an outdated system. After all, Devuan did not make it to the top 10 of Distrowatch's popularity page in the last year. Or the top 20. Or to top 30. Or even the top 40. There must be a reason...
"Losing side" is meant in absolute terms here: systemd had been widely adopted now, whatever the "opposition" would like to see.
But if you are so much against systemd that you'd rather die than to use the product of a developer who is fashionable to hate (because let's be honest, much of the hatred toward his software is meant for the person who developed it), there are systems that probably do better. Go Arch, or Gentoo. They might even make you feel like a power-user.
But if you are a happy-snappy Debian Oldtsable user, who does not mind the uncertainty, and loathes systemd for either good reasons (like real reasons which do in fact exist), or a perceived popularity of the statement, Devuan might be the absolute best choice of distro for you.
Meanwhile, Debian had seen its second stable release based on systemd, and somehow the horror stories about it still fail to occupy the news-feed. It's probably best decide for yourself.
Although there might be light at the end of the tunnel. If the promised alternative init systems ever make it to Devuan, it might as well begin to make sense. If its developers can ship any important security patches on time, the oldstable basis might mean really unparalleled stability, but for that to be appealing, they should make it so, by offering a real alternative. For that, they need considerable resources and loads of developers that can commit loads of working hours. To achieve that, it might be worth to further calm the confrontational attitude and begin to base Devuan onto something more substantial than an argument around an init system, and that does not mean inventing lofty phrases like "init freedom" that have neither weight nor substance. Until such time, the measured and time-tested ways upstream had long been famous for, will continue to rule the landscape.
I absolutely have to add this note here: I am not some pro-systemd troll. I dislike about it just as many things as I do like. What I tolerate less than systemd's flaws is the pointless flame-wars, finger-pointing, and copy/pasted "reasoning" of people who react out of herd mentality. That irritates the hell out of me.
So, if you are planning on doing just that in the comments for whatever reason, please know that the comments are, and will remain moderated (on this page only, as of the time of writing). So keep it respectful, and original. Any irony and even sarcasm are OK, just don't go personal, and please for gods' sake, stop simply repeating other people's opinions.
Liked what you've read? Sharing this article on your favourite social medium helps a lot with discoverability. You know, sharing is caring.
Got something to add? Comment is free, so please leave your thoughts below, and don't forget to "like"/recommend it on Disqus.