Slashdot Mirror
Fedora Project Releases Fedora 24 Beta; Stable Version Comes Next Month (betanews.com)
A month ahead of its final release, Fedora Project on Tuesday released Fedora 24 beta for users and enthusiasts to try. An anonymous reader writes: The workstation version -- the one most home users will target -- offers GNOME 3.20 preview as a desktop environment. The GNOME environment has improved leaps and bounds over the years, becoming one of the best UIs of any operating system. Wayland is available as preview, but not default. The display server protocol is still poised to replace X, but it will not yet be ready for Fedora 24. The team explains that it should be ready for 'future versions'. Whether that means version 25 is something that remains to be seen."We're pleased to announce that Fedora 24, the latest version of the Fedora operating system, is now available in beta. The Fedora Project is a global community that works together to lead the advancement of free and open source software. As part of the community's mission the project delivers three editions, each one a free, Linux-based operating system tailored to meet specific use cases: Fedora 24 Cloud Beta, Fedora 24 Server Beta, and Fedora 24 Workstation Beta," said Matthew Miller, Fedora Project Leader.
For those of you wondering if the Beta is okay to use, I'll share that I've been running Fedora 24 Alpha since it was released at the end of March, and the Alpha has been stable for me. I'm looking forward to installing the Beta this weekend.
(My Linux system is a Thinkpad X1 Carbon, 1st gen.)
"one of the best UIs"
Yeah... that's why the first thing I do on my Fedora and RHEL installs is switch to Cinnamon. Yes, it is/was Gnome-based. But with all the crack-addled BS fixed, and everything that should work just does. No need to install a bunch of shell extension to get semi-sane behavior from a desktop.
the growth in cynicism and rebellion has not been without cause
Only remaining good window manager on any platform.
A bit pedantic, but it's a "Desktop Environment", not a "Window Manager".
KDE includes a window manager, KWin. I don't think anyone can really say it's the only remaining good window manager, but it's arguably one of the good ones.
You're a vocal minority that is shrinking. I know several Linux admins who have changed their minds about systemd once they started writing scripts for it.
Ubuntu, RHEL/CentOS, Fedora, Debian, openSUSE, Arch, and Mint all default to systemd, are planning to, or have it as the only option in the most recent versions. Gentoo maintains it as an option. Among major distros (for various definitions of "major") only Slackware seems not to have moved yet. One could call Amazon Linux AMI a major distro given its relatively widespread use, but IIRC it's based on RHEL6, so the next version could easily use systemd.
That list will grow as other software starts using systemd by default instead of an option. You can continue resisting systemd, but it will require a great deal of ongoing work to do so.
You can never go home again... but I guess you can shop there.
"one of the best UIs"
Yeah... that's why the first thing I do on my Fedora and RHEL installs is switch to Cinnamon. Yes, it is/was Gnome-based. But with all the crack-addled BS fixed, and everything that should work just does. No need to install a bunch of shell extension to get semi-sane behavior from a desktop.
Look into Budgie instead. Cinnamon is unsustainable. The Mint folks rarely make commits to it, mostly because they lack the expertise. Budgie is a more realistic alternative to GNOME with longer term ambitions. Its very usable right now also.
So, why aren't there any protesters outside Red Hat's office? Why do people keep coming to Red Hat Summit? Why do they keep switching to RHEL 7? Maybe you're actually wrong and that not that many people actually have a problem with systemd. I mean if it actually pissed of so many then maybe we would see more of that outside of Slashdot and the likes, you know out there in the real world.
Don't really want to stir the pot here but my 2 cents are this. People are protesting in their own way. I've seen a huge uptake in BSD lately. But you're right that RHEL is growing. I attribute that to a lot of people simply not knowing or caring about the issue.
I happen to like systemd. I do not like the developers. I find them to be childish amateurs with extremely poor people skills. I also have huge problems with Lennart trying to replace the core OS with systemd. He's given presentations to this effect. He mocks rather than debates and actively belittles people who disagree with him. This is in stark contrast to Linus Torvalds who will call something stupid based on its technical merits or demerits. Linus doesn't show up to other peoples presentations, actively disrupt them and belittle the speakers on a regular basis for example.
If systemd was being developed by anyone other then Lennart the controversy would not be an issue. Rather than get out ahead of this Red Hat has stood back and let their guy make a mockery of the open source community. Had Lennart been working where I work instead he would at best have been told to step back and stop speaking and at worst been giving his walking papers.
There are legitimate questions about systemd. Many distros have adopted it because it makes creating and maintaining a distro easier. No other reason. It adds nothing to security and very little to functionality that wasn't already being done elsewhere. That said, I don't actually care. If the systemd shit hits the fan I can replace it very quickly in my systems. I have other security controls in place to negate it (layered approach) and I rather enjoy using it (although I have issues with journalctl).
"one of the best UIs" - my favorite annoyance.
It is among the top 100% of interfaces available, ever. Even far into the future.
It's not a lie, it just means literally nothing. I frequently use it to backhandedly compliment someone's fave band or food.
"One of the better UIs" would mean it is better than at least one, so I use that instead.
This is such a mindboggling position for people to take. The entire concept of open source is about flexibility but people think it's fine to blindly force one option down everyone's throats, regardless of what they want. It's surreal to watch. We have 9,000 distributions but only One True Init, apparently.
I use Fedora and CentOS 7, so I'm using systemd on every system I touch now. And it's annoying, I certainly prefer the simplicify of the old init system. I'm really not sure what problem this was supposed to solve. Everyone talks about fast bootup times, but my servers uptimes are measured in years. Why should I add all this complexity to save a couple of seconds a year? I've spent hours reading about and learning systemd, to gain what exactly?
And I'm not saying no one should use systemd. If it solves some problem for you, great, use it. I completely support you. But why don't you also support my desire to use the init system that I want to use?
So do your own. The entire concept of open source is flexibility, and absolutely nothing is stopping you. Distros making decisions inherently remove some flexibility for the sake of delivering a functional platform.
Is it existentially annoying, in that "it's there, and it bothers me" sense or is there a tangible criticism you have against it?
Reduced resource usage, reduced system overhead, increased response time, increased manageability, etc.
Well, fast (parallel) bootup is one bonus, mostly for desktops, laptops, and embedded platforms. Maybe not for servers, because any with sufficient RAM will spend several minutes in the BIOS doing POST.
Use a distro that caters to your desires. Not necessarily needs, obviously, since you failed to make against systemd that isn't the same "It's there, and it bothers me" that has been howled for years now.
A lot of the things that make GNOME Shell suck are fixed in 3.20, but it's still rather agonizing to have to use the stupid activities page (clearly ripped off from OS X's launchpad...), and the lack of menu buttons is equally frustrating. Plus Nautilus still pales in comparison to Nemo and Dolphin.
Yeah, right. Doesn't sound suspicious at all. No sir, I'm totally going to take your word for it. Or maybe your "Linux admins" are converted Windows hacks who know as little about Linux as they did about Windows? That could explain it because something about systemd makes me think of svchost.exe more than anything else.
Perhaps you haven't noticed, but Red Hat, SUSE, and Canonical are all growing, and the usage share for RHEL/CentOS, SUSE, and Ubuntu have all risen as well. Where are all these hordes of systemd refugees that anonymous cowards on Slashdot keep telling me about?
The entire concept of open source is about flexibility
"flexibility" does not mean "the community will do your work for you." It means that you have the means and the right to modify the software that you use so that it works the way you need/want it to.
You still have those means and rights.
While your comment is true, the Unix/Linux philosophy grew from do one small thing well and interoperate. Even extra keystrokes were omitted in the interest of keeping things small and fast - rm, cp, mv, ls and the like come to mind.
systemd is getting its paws into everything. I too see a use case for some of its features - particularly for mobile users. But their ease of use has really made a mess of some interconnected servers. I'm sure that this will settle down as the code base gets more stable and quits changing, but it has had an impact particularly in trying to find out if an interface is up for some definition of up and waiting till that is true before starting other critical services.
As systemd impacts and worms its way into more and more packages, the ability to "write your own" becomes just about impossible for any small group of people. If any software package maintainer decides - hey - I could use systemd this way and makes the changes, that person understands the software and knows what to change for their package. Multiply that by hundreds of packages. From the other side, in order to undo the encroachment - you not only have to understand how to write an init system, but you have to understand every package that you need that has built a dependency on systemd and unroll all of the systemd related changes which may have occurred over months or at some point years and put your own hooks in without breaking it and then maintain those hundreds of packages. As it works its way into desktops, the number of packages is growing.
The get the source code and modify philosophy was really meant for fixing a problem or two before the end maintainers got around to fixing it or modifying a package or two and taking over those packages for internal operations. Duplicating the programmers of RH or the like really isn't in the cards for any normal person or group. Maybe it is for Mr. Pottering, but not for most everyone else.
A decade from now when systemd has all the bugs worked out and the system admins are happy again, somebody will decide to build a systemg and it will start all over. It would be nice to think that they will remember the systemd mess and do better, but then again they're probably in grade school now - or their parents or grandparents are. That's the sort of init history that got flushed away in the name of new and shiny.
So do your own. The entire concept of open source is flexibility, and absolutely nothing is stopping you. Distros making decisions inherently remove some flexibility for the sake of delivering a functional platform.
This is highly specious reasoning. systemd's developers' stated intent is to standardize (read: take over) all userspace initialization. They've accomplished (much of) this through a combination of embrace and extend, land grabs, intentionally breaking unapproved setups, tying functionality together unnecessarily or for highly trivial reasons, arm-wrestling other distributions with self-fulfilling prophecies, and generally being assholes. In fact, they've been doing many of the same things that Microsoft did in the late 90's with browser integration and which we all fought heavily against.
As a result of this, though, and partially as a specific result of their policy banning the packaging (and even OPTIONAL packaging) of initscripts at distro level in Fedora (the subject of this article), that it's become more and more difficult to revert out of it. Nothing may be *stopping* us (hence forks), but it's being intentionally made more and more difficult by a complete ass. Don't pee on my head and tell me it's raining.
Hire a Linux system administrator, systems engineer,