Slashdot Mirror
Debian Forked Over Systemd
jaromil writes: The so called "Veteran Unix Admin" collective has announced that the fork of Debian will proceed as a result of the recent systemd controversy. The reasons put forward are not just technical; included is a letter of endorsement by Debian Developer Roger Leigh mentioning that "people rely on Debian for their jobs and businesses, their research and their hobbies. It's not a playground for such radical experimentation." The fork is called "Devuan," pronounced "DevOne." The official website has more information.
I think it's fair to say that this fork is far more significant.
I certainly wish them luck, but I am concerned that they may not be able to get the resources needed to successfully compete against the Redhat/Debian agenda.
More to the point, switching to FreeBSD/OpenBSD was always an option anyway. FreeBSD last forked at version 4/5 with DragonflyBSD coming out of it.
Forking the OS is something that should happen when idiotic choices are made, but popularity of a fork doesn't mean it's better. VHS vs Beta.
My personal opinion of the matter is that all OS's need to go back to the drawing board and redesign around multicore and the architecture of GPU's, as only Windows ever really cared about multithreading, and decades of single-threading stupidity has finally hit the wall because nobody knows how to do a multithreaded renderer for games. Likewise idiots at Google and Firefox still single-thread crap in the browser, so crappy javascript has no problem locking up the user interface.
I think it's fair to say that this fork is far more significant.
I think this fork will be fairly insignificant, and, further, that it will increasingly run into problems as desktops and other packages depend more and more on systemd components (that trend was one of the major factors in the Debian decision to adopt it).
I actually wish the Devuan guys all the best; I'd love to see another solid server-focused distro (server focus may help them avoid the issues with DEs). But I'm really glad to hear about this fork because the systemd debate has been a huge distraction to Debian. Hopefully this will finally put it to bed as all of the systemd opponents leave Debian for Devuan. I think that will be a net win for Debian because most of the vocal opponents don't contribute much code anyway.
Personally, the more I learn about systemd, the more I like the ideas behind it, and both code and documentation seem to be of high quality (documentation in particular is much better than is typical of open source projects). I'll be sticking with Debian.
Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
Pottering doesn't work for MSFT, he works for the 3 letter agencies. Considering that MSFT would probably be a step up on the trust scale. Where does Pottering get his money? Red Hat...okay so where does RH get THEIR money? NSA,DoD, FBI,CIA, DoJ, something like 85% of their income is from .Gov institutions, most in the Intelligence community. if the 3 letter agencies quit buying RHEL tomorrow that company would be on the ropes.
So call me paranoid but I can't really blame anybody for not wanting something controlled by a company that is so tied to the spooks, after Snowden its just common sense.
ACs don't waste your time replying, your posts are never seen by me.
Except there is plenty coming out of the RHEL customer base - we're being told to shut up. Also notice how many RHEL shops are not moving to RHEL 7 (us among them).
I'm starting to think GNU is the problem with "GNU/Linux" these days.
I disagree. Uselessd shows that systemd's parts are not as tightly coupled as people suppose. Just because they are all part of one umbrella project does not, in fact, mean they are tightly coupled and integrated in some sort of orwellian fashion. Uselessd proves this fact. And Uselessd is a good thing to have. Provides competition for systemd, provides a few features people want, and could pave the way for modern desktops like Gnome to run on non-linux systems such as BSD. Gnome isn't bent on having *the* "systemd" just the capabilities that systemd provides. If Uselessd can do it, so much the better.
Historically, Debian unstable was usually absolutely solid.
Except for all the major changes it goes through. The introduction of udev, change to 2.2 and 2.4 kernels, all of those broke for me (though switching to 3.0 was fine but I think that was more of a marketing move than a major version change). Then there's application level problems such as config utilities breaking things like Apache when it jumped a major version.
The system is labelled as unstable. If it's stable it is a bonus. What you *think* it should be is irrelevant, it is provided without any guarantee to be bug free, or even a guarantee that it will boot.
Um, what I think it should be is entirely relevant. I was primarily responsible for maintaining sysvinit and the initscripts from squeeze through to the wheezy release and after, doing the testing and providing the guarantee that it would boot. I was the one who did the testing before uploading. Different VMs, different upgrade scenarios, bare metal on different architectures, Linux, kFreeBSD and HURD kernels. If I'd screwed up, people would have had unbootable systems and come shouting. The quality bar was higher then and we did pretty thorough testing; I'd like to think we did a pretty good job. I certainly was never responsible for systems becoming unbootable on upgrade.
Some people dislike systemd because they can see where it is headed. Here is your sign.
"Somebody has to do something. It's just incredibly pathetic it has to be us."
--- Jerry Garcia
Then why aren't you hearing anything from the Red Hat customer base? If anyone wants reliability it's the enterprise which is Red Hat's entire market. The fact that nothing is coming from that side tells me that this is about something else entirely where people are more concerned about the political process and symbolism than the technical merits.
Maybe there is a big demand for a very stripped down low feature server distro, but I suspect this isn't going do become a big player.
We're a Red Hat shop. 500+ servers by current count with capacity expected to double in the next 6 months. Not one single system 7 server yet and the more official Red Hat training we get the less we like the changes in system 7. But you won't hear our complaints because our concerns go directly to Red Hat.
Let's be clear here for the 'people more concerned about desktop users', systemd has absolutely NOTHING to do with desktops. The servers we manage are headless boxes (all of them). Red Hat is pushing systemd because they need it to manage 'containers', the new cloud buzz word of the year. The Red Hat reps practically scream 'containers! containers! containers!' like another well know monkey dancer at each and every presentation. We in turn want to throw chairs out the windows on the way out. We don't give a damn about containers. Neither do our clients or developers. We're not a cloud provider and neither are desktop users.
Red Hat can't provision (sell) containers without systemd. It's as simple as that.
Now we're considering 'off standard' distros which may very well mean an end to new Red Hat 7 builds.
Ubuntu also moved to systemd because everyone was moving to systemd. Before that, Ubuntu has their own init system called Upstart, and there was much debate in Debian on whether to use systemd or Upstart.
It's my understanding that there was an attempt to affect the voting by limiting who had the ability to vote, simply because one of the lead developers was a prominent Upstart supporter. One interesting reference is here, though this is not the source I read about the vote manipulation from.
That said, I'm not overly familiar with how Debian elections are carried out. I only know what I came across in the last couple weeks when I was trying to get a grip on why major distributions were going so solidly with systemd, given issues that so many have found in the package. The trick to remember is that systemd is not the only solution to any {real|perceived} issues that sysvinit may have: There's also openrc and Upstart, to name two other alternatives, and they each have different solutions to bring to the table. Part of what made Linux what it is is the ability to choose what you want in your distro, to determine what you think is really "broken" and what the solution should be.
Honestly, I started getting migraines trying to wade through all the political crap. Proponents of systemd started to sound like American politicians (Democrat or Republican, take your pick; they both tell lies and break promises). It's mind-numbing, which I think is the point. I couldn't find a distro without systemd at all (this was a couple weeks ago, before I head of Devuan) so I wiped my Linux (Fedora) box and put FreeBSD on it.
Yeah, I'll have to learn how to deal with 'ports', but I won't have to deal with the nightmare that appears to be systemd.
When politicians are involved, everyone loses.