SoylentNews is people
SoylentNews
Despite the previous announcement at the Ubuntu Wiki, that said
Martin Pitt announces the date to switch Ubuntu Vivid to boot with systemd instead of upstart as Monday, March 9th. He says that the switch will affect the desktop, server, cloud, all flavors but not Ubuntu Touch, and that if there are too many regressions there is a simple upload to revert to upstart.
regarding Lubuntu, the Ubuntu Wiki now reports
LXQt is still in development, so Vivid Vervet [(*buntu 15.04, slated for release in April)] is another bug fix release. A late regression in the desktop installer for 32 bit means there is no Desktop installer for this milestone but it does not affect the alternate installer. Systemd is not the default init system.
LXQt is a light-ish desktop environment built with the Qt toolkit usually associated with KDE-compatible apps.
This announcement indicates that, for the time being, Lubuntu will be sticking with LXDE (Lightweight X11 Desktop Environment) built with the GTK+ toolkit usually associated with GNOME-compatible apps.
Upstart will remain the init system for Lubuntu for now.
(Score: 5, Informative) by hemocyanin on Monday April 06 2015, @04:50PM
List of no-systemd distros:
http://without-systemd.org/wiki/index.php/Main_Page [without-systemd.org]
(Score: 1, Interesting) by Anonymous Coward on Monday April 06 2015, @05:05PM
List needs to be updated -- Doesn't include Kwort Linux: http://kwort.org/ [kwort.org]
(Score: 4, Informative) by kaszz on Monday April 06 2015, @05:14PM
Alternative for those on the Debian poison:
https://devuan.org/ [devuan.org]
(then there's always that software from some university in Berkeley ;-) )
(Score: 5, Interesting) by CRCulver on Monday April 06 2015, @07:03PM
I don't think Devuan is ever going to pick up steam. They made some bad decisions early in the project that might have crippled them for good: an ugly website, a name that unfortunately evokes derision among North American users, and they continued to release English-language publicity in ungrammatical or unidiomatic English even as native English speakers were offering to help. Here on SN some users have responded to those criticisms by saying that only the quality of the code matters, not the project's presentation, but sadly in the real world we live in, a presentation that doesn't draw such scorn does matter for getting a critical mass.
I'll probably upgrade a couple of my Debian boxes to Jessie, just trying to keep sysvinit on there [skolelinux.org] instead of moving to systemd. However, I installed Slackware on an old netbook and on a Raspberry Pi I had laying around, and I've really enjoyed the experience. It has been a pleasant trip down memory lane, as so much of what's in a default Slackware install are *nix utilities that haven't changed in decades, and it reminds me of my first experiences with Linux around the turn of the millennium (though all the "modern Linux" stuff can still be installded if you want it). However, I find the Unix philosophy perennially attract, so I've found no problems using a Slackware distro to get stuff done in a satisfying way.
Why does Slackware have that attraction for this systemd refugee when Devuan doesn't? Perhaps the secret is that Slackware has a minimal presentation, really lettting the software in the distro speak for itself. With his quiet, mysterious air, Patrick Volkerding has never put himself out in the limelight enough to risk his own and his project's reputation like the Devuan folks.
(Score: 2) by CRCulver on Monday April 06 2015, @07:06PM
(Score: 5, Interesting) by kaszz on Monday April 06 2015, @08:02PM
Declining help seems like a really bad decision when it's that obvious. Is there any other specific decisions that may have destroyed the project?
What alternatives to Debian and Devuan are there?
(the question one might ask is where the systemd buck stops - preferably the hard way)
(Score: 2, Insightful) by Anonymous Coward on Tuesday April 07 2015, @12:09AM
Here on SN some users have responded to those criticisms by saying that only the quality of the code matters, not the project's presentation, but sadly in the real world we live in, a presentation that doesn't draw such scorn does matter for getting a critical mass.
A working distribution will attract people who can work on the presentation. Distracting developers with presentation or just organizing the presentation effort just delays the point where they will have a healthy pool of people who can work on presentation. You do not paint the walls before the foundation, frame, floor, or ceiling has been put in place, would you? You might if you were a Silicon Valley startup, but there are many reasons why a large percentage of those companies fail. Making the place look good before having a working product is one of those reasons.
(Score: 5, Informative) by NCommander on Tuesday April 07 2015, @06:37AM
When Devuan got started, I dropped into their channel on Freedom to see what I could to do to help. I do not look forward to having to migrate my systems (or the SN servers) from Ubuntu, and if I had an easy option to remove systemd from the stack, I would take it. As such, I had both a professional and personal interest in the project. For those unaware, I'm both an Ubuntu Core Developer, and a Debian Developer, and am *well familiar* with the underlying infrastructure that underpins both distributions (I've committed a fair bit of code to the Debian Archive Kit which is the magic piece of software that underlies the Debian archive as well worked on various Debian ports and spinoffs).
While I understand the project leaders mean the best, they've chosen to essentially throw out everything related to Debian and go to entirely git based system, only the binary package format (deb) is essentially being retained. For those who don't quite get what I'm saying, as in every piece of code, every package, and solution will be built with git and use a jenkins backend. Those who are lead developers seem to have no knowledge of Debian internals, and are making mistakes that will make trouble down the line. For those who aren't involved with Debian, there is an extensive test that has to be completed that makes sure you understand most of Debian policy, why that policy exists, as well as the fundamental internals of packages and such.
I tried heavily to explain why this was a bad idea, as explaining from personal experience that jenkins will not scale properly to a project of this size or handle many of the fine details that go into package building such as dependency handling, and that reinventing the wheel here is a very bad idea, as one isn't careful, you will cause numberous issues with ABI breaks, soversioning, and drastically increasing the difficulty of collobration with Debian and another Debian based distros. In the end, I gave up. I do hope they success despite my misgivings, but IMHO, I do not have high hopes.
Still always moving
(Score: 2) by digitalaudiorock on Tuesday April 07 2015, @08:17PM
Sounds like a really odd decision. I mean really, when it comes down to it, most of the code is upstream, and a distribution pretty is it's package management system for the most part. Seems a bit like taking off the gas cap and driving a new car under it ;).
(Score: 2) by kaszz on Wednesday April 08 2015, @01:15AM
What do you consider a good alternative to Debian that is systemD free?
Seems weird that it's so hard to just fork it like people did with XFree86 when it went sour and went on with x.org.
I'm really open to suggestions.
(Score: 2) by NCommander on Wednesday April 08 2015, @06:28AM
If you want to stay with Linux with a major distro, IMHO, Slackware, though even Patrick has said its not out of the question. Gentoo also supports their own system if you don't mind a rolling release.
The problem is Red Hat basically controls the vast majority of the desktop stack through GNOME developers, and the lower stacks via maintenance of PAM and such. They can force through basically any stack change they want above the desktop. In all seriousness, I'm looking at migrating to a *BSD for everything except my desktop; I do enough gaming that it would be painful to migrate the desktop.
Still always moving
(Score: 2) by kaszz on Wednesday April 08 2015, @09:33AM
Time to fork GNOME and kick RedHat out of the ecosystem?
Not out of question that Slackware will adopt systemd?
Won't all this systemD adaption affect *BSD too?
(Score: 0) by Anonymous Coward on Wednesday April 08 2015, @10:35AM
Or just abandon Gnome fully as a lost cause and adopt XFCE.
Nof if only Gnome didn't have GTK by the balls...
(Score: 2) by NCommander on Wednesday April 08 2015, @02:25PM
GNOME has been dead to me since 3.0. I actually use (and like) Unity on the desktop, and have used it now for a few years. The systemd API dependence will affect the BSDs, but the OpenBSD project is working on creating a systemd "proxy" so to speak to let it be emulated. Honestly, if it was just coded to POSIX vs. the Linux API, it wouldn't be an issue; I find it rather insane that my system will fail to boot if I compile a kernel without CONFIG_CGROUPS.
(/sbin/init or equivelent should basically work on any generic kernel. Even if I boot with an antique kernel which my udev does NOT support, at least I can still reasonably get to bash).
Unfortunately, with it looking like udev will soon start depending on systemd, I'm guessing avoidance will become impossible, hence why I'm looking at non-Linux choices. Unfortunately, Linux just recently starting getting SOME momentium with Steam being ported over and such, and I dislike having to change because of such a horridibly designed bit of code acting like a cancer.
Still always moving
(Score: 2) by kaszz on Wednesday April 08 2015, @07:41PM
Perhaps it so that Linux has got too many developers that don't get the Unix philosophy with regard to orthogonal interfaces (API), modularity and safe approaches ie not run on assumptions etc? Otoh, there's always been a different paradigm in programming philosophy between Linux and the Berkeley sphere. When it works it's alright, when it's right and works it's alright.
Perhaps there's been a balance shift between people with a science background vs autodidacts or something alike?
This whole circus smells Psyops with systemD, derailed projects, the need to engage people in pseudo problems (systemD compatibility, Gnome replacement), .. There are players that has an interest in derailing FOSS. Security organizations that can get rid of a platform outside corporate control and with security, Corporations like MS to eliminate the competition, Some companies may want to co-opt projects for gain etc. If so, abandoning Linux will only allow the cancer to grow and take on the next sphere.
(Score: 4, Informative) by Subsentient on Monday April 06 2015, @09:19PM
I just might want to point out that there are other init systems [universe2.us] out there. Like the undying glory of the Epoch Init System.
"It is no measure of health to be well adjusted to a profoundly sick society." -Jiddu Krishnamurti
(Score: 0) by Anonymous Coward on Friday April 10 2015, @09:11PM
At this point in time it is flat out wrong to think about systemd just in terms of init. It has become a whole host of interlocking processes that handling anything from task scheduling (think cron) to networking (firewall included).
In essence it reeks of Embrace, Extend, Extinguish.
(Score: 2, Interesting) by Anonymous Coward on Monday April 06 2015, @11:42PM
The choice to stay with LXDE makes sense to me, it has for quiet a while been my go to DE. PCmanfm is a good file manager and the whole of LXDE runs nicely on aging hardware. (set up a perfectly smooth system with debian 7 and LXDE on an Athlon64 3000+ with 512MiB DDR2) .
I have been playing with LXQt on my Arch box and it is just not quite there, I am liking how it's looking though but there is a shortage of Qt apps that don't pull in the whole of KDE, and some nice features are missing (auto hide panel).
That said it has lots of promise and i prefer Qt to gtkmm. I will argue the C is a bad language for GUI stuff as objects just make sense for a GUI. I can't wait for LXQt to really get to the the point where it's feature equivalent to LXDE, when that happens it will be my goto DE.
(Score: 2) by NCommander on Tuesday April 07 2015, @06:40AM
I won't be surprised if a lot of the decision to go to systemd deals with the fact that Unity still uses a large number of GNOME components, which in turn depend on systemd. Using upstart was reasonable as long as the vast majority of the stack didn't break w/o systemd, but without Debian making sure the core components are systemd-free, the delta between Debian and Ubuntu would drastically increase. LXDE, which, to my knowledge, does not ship with GNOME based components would not have this underlying issue.
Still always moving
(Score: 0) by Anonymous Coward on Tuesday April 07 2015, @10:45AM
Weird thing is that supposedly Canonical was poised to take over development of Consolekit, that used to do what Logind (the main link between Gnome and Systemd at this point) does now, but for some reason it never happened.
But then there is all kind of weird passive-aggressive stuff regarding the move from Consolekit to Logind. Not helping that Poettering was the de-facto maintainer of Consolekit towards the end.
What is really striking tough is that apparently the major reason for discontinuing Consolekit was that they could not get multi-seat to work to their satisfaction. And multi-seat basically means having multiple displays and input devices hooked up to a single computer, that is then logically bundled into a "seat". In essence it recreates terminals via different means. And they are trying to get it working "securely" while still allowing for hot-plugging of devices.
Their idea with Logind seems to be to use cgroups, via Systemd, to limit what device files a seat can see, never mind interact with.
All in all the whole thing is starting to look like netsec erotica...