SoylentNews is people
SoylentNews is powered by your submissions, so send in your scoop. Only 15 submissions in the queue.
SoylentNews
Debian 8 "Jessie" was released on 25 Apr. A link to the Debian release page shows the changes and you can follow the release in 'real-time' should you desire to do so.
This release will be supported for 5 years and includes "improvements" to the UEFI software (both 32- and 64-bit) introduced in the previous version, "Wheezy". It also is the first release to use systemd as default init system replacing the earlier sysvinit, which is still available in the repos should you wish to revert the change. What effects such a change might have on the remainder of the system is not clear. Improvements to the support of Debian software include the ability to browse and search all source code distributed in the latest release.
This discussion has been archived.
No new comments can be posted.
Follow the Debian 8 "Jessie" Release Activity in Realtime
|
Log In/Create an Account
| Top
| 140 comments
| Search Discussion
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
We lie loudest when we lie to ourselves.
-- Eric Hoffer
(Score: 2) by kaszz on Sunday April 26 2015, @09:41PM
Makes me wonder. What is it that make the BSDs to resist systemd and other things like it. And for how long will they resist such infiltration?
(Score: 5, Informative) by K_benzoate on Sunday April 26 2015, @10:08PM
It's more that systemd isn't compatible with the BSDs, not the other way around. Systemd is, practically by design, unportable to other Unix-likes. It's Linux specific. It's designed to serve Red Hat's needs to create a complete Linux standard that they can package, sell, and contract support for. Everything about systemd/btrfs/gnome over the last few years has been geared toward making Red Hat's job easier, without regard to what the larger community wants or needs. Everything about systemd makes sense when you realize that Red Hat doesn't view its product as "a Linux distro" but as their own proprietary operating system that they want to control and profit from. They're taking the Linux kernel and creating their own operating system, and the tools they pull in from the community then become wedded to their own standards because they have the dollars to pay the programmers to make the code fit their needs. Tight integration is good for Red Hat, and bad for the rest of the community which needs modularity and options.
Red Hat wants a full operating system stack that they can sell and offer support contracts for. That's why systemd seems to consume everything it touches. They need in-house tools for an entire OS so they can do quality assurance for their customers (mostly the DoD and other large government agencies).
Red Hat is the Microsoft of the Linux world.
Climate change is real and primarily caused by human activity.
(Score: 3, Interesting) by kaszz on Sunday April 26 2015, @10:35PM
So we need to shitlist Red Hat?
(Score: 0) by Anonymous Coward on Monday April 27 2015, @03:36AM
It wasn't already?
(Score: 3, Informative) by Marand on Monday April 27 2015, @03:39AM
So we need to shitlist Red Hat?
Or at the very least, be extremely wary of anything they introduce, because it's entirely for their own benefit, even at the expense of everybody else. RedHat backs GNOME, employs GNOME devs, employs Poettering, etc. They're all known for a heavy "not invented here" mindset, refusal to adopt existing designs and instead preferring to build their own. They also like tying all their disparate parts together and presenting it with a "take it or leave it" attitude.
It works because most people want to avoid confrontations and splits and the like, so they give up and either scrap their work or do extra work to make their code interoperate with the RH/GNOME bits (because RH and GNOME won't do that work). It's not just the big pieces like PulseAudio, NetworkManager, systemd-init, either; they also ignore things like existing notification and systray work (such as done by KDE) to instead create their own "standard" for others to use, along with doing crap like refusing to play nice with non-GNOME apps in regard to window decorations and widget themes, resulting in fucked up situations like the KDE devs creating Qt and Gtk themes to make apps from either toolkit act native in KDE or GNOME.
Hell, GNOME itself only exists because KDE came first and contrarian folk decided "we don't like your license so fuck you we'll make our own" and then stuck with it long after the licensing became a non-issue.
The goal now, which has been stated in the past[1], is to have Oracle-style control of the entire stack, top to bottom. They want GNOME to be the OS, and anything that isn't GNOME or GNOME-created is an obstacle to that. Conveniently enough, the "systemd cabal" (as Poettering called it), wants to obsolete the idea of a Linux "distribution" [0pointer.net], which would conveniently leave RedHat the gatekeeper of modern Linux.
That should be reason enough to want to shitlist RedHat anywhere possible. If I liked Redhat, I wouldn't be using Debian and its offspring distros.
[1] It was in some GNOME dev presentation a few years back. I tried finding it again to link but I can't think of specific enough search terms to filter out unrelated junk and it's not worth spending more time searching than I already have.
(Score: 2) by kaszz on Monday April 27 2015, @07:08AM
Perhaps it's time for the community to screw around with everything Red Hat makes such that it makes corporate life hard. Refuse to accept their APIs rip their GPL code and implement in other ways than they thought of etc. New kernel feature? then fix it up and then make Red Hat to accept or leave it etc.
Someone else have a better idea?
(Score: 4, Funny) by hash14 on Sunday April 26 2015, @11:12PM
Red Hat is the Microsoft of the Linux world.
Going forward, I think the most appropriate name for them would be Redmond Hat.
(Score: 1, Insightful) by Anonymous Coward on Sunday April 26 2015, @10:49PM
The BSD developers are generally Generation X'ers, with some even being Baby Boomers. They aren't Millennials (aka, hipsters). That's the big difference.
Maybe it's due to their age or the naivety they were raised with, but Millennials make some really fucking stupid decisions when it comes to software. Ruby on Rails, JavaScript, NoSQL and systemd are superb examples of this. Developers from earlier generations aren't dumb in the same way. They don't make obvious mistakes so readily.
(Score: 2) by Marand on Monday April 27 2015, @03:44AM
The BSD developers are generally Generation X'ers, with some even being Baby Boomers. They aren't Millennials (aka, hipsters). That's the big difference.
Maybe it's due to their age or the naivety they were raised with, but Millennials make some really fucking stupid decisions when it comes to software. Ruby on Rails, JavaScript, NoSQL and systemd are superb examples of this. Developers from earlier generations aren't dumb in the same way. They don't make obvious mistakes so readily.
You might want to rethink your stance on this: Lennart Poettering, at 34, is considered "Generation X" and is responsible for systemd, PulseAudio, and Avahi (among other things). Lumping his decisions into the millennials group is an insult to the millennials.
Bad decisions happen at all ages.
(Score: 0) by Anonymous Coward on Monday April 27 2015, @05:44PM
No, somebody who is 34 would be a millennial, not a gen x.
Baby boomers were born between 1945 and 1960. Gen x were born between 1960 and 1975. Millennials were born between 1975 and 1990. They're called millennials because their formative years (10 to 25) were around the turn of the millennium. Those born between 1990 and 2005 are gen z. Those born after 2005 are commonly called recessionistas, due to living most or all of their lives during a global economic downturn.
(Score: 2) by Marand on Tuesday April 28 2015, @02:41AM
No, somebody who is 34 would be a millennial, not a gen x.
Baby boomers were born between 1945 and 1960. Gen x were born between 1960 and 1975. Millennials were born between 1975 and 1990. They're called millennials because their formative years (10 to 25) were around the turn of the millennium. Those born between 1990 and 2005 are gen z. Those born after 2005 are commonly called recessionistas, due to living most or all of their lives during a global economic downturn.
Source for this definition of GenX? and Baby Boomer and Millennial actually...
Wikipedia mentions multiple age ranges that have been used to define GenX and they all include the early 80s.
In a 2012 article for the Joint Center for Housing Studies of Harvard University, George Masnick wrote that the "Census counted 82.1 million" Gen Xers in the U.S. The Harvard Center uses 1965 to 1984 to define Gen X so that Boomers, Xers and Millennials "cover equal 20-year age spans".
Jon Miller at the Longitudinal Study of American Youth at the University of Michigan wrote that "Generation X refers to adults born between 1961 and 1981"
In 2011 "The Generation X Report" (based on annual surveys used in the Longitudinal Study of today's adults) found Gen Xers, defined in the report as people born between 1961 and 1981
Poettering, born in 1980, counts as GenX in every definition except yours, conveniently.
Related: there's also some variation in Millennials, but the definitions for it mostly stick to 1981-1982ish for the beginning and ending around 2000-2004, and "baby boomer" is mentioned as being 1946 to 1964. Every label you defined has been off by 4-10 years.
In my opinion, using your year of birth as a catch-all for negative stereotyping is lazy and ignorant. When the boomers did it to later generations it was considered ignorant, but now that genX is getting older, many are doing it to the next generation down. It's just as dumb now as it was then, even if you're using widely-accepted definitions (which you aren't even doing).
Point is, if you can't find a better foundation for your argument that BSD development is superior than lazy stereotyping based on blaming millennials (that you can't even define properly), then there's probably something wrong with the argument itself. Not saying that you're wrong about BSD devs being superior and the tendency for dumb design decisions elsewhere, just that your reasoning is faulty and you should look for a better answer.
(Score: 0) by Anonymous Coward on Tuesday April 28 2015, @03:49AM
Poettering looks like he's 12.
He should date 12 yr old girls.
(Score: 3, Interesting) by hash14 on Sunday April 26 2015, @11:40PM
In the software world, you generally live on a spectrum of two extremes: on one side, you can do it quick. On the other, you can do it right. Though I should note before going forward that the biggest reason I'm still on Gentoo instead of FreeBSD is because of KVM - from what I've heard, bhyve isn't quite up to par yet, though if anyone has experience with this, please let me know.
I think the heads of the BSDs tend to go more for the latter side. Their products don't have quite as many features or the same degree of performance as Linux, but they are stable and well-designed as hell. The fact that the FBI has NDAs with companies to backdoor OpenBSD [arstechnica.com] suggests that OpenBSD is probably doing things right. The fact that no such issue has come forward in the Linux development community might suggest that they really don't need backdoors to hack it.
On the other hand, Red Hat, driven by their publicly traded corporation-mindset, has lately been sounding off on how they need to support all their customers' wishes and desires. Even worse, they talk like it's the kernel's obligation to provide and maintain these services, rather than their own downstream packages (see the debug fiasco and the kdbus merge which hasn't been going very well either). Their software is increasingly half-baked as they race to get into more fields like cloud services, IaaS, etc. From the outside looking in, it seems that it's a political shithole where the name of the game is to get your product into the market and make money (and to hell whether it's actually good or not).
In Red Hat's world, it's all about the money (which is why they act and sound so much like MS lately). In the BSD world, it's more about the software and the product. So that's where I think the distinction lies.
(Score: 2) by fritsd on Monday April 27 2015, @12:14PM
The way I understand it, is that systemd depends on a Linux-specific kernel feature called "cgroups", a compartmentalization(sp?) of processes.
The BSD's don't have this, so they can't have systemd.
Apparently,systemd uses the cgroups feature to determine which running processes are descended from an init service, so that killing or restarting a service becomes much more accurate (if nothing is alive anymore in the cgroup of that service, then it probably is time to restart it, if it is so configured). DISCLAIMER: I have no practical experience with this.
(Score: 2) by kaszz on Monday April 27 2015, @10:36PM
There's something called jail(8) instead. Perhaps not good enough for systemd usage.
New idea for a license.. "This code cannot be run on a CPU that has been running systemd the last five minutes" ;)
Or "if( sys.systemd ==1 && random(10) >8 ) { panic("unexplainable panic happened!"); }" :D