Ubuntu 16.10 Reaches End of Life (softpedia.com)

prisoninmate shares a report from Softpedia: Today, July 20, 2017, is the last day when the Ubuntu 16.10 (Yakkety Yak) was supported by Canonical as the operating system now reached end of life, and it will no longer receive security and software updates. Dubbed by Canonical and Ubuntu founder Mark Shuttleworth as the Yakkety Yak, Ubuntu 16.10 was launched on October 13, 2016, and it was a short-lived release that only received nine (9) months of support through kernel updates, bug fixes, and security patches for various components. Starting today, you should no longer use Ubuntu 16.10 (Yakkety Yak) on your personal computer, even if it's up-to-date. Why? Because, in time, it will become vulnerable to all sort of attacks as Canonical won't provide security and kernel updates for this release. Therefore, all users are urged to upgrade to Ubuntu 17.04 (Zesty Zapus) immediately using the instructions here.

Comments

non-remarkable non-LTS

[KiloByte]

Not sure why we would care -- it's just an old already-replaced short lived release. The release Ubuntu users should care about is 14.04 (supported until 2019-04) as it's the last one with a sane init.

Re:non-remarkable non-LTS

[Anonymous Coward]

I have been trying not to care about this init stuff. Through a series of upgrades I've ended up with one machine still stuck with upstart and another on systemd, and I didn't want to get involved in this discussion. I figured I'd just adapt to whatever.

Let's just say I have reached the point of caring. :-/

One small thing to start: how the fuck is it not the default behaviour of journalctl to linewrap so you can actually see all the errors?

I am no sysadmin. I've been using Linux for about 24 years, day-in, day-out, in one capacity or another (back from the 0.99pl12 days, stack of floppy disks, 486 with 8 megs of RAM and a tiny hard disk) and I'm still confident and happy saying I am no sysadmin. I accept my limits; I know I am a developer just *using* it with admittedly pretty significant day-to-day acquired knowledge, I'm not operating it as a sysadmin with studied expertise. I look like an expert to others; I don't feel like an expert.

So I expected some relearning and some frustration, but fuck me I didn't expect to feel patronised.

Re:non-remarkable non-LTS

[Anonymous Coward]

how the fuck is it not the default behaviour of journalctl to linewrap so you can actually see all the errors?

Why the fuck do you expect anything shat out by Lennart Poettering to be user friendly?

Re:non-remarkable non-LTS

[Anonymous Coward]

One small thing to start: how the fuck is it not the default behaviour of journalctl to linewrap so you can actually see all the errors?

Lennart Poettering should be hit on the head with an actual log until he recognises the importance of logging (real, or traditional).

Re:non-remarkable non-LTS

[GoingDown]

One small thing to start: how the fuck is it not the default behaviour of journalctl to linewrap so you can actually see all the errors?

It is definetely not journalctl's job to linewrap anything - your terminal should do that.

Re:non-remarkable non-LTS

[Anonymous Coward]

And it isn't journalctl's job to truncate log lines to the width of the terminal. Yet it does.

Re: non-remarkable non-LTS

[Zero__Kelvin]

The output is paged through less by default, and long lines are "truncated" to screen width. The hidden part can be viewed by using the left-arrow and right-arrow keys.. You could also use 'alias' to change the default behavior, or use other methods. As usual when people hate on systemd, you are complaining about systemd because you don't understand it and couldn't be bothered to learn what you could have learned with 2 minutes of googling. It literally would have taken less time to learn about it than complain about it.

Re: non-remarkable non-LTS

[Anonymous Coward]

Considering that the default behaviour for the past forty years, for all Unix systems, has been to print out lines unaltered, wrapping them when necessary, why the hell do users have to adjust to new behaviour? This is altering the system's behaviour, contrary to end user expectations, for no good reason that I am able to discern.

Two minutes of googling? Multiply that by however many thousands of sysadmins are out there and having to deal with this bullshit. Multiply that by however many times a sysadmin gets tripped up before baking it into a system image as a default, plus the number of times they get caught with a new release.

Arguing "you can change the behaviour back, quit whining" simply doesn't cut it in the context of systems that are managed by the thousands.

Re:non-remarkable non-LTS

[thegarbz]

One small thing to start: how the fuck is it not the default behaviour of journalctl to linewrap so you can actually see all the errors?

If you pipe journal control through anything like you would have done previously it defaults to linewrap.
Given how much more it actually displays per line the no linewrapping is a bonus. It makes it much easier to read.

Why do some distributions default to coloring ls and others not?
Why do some distributions provide a short hand for ls -l --color and others not?
Why is it that people get so upset about something when the new option is more configurable than the previous options. Add "export SYSTEMD_PAGER=less" to your environments and your incredible problems of personal preference will be a thing of the past.

Re:non-remarkable non-LTS

[thegarbz]

Journalctrl is not a grep of a dumb text file. It's job is to do whatever it was designed to do by the author.

Fortunately the author made it quite configurable. Just export SYSTEMD_PAGER=less, and journalctl will look 100% identical to your previous ways of working. Or just ignore journalctl and set it to output to syslog and it will actually be 100% identical to your previous way of working (with the addition of boot messages in the syslog).

Complaining about something more configurable that offers a complete compatibility with your own way of working looks childish.

Re: non-remarkable non-LTS

[thegarbz]

This! Not only can you change it through aliases but there are also specific environmental variables that can be set so journalctl adjusts its behaviour on a per use basis. The option the people who don't RTFM are looking for is "export SYSTEMD_PAGER=less"

Re: non-remarkable non-LTS

[thegarbz]

Considering that the default behaviour for the past forty years

Oh dear, someone moved your cheese.

Re: non-remarkable non-LTS

[Anne Thwacks]

This is altering the system's behaviour, contrary to end user expectations, for no good reason

which should be considered a crime against humanity (well, against Unix philosophy, anyway), and punished by people using another distribution (but nuking from high orbit is fine in my books).

Personally, I like the *BSDs.

Re:non-remarkable non-LTS

[Anonymous Coward]

Nice strawmen there. Nobody is upset about distro differences or more flexible configuration. People are upset about the default changing. Existing infrastructure breaks, with very little justification. The correct approach is to use the new and flexible configuration options to default to a configuration that acts identical to what is already there. People that want the new shiny can change the config, people that want to get things done do not need to stop and wipe up the mess someone else made.

Re:non-remarkable non-LTS

[Anne Thwacks]

Why is it that people get so upset about something when the new option is more configurable than the previous options.

NO, You just don't get it at all why should adding more configurability fuck with the well established defaults which zillions of lines of code and millions of dumb users totally depend on to get things done?

This is something that is in the same league as casually kicking people in the shins for no reason, or spilling other people's drinks in a public bar, and you can't be surprised if it leads to a bar brawl - it is the conventional way to start one.

Re:non-remarkable non-LTS

[butzwonker]

Maybe they should use sane defaults?

Re: non-remarkable non-LTS

[F.Ultra]

Since you executed the script it actually doesn't sound like systemd was involved at all.

Re:non-remarkable non-LTS

[thegarbz]

I agree. You should tell that to the package manager of your distribution.

Re:non-remarkable non-LTS

[MangoCats]

I suppose the 15.10 that I'm typing this on is even worse off, then?

Funny that it's in better shape than the 14.04 LTS running on another machine in my house - that 14.04 has patched itself into all kinds of system-error complaints, and seems to be more crash prone - maybe because it runs Kodi all the time, but I like to think most of the problems are in the updates. The last update to Kodi that changed the whole user interface was particularly annoying.

Re: non-remarkable non-LTS

[F.Ultra]

Tell that to your distribution because it's them that changed your default. Yes the journald default is this but since it's configurable the distribution choose to not change it to match their old behaviour.

Re: non-remarkable non-LTS

[Anonymous Coward]

And people wonder why real sysadmins hate systemd. (hint... douchebags like yourself.)

Re:non-remarkable non-LTS

[Anonymous Coward]

> Add "export SYSTEMD_PAGER=less" to your environments

Why doesn't it respect the PAGER environment variable? Why do we have to use a NEW variable in the SYSTEMD namespace?

I mean... seriously? That's been the way you specify your desired pager for _decades_.

> ...your incredible problems of personal preference will be a thing of the past.

Heh. Maybe PAGER is a thing of the past. We all know that Poettering and the Systemd Cabal say that everything from the past must be swallowed into the systemd project and namespace. :(

Re:non-remarkable non-LTS

[thegarbz]

NO, You just don't get it at all why should adding more configurability fuck with the well established defaults

Implying your retarded* well established defaults are the best way.

which zillions of lines of code

That was automatically handled via the backwards compatibility. If this change broke anything then the person shouldn't be coding let alone coding millions of lines.

and millions of dumb users totally depend

And "dumb" users are the reason we make changes and improve things rather than staying locked in some course because of ... reasons.

This is something that is in the same league as casually kicking people in the shins for no reason, or spilling other people's drinks in a public bar, and you can't be surprised if it leads to a bar brawl - it is the conventional way to start one.

If that's what you think then you should seek psychiatric help. Just because someone moved your cheese doesn't mean the world is ending, and unlike any of your examples they didn't do it *to you*, they did it *for everyone else*. Learn the difference. "Change for changes sake" is equally as bad as staying the course because "It's always been like that". But as usual in cases like this, people aren't interested in reasons, they just want it the way it always was and will seek the blood of anyone, regardless if that's the programmer who specifically added backwards compatibility options so the distribution maintainers can set the defaults the way their users expect.

*my opinion. I hate line-wrap.

Re: non-remarkable non-LTS

[Hognoxious]

So instead of users working around Lennart's shitty designs the distro maintainers should do it?

While I agree that's an inconvenience to fewer people and is therefore a slight improvement I can think of another step that would be even better. Can you?

Re:non-remarkable non-LTS

[GoingDown]

Yeah, I actually realized now what is the issue.

If you run
>journalctl | less

You will get normal less pager with wordwrap

If you just run
> journalctl

You will get still paginated output, without wordwrap (not sure what this is using as pager, is this something built in?).

Finally
>journalctl --no-pager

Will show you plain output without pager.

Yeah, it seems that defaults are little bit strange, and I do not understand why there must be that default pager (without wordwrap) at all. And I don't know where the pager is choosen from, I do not have either PAGER or SYSTEMD_PAGER set (this is Ubuntu). And actually, setting pager does not seem to help

Re: non-remarkable non-LTS

[drinkypoo]

It literally would have taken less time to learn about it than complain about it.

The default in basically everything is to show full lines. But systemd wants to be different. Why? Because Poettering thinks Unix was done all wrong, and he's smarter than everyone else, and changing default behaviors on the console to be more like a GUI is a great idea. That's because he's a fucking child. His development stopped at maybe age twelve. It's all what he wants and what he thinks, and everyone else is dumber than him, right? Except a review of his code proves that's false. He's actually a shit coder with shit ideas and you're defending them because you're a shit person.

Re: non-remarkable non-LTS

[Hognoxious]

Shittest book on anything ever.

But I must take my hat off to you for your time management skills. Most people would find shilling for the 1% or Lennart Poettering a full time job on its own (no apostrophe - is it that hard?) but you manage to combine both.

Re:non-remarkable non-LTS

[dbIII]

Careful - that's far more explicit than what Lennart screamed was a "death threat" a few years back.
He takes comments like yours seriously. It's apparently OK for him to describe a fantasy about people like you collecting bitcoins to pay a hitman to go after him (and for him to say it "really happened" - yeah right) but he's got too thin a skin to allow you a fantasy.

What I'm trying to say is "jokes" like that just reinforce his "us and them" attitude, where everyone who isn't his fanboy is apparently not worth listening to because they apparently want to be violent towards him.

Re:non-remarkable non-LTS

[Hognoxious]

Implying your retarded* well established defaults are the best way.

It implies nothing of the sort. However breaking expectations is rarely a good idea.

Is brake on the left & gas on the right intrinsically better than the other way round? No idea. But it's what people expect, and if you're going to change it then your awesome solution doesn't just need to be better - it needs to be vastly better.

See also: qwerty.

Re: non-remarkable non-LTS

[dbIII]

More that Lennert never knew where the cheese was before - but yes you have a point.
The annoying thing is the number of changes and the ones (like killing all user processes on logout!!!!!!) that show he just didn't ask anyone before making the changes. Things like his comment "what tool was used to create a username with a number?" show he's not getting good advice about the environment he's working in - on top of things like the newbie mistake of not checking for valid inputs. That's pretty fucking cheesy.
It's only really a problem because so many of these things are going live before others can do something about it.

Re: non-remarkable non-LTS

[Zero__Kelvin]

The default is what your distribution says it is, and your jealousy is telling. You aren't a mind reader, psychologist, or qualified / competent software engineer, so your ramblings about why it is that way are just that .. ignorant ramblings.

Re: non-remarkable non-LTS

[drinkypoo]

The default is what your distribution says it is,

Yes, and that's a problem, because the default should be to provide expected behavior. Sun's former use of csh instead of sh was a problem, too. Just do the classic Unixy thing and leave it the hell alone. I for one do not comprehend why anyone would write even a bash-specific script for something like an init script.

and your jealousy is telling.

If I have to be him to have what he has, I don't want it.

You aren't a mind reader, psychologist, or qualified / competent software engineer, so your ramblings about why it is that way are just that .. ignorant ramblings.

And you aren't a functioning human being, but you're still permitted to post to Slashdot.

Re:non-remarkable non-LTS

[khandom08]

not sure what this is using as pager, is this something built in?

It uses less. Type 'h' in the terminal and you will see the help screen entitled " SUMMARY OF LESS COMMANDS"

Re: non-remarkable non-LTS

[drinkypoo]

I don't even have to comment now.

You don't have to comment ever, but you keep coming back anyway

Re:non-remarkable non-LTS

[Dog-Cow]

Has the default behavior of journalctl changed? No? Then take your computer and use it to bludgeon yourself to death, you ultra-conservative whiny shit-face.

Re: non-remarkable non-LTS

[Dog-Cow]

You seem to have a serious problem with changes that don't affect you. Use your power cable as a noose and hang yourself. You'll never be bothered by systemd again, you whiny shit-bag.

Re: non-remarkable non-LTS

[Hognoxious]

I love a well-reasoned argument. Could you point me to someone who could provide one, you smelly fat kiddy-diddler?

Re: non-remarkable non-LTS

[thegarbz]

shill

verb
verb: shilling; plural noun: shillings
1.
A term used on Slashdot to describe anyone who disagrees with you or has a differing opinion.

Re: non-remarkable non-LTS

[thegarbz]

More that Lennert never knew where the cheese was before - but yes you have a point.

That's the principle behind it. Someone moved my cheese is about your reaction to things that others do to you.

Things like his comment "what tool was used to create a username with a number?

Yeah he's a dick to users. Yeah his software has bugs. All of which has nothing to do with the fact that many of the complaints are people's inability to RTFM or to blame him for defaults which are in the domain of the distribution provider and nothing to do with him. I mean do you go and complain to GNU because the latest version of Raspberian ships by default without ls aliased to "ls --color"?

Maybe if half the people directed their venting at the people actually in control rather than just blindly bitch on slashdot it wouldn't be as bad. But they don't. Also explained by someone moved my cheese, people make irrational decisions when something happens to them which doesn't at all affect their situation.

Re:non-remarkable non-LTS

[thegarbz]

Is brake on the left & gas on the right intrinsically better than the other way round?

Funny you mention that. No problem with moving indicators and wipers, or gear sticks from the left to right. People adjust to that within a few minutes. People also didn't have a problem back when Ford had the accelerator on the steering wheel. But then only until very recently you could find motorbikes with the throttle on the left instead of the right. And the standards for bicycles require the front / back brake to be different on left hand drive vs right hand drive countries so you can indicate your turn without slamming on the front brake and kissing pavement.

Yet everyone survives through this transition just fine. There are people that bitch about everything, such as the pedals not being cantilever from the top and instead from the bottom, or that the acceleration response is different for some cars (Tesla gets complains about this), or the different spacing between pedals, different sizes of pedals, if you actually looked into it you'll realise just how non-standard car systems are, and if you look through history you'll find many examples of people changing, altering or playing with it trying to find something "better".

Re:non-remarkable non-LTS

[thegarbz]

Why doesn't it respect the PAGER environment variable? Why do we have to use a NEW variable in the SYSTEMD namespace?

Because log file outputs especially with the additional information added by systemd are now wider than before so it makes sense not wordwrap, the same can not be said for many other uses of less.

Also because journalctl is a different program to less so why shouldn't it have its own config?

I mean... seriously? That's been the way you specify your desired pager for _decades_.

Nope. It's been the way you specify your desired pager settings under "less" for decades. Maybe journalctl isn't for you. Just set it up to dump to syslog and use less like you used to and let everyone else who is comfortable reading the manual use a different program.

You can do anything if you set your mind to it. Or if you post on a forum and get someone else to RTFM out to you.

Re:non-remarkable non-LTS

[thegarbz]

These dumbfucks seem to not understand conformity through established standards.

Oh I didn't realise there was an established standard for how journalctl should display it's output. I just tried running that command on and older unix machine and I got journalctl: command not found.

Maybe you can help me and tell me what the "established standard" should be for this program that didn't exist until a few years ago.

Maybe learning to use new programs are too hard for you. Give your computer to a 15 year old so he can setup journald to output to syslog and you can awk grep less your way through a text file the "better" way like you used to.

Re: non-remarkable non-LTS

[dbIII]

That's the principle behind it. Someone moved my cheese is about your reaction to things that others do to you.

I am very aware of the phrase as should be clear from me taking it one step more above. It's not just that "the cheese was moved" as in different behaviour, it's that Lennart is not taking the prior behaviour into account at all - he doesn't know where the "cheese" was in the first place and is not going to listen to anyone who does. It's not about improvements just different. It's a quick choice not based on experience, reinvention for the sake of it so almost always likely to be less than ideal on the first attempt. That has annoyed people.

If half the people directed their venting at the people actually in control

Hence my comment "It's only really a problem because so many of these things are going live before others can do something about it."
Pulseaudio mostly works these days probably because the team that took over a decade ago listen. Systemd is likely to benefit to being run in a similar way instead of the it is - the inevitable design flaws and bugs that crop up in a project appear to persist for far longer than they should due to the insular approach taken. I've still got a lot of stuff on CentOS/RHEL6 because a couple of software vendors are having mostly systemd related issues on RHEL7. That wouldn't happen with sensible defaults to fall back to so the new software can act like the old. The attitude is that the world has to change for Lennart.

Re: non-remarkable non-LTS

[F.Ultra]

This is no different from all the thousands of other project where distributions change the default by supplying their own configurations or patching the project in question. Since the maintainers on Ubuntu have decided to not change the default behaviour of journalctl then they have decided that this way is not shitty but better. You obviously does not agree, but then no one is forcing you to either. Myself I have yet to form an opinion on the matter, sometimes I think that it's better and sometimes I think that it's worse and when I find it to be worse, I use the "--no-pager" option to disable it.

Re: non-remarkable non-LTS

[F.Ultra]

If you execute "/etc/init.d/script" then you execute the script with your shell and not with systemd. This is why i.e upstart required you to make a link from /etc/init.d/script to /lib/init/upstart-job so that upstart could execute the script regardless of location, systemd does not do this so you have to explain further how systemd could interfere with your script if you executed it directly with bash.

Re:non-remarkable non-LTS

[UltraZelda64]

Awww... so you're saying I might make poor 'lil Poettering (however the hell you spell his name) cry. Awww, such a shame. Waaaaahhh....

If he can't take a joke, that's his problem. If he, as a supposedly respected software developer, has nothing to do but scour Slashdot topics and find discussions that include him and whine because people make fun of him... then, well, that's actually pretty pathetic. He can cry all he wants.

Also, today

[I'm New Around Here]

You should replace the batteries in your smoke alarm.

Re:Also, today

[tehcyder]

You should replace the batteries in your smoke alarm.

Wait, they use batteries?

Nine Whole Months

[Anonymous Coward]

But sixteen years is not enough for Windows XP?

Bring on the excuses...

Re:Nine Whole Months

[NoNonAlphaCharsHere]

Yeah, but at least XP didn't have systemd.

Re: Nine Whole Months

[Zero__Kelvin]

Why would we make excuses for your stupidity?

Re:Nine Whole Months

[johnsie]

So basically they went to all the effort of packaging up a whole version of a distro and releasing a version that only lasted 9 months. You're the twat if you think that isn't a giant waste of time.

Misleading

[Anonymous Coward]

Why? Because, in time, it will become vulnerable to all sort of attacks

This is misleading. The software is already vulnerable to all possible attacks. Over time, existing vulnerabilities might be exploited. Software does not become vulnerable because it is not 'supported'. That's not to say there is a risk, but the risk is not directly that the software is not supported.

Re:Misleading

[tehcyder]

Why? Because, in time, it will become vulnerable to all sort of attacks

This is misleading. The software is already vulnerable to all possible attacks. Over time, existing vulnerabilities might be exploited. Software does not become vulnerable because it is not 'supported'. That's not to say there is a risk, but the risk is not directly that the software is not supported.

You seem to have overlooked the phrase "in time". No one is saying that software magically becomes vulnerable the second support stops.

Re:Why we don't use Linux

[Kjella]

Personal preference has nothing to do with it. We're running a business, not a kindergarten. We need professional tools.

So availability of applications have what do to with support window? Maybe you shouldn't be running anything, since you can't seem to make a coherent argument.

Re:Why we don't use Linux

[ckatko]

You... you realize this is an intentional short lived release... right?

And... that upgrading Linux is as simple as running a simple command line and it doesn't break compatibility with everything...

Re:Why we don't use Linux

[DogDude]

Install a new OS without regression testing the apps you need to run on it? Really?

Re: Starting today,

[Joe_Dragon]

and NetworkManager does it still does not properly support bridging or bonding?

Rolling Release

[Philotomy]

This kind of thing is one reason I switched to a rolling release distro (Arch, in my case). I won't be going back.

Re:Rolling Release

[somenickname]

Meanwhile the people who use their computers to get work done use the LTS releases, Debian stable, CentOS, etc.

Re:Rolling Release

[thegarbz]

Because the ever moving target of rolling releases which could change at any moment are so much better than running the command "do-release-upgrade" every 6 or so months?

Re:Rolling Release

[Kjella]

Back when I used Ubuntu, I dreaded every upgrade because I knew something would konk out and force me to either reinstall or hammer the configuration until it worked. It's been ages, though, so maybe things are better now?

Doesn't that describe like every upgrade, ever? Even with the best of intentions things can break and then you have all those pushing the new shiny who wants to break things. The question is how often do you want/need it compared to the benefit of getting new hardware support, new software features and fixes quicker. Every day, taking it in stride (until it breaks in the most inconvenient way at the most inconvenient time), every six months, every two years... you can skip an LTS and do it once every four years, the choice is yours.

The only real downside I saw was that even with backports and PPAs occasionally sorting it out you sometimes had cascading dependencies where you wanted to upgrade just one software package and it set off a chain reaction. I mean the software already exists in a newer distro, if there was a simple way to say "Take application X from distro version N+1/2/3, put it in an AppImage/Flatpak/Snap/chroot container and let me run that isolated from the rest of the system" that would be great. Ideally integrated into apt-get so that branch would get updates along with the rest. If you upgrade your distro and "catch up" later, you can uninstall the container and install it the ordinary way. That way you'd have a vast library of "free" backports.

Note: I might be unaware what actually exists now, there's some years since I last ran Linux.

Re:Rolling Release

[arth1]

Meanwhile the people who use their computers to get work done use the LTS releases, Debian stable, CentOS, etc.

Very different definitions of Long Term for those.
Debian: 5 years
CentOS: 10 years upstream support
Red Hat: 10 years plus 4+ years of extended lifecycle support at extra cost.

In many fields, 5 year isn't enough. The cost and complexity of replacing legacy software and hardware can be a showstopper. If you work in manufacturing, you have to be able to support customer installations that's more than 10 years old. Including being able to run development environments that actually work with the hardware in question, even if newer OSes have dropped all support. Say someone publishes a way to hack all 2006 model cars from a big manufacturer. They want to provide a fix, but may need their 2006 systems running to do so.
Or banking, where they sometimes need to scan old microfiches or even punch cards, especially when one bank in the past engulfed another and not everything got nicely integrated. Got a handy hollerith driver or SANE fiche scanner plugin for Debian?

Re:Rolling Release

[Bite The Pillow]

I also switched to rolling release. It's called Windows 10.

Re:Rolling Release

[Philotomy]

I know what you're saying (I run CentOS servers, too), but for my workstation (yes, for professional use) and for personal (non-work) computers, I much prefer rolling release.

Re:Rolling Release

[Philotomy]

I've run both rolling and non-rolling distros over the years. All I can tell you is that, yes, I prefer a rolling release distro for my workstation. I've been running the same install of Arch for several years, now, and there have only been a couple times that applying updates caused me any issues. I run updates once or twice a week, and before running them I check the Arch news for any gotchas or pre-update tweaks I might need to perform (uncommon). Personally, I wouldn't run Arch on a production server, but for a workstation or a home system I think it's great.

Re:Rolling Release

[Philotomy]

I'm a professional developer. I prefer Arch (and don't have any problem keeping it up to date and stable). I think you're mistaken about it being "trash." That said, if you prefer something else, I don't mind.

Re:Rolling Release

[thegarbz]

Back when I used Ubuntu, I dreaded every upgrade because I knew something would konk out and force me to either reinstall or hammer the configuration until it worked.

Mine hasn't conked out since I've been using it. Hardy. Phwoar I just realised that's version 8 so I've been running it for 9 years. Even the gnome > unity or the init > upstart > systemd didn't break anything.

But then judging by the arch mailing list it sounded like the move to systemd there caused a shitload of problems with broken startup configurations.

Re:Rolling Release

[ottdmk]

The "ever moving target" can definitely be an issue. I'm not a Linux guy primarily (although I do run OpenSuSE and Steam-OS at home.) I'm mostly a FreeBSD guy (currently 11.0-RELEASE). I like the balance FreeBSD has struck. The base system is quite stable and is supported for long stretches of time. The ports system, on the other hand, is continually updated with the latest & greatest. Works for me.

Re:Rolling Release

[iggymanz]

things that roll hit bumps; rolling releases always break a given configuration and need fixing

Re:Rolling Release

[Philotomy]

That's possible (although I'd also say it's possible for non-rolling releases), but it hasn't been my experience with Arch. YMMV, of course. Some of it may be hardware and driver dependent.

Re: Why we don't use Linux

[Zero__Kelvin]

Your ignorance? This is no different than "SP1 is no longer the latest; there is now an SP2. In order to continue with this update you must first apply SP2"

Re:Why we don't use Linux

[turbidostato]

"So availability of applications have what do to with support window?"

Everything. Neither people nor companies "run" operating systems. They run applications and services built upon those applications. That's why operating systems with strong implantation on corporate environments expend quite a lot of effort making sure the most popular applications and services' support windows match theirs.

Re: Why we don't use Linux

[Zero__Kelvin]

... and yet, by your own admission, you refuse to use them. Interesting.

Re:Why we don't use Linux

[Tranzistors]

16.06 has 5 years worth of support. What kind of company is going to re-do their entire infrastructure every 5 years?

If you need 10 years, Red Hat Enterprise Linux has that. Add ~5 years of “extended” support, if you really need it. If somehow you have managed to get yourself in a real pickle and need to run it longer than that, you can maintain it yourself (as a company, using contractors, probably ex Red Hat employees) or use hardened (virtualized, in separate network etc.) unmaintained versions.

In any case, I am not aware of any applications that are supported longer than 10 years.

Re:Why we don't use Linux

[butzwonker]

That's exactly the reason why I use Linux for work and Windows only for gaming...

Re: Why we don't use Linux

[F.Ultra]

Or 'editions' as they apparantly are called now since Windows 10. Don't know if SP still are a thing on their server line though. But you are correct and more people should understand this, people running say Windows Server X are not running the initial release anymore.

Re: Starting today,

[drinkypoo]

and NetworkManager does it still does not properly support bridging or bonding?

That's not NM's job. That's ifupdown's job. It works fine there.

Question on Posting Guidelines

[ytene]

This might be a bit of a novice or silly question, but is there a reason why this specific release of this specific GNU/Linux distro has a dedicated posting here on Slashdot?

This isn't flame-bait and I'm not trolling... I'm actually trying to understand the process of selecting news-worthy submissions for posting. I can't recall seeing similar articles like this [either for earlier ubuntu distributions, or others] and I didn't see anything in the article that highlights this as special other than the unusually short lifespan of this particular release. Is that the reason for the posting, or could it be something else?

Genuinely curious.

Re:Question on Posting Guidelines

[SeriousTube]

Probably because it's the last release of the 16 series?

Re:Question on Posting Guidelines

[Anonymous Coward]

>the unusually short lifespan of this particular release

It's not unusual. Ubuntu has been following a 6 month regular release schedule for years.

>Is that the reason for the posting, or could it be something else?

Judging from the submitter's posting history, he posts stories that appear on softpedia.com. He probably works for them and perhaps is the author of those articles.

Once a upon a time, Slashdotters frowned on submissions by shills.

Re:Question on Posting Guidelines

[hackel]

lol, what? "16 series?" There are two releases every year, and it's been that way since the beginning. It's not a "series". They aren't connected. If anything, it was the *second* release in the 16.04 LTS series if you consider 18.04 to be the start of a new LTS series. Still not remotely newsworthy.

When asked for a comment,

[Chris Mattern]

the developer said, "Don't talk back."

Re:Why we don't use Linux

[zwarte piet]

You don't have your ~home in a separate partition? :O

Re:Why we don't use Linux

[footNipple]

That's exactly the reason why I use Linux for work and Windows only for gaming...

Then Linux and Windows together is like a mullet; business in the front and party in the back.

Just use LTS release

[iamacat]

Constant churn is only good for development or hobby systems, otherwise use an LTS release supported for 5 years.

Stop posting non-stories!

[hackel]

Yes, every incremental Ubuntu release only gets 6 months of extended support. It's been this way for many years. This is not a news story. Shame on you for reporting it, BeauHD. What has happened to the editorial standards on Slashdot?

Re:Why we don't use Linux

[DogDude]

Plenty of applications are supported more than 10 years. The same version of our point of sale has been supported for the past decade, and will be supported for another 5 years.

Re:Why we don't use Linux

[Chris Mattern]

Because other OSes never go out of support. How's using XP going for you?