Ubuntu, the platform used in the majority of cloud deployments worldwide, today released version 16.10 with hybrid cloud operations, bare-metal cloud performance, the ability to lift-and-shift 80% of Linux VMs to machine containers, Kubernetes for world-leading process-container coordination, full container support in OpenStack, and telco-grade networking latency enhancements.
…this isn’t really about the desktop anymore, is it?
I’ve used Linux exclusively for over a decade and I have no idea what those words mean. O_O
I have been around computers since my mid teens – forty years ago; and I grok about a third of the buzzwords they are using.
There really is room for a few webpages that operated as an expert system guiding you to a suitable blend of container/ VM/ cloud technologies to meet your needs.
Figuring it out using the web is darn hard work. Buzzwords are often only explained through other buzzwords and this goes on until your brain explodes.
Buzzword compliance is cool but plain language really helps. Plain language that can communicate to incomers as well as resident experts.
Here in the UK we have a “Plain English Campaign”; who offer publishers a “Crystal Mark” compliance logo. If our national tax guidance is understandable then it is down to these guys.
Perhaps the tech world could do with a similar scheme? It is part of how I see wikipedia in practice, I guess.
What a great idea! I’ll make a fork
Basically, web apps are so horrendously shitty and hard to install and manage that all this other crap springs up around them trying to make them easier to scale up and out, or to survive hardware failure. If you just have a simple app, or rarest of all an actual web SITE that’s about presenting content via static HTML instead of six-axis rotating websocket webgl teapot benchmarks or mining user data, then traditional web stuff continues to work fine.
Ruby is one such a thing as the most painful language to install. So I pass.
Make your Scripting/language easier to install, please.
Well:
MAAS = Machine As A Service
Plus
PAAS = Platform As A Service
Equals:
PIAS = Pain In The Ass
is enterprise server stuff. myself, as a long time *desktop* user, I don’t know and not care about those either.
Do you feel yourself identified with this blog post?
https://circleci.com/blog/its-the-future/
It is pretty funny actually and it is so true.
🙂
I’m relieved.
I thought it was just me going crazy with all “layers upon layers” some guys want us to code for.
Can’t come early enough the movement to simplify app development/deployment.
With Canonical going off track with their whole Mir NIH, I’m not so interested in their plans. KDE on the other hand should be ready with Wayland switch next year. Now the main remaining piece is pointer containment.
Edited 2016-10-13 20:32 UTC
I for one think Mir is good. There needs to be competition in the display server field–isn’t that what we want? To prevent a second situation like X11, where there is only one solution, and it becomes long in the tooth?
I for one hope that Mir and Wayland both coexist. It’s not like being cross-platform hasn’t forced frameworks to support 3+ display servers in the past, and Mir and Wayland are conceptually rather similar.
Funny you say that, historically we’ve gone from old X11R# where there were quite a few competing display servers, until XFree86 won out, I think mostly because it was free and could be included everywhere. Then it was more or less forked into X.org, due to XFree86 stagnating… X.org I think got renamed to just xorg, but it is still split and modularized, but there simply wasn’t many ways to improve upon it, from what I understand, because of all the old X11 protocol requirements. Enter Wayland that has simply taken forever.. then Mir that Ubuntu jumped on board with a ‘me too!’. The problem is that A) we don’t really need competition to drive innovation here, since it’s the display server, most people don’t even know what that entails. B) it really would be better for them all to work together.
I can almost guarantee that Mir is going to end up being Ubuntu only for a few releases until Wayland is supported by all the other distributions, then Ubuntu is going to adopt it, just like they did systemd over upstart.
So what if it does? That doesn’t mean it was for naught.
I mean, it’s paying a few developers bills for a while?
I personally don’t get the issue at all, even Chrome OS has its own display server… although it got done a little quicker than Mir.
I don’t know if it has hurt the Unity 8 development, although certainly I think the resources would have been more useful there… we might not have had to wait 2-3 years for any product if they weren’t developing the whole stack in-house.
I just don’t think Canonical is big enough to pull off what they’re trying to accomplish. I also find it strange how much you go on justifying them on here… for me, the whole appeal of Open Source is that I don’t have to care about companies, they’re all working together on these parts.
People can do whatever they want, of course, but ultimately the Canonical CLA’s make it impossible for their work to be widespread in the community.
At this point, it is very difficult to care at all about Ubuntu.
I agree 100% with that, Lunitik.
I stopped caring about Ubuntu probably around the same time they decided Gnome 3 wasn’t for them and created Unity.
Namely because to me, the initial Ubuntu goal of “Be Debian, with a 6 month release schedule and latest Gnome” was broken, much like various Ubuntu releases have been since then.
They have some of the worse NIH problems I’ve seen with any distribution. Even their latest foray into a ‘universal packaging’ system is going against what everyone else is working on. I understand the need to try to stand away from the crowded line up of other distributions, but then we now have the mess of Ubuntu being Debian based, but then modified enough that there is a huge list of Ubuntu based ones now.
While I’ve seen a bunch of Arch based or Gentoo based systems, you don’t see distributions based off of those! This makes things like Linux Mint (non Debian Edition) being sort of a third tier. As is fondly said, ‘use the source, Luke!’ But whatever…
rant off..
A universal packaging system to make up for the lack of and SDK that guarantees programs “just work” was needed about 10 years ago, but the Linux people were touting the superiority of package management then and how it used less disk space by re-using code. Now terabyte drives are available and games are over 5GB in size. Fixing this 10 years ago would likely of had an enormous effect on spreading Linux for the desktop, but too many superiority complexes in play inside the Linux community.
It means they wasted time of graphics toolkits and drivers developers, who instead of fixing bugs had to deal with supporting Mir. Intel driver developers outright refused to waste their time on it.
Edited 2016-10-16 04:53 UTC
Hi,
That depends who “we” are.
If we are Microsoft, then the answer is “Yes, we definitely do want a “divide and conquer” situation that wastes the competitors development effort while maximising end-user confusion and/or risk of compatibility disasters“.
– Brendan
The GNU/Linux community is very good at doing that alone, no need for external help.
I lost count how many technologies were finally getting mature, just to be forked and started from scratch.
Forked and started from scratch?
That is IMHO wrong.
When you for something, you take the source code and then work on developing from that point using your ‘forked’ code. You don’t start from scratch with a fork.
Well, that’s my opinion and I’ve only been writing software for 44 years so what so I know eh?
Excepting very few fortunate exemplars, FOSS efforts are not politically structured to prevent Diasporas. On growing, average FOSS structures tend to fragilize.
The dual answer is {‘Small is Beautiful’ | ‘Power and Empire go together, (anoint YOUR own)’}.
You Can Smell Problems on the distance if it’s a project about THE Few, THE Proud.
As a rule of dumb do not expend your life in projects ‘too big to fall down’. [From someone who did that].
Edited 2016-10-17 19:30 UTC
I think that choice and variety are what makes Linux and BSDs better than Windows. Having only one choice of anything seems like a bad situation for such OSes.
The attitude that we need to work on only one project of any one kind is counter to what made Linux so successful in the first place.
Hi,
The “divide and conquer is an advantage” mentality is why Linux has been such a massive failure (and why Linux only succeeds if a large company like Google steps in and discards the “divided by retards” user-space).
The ability to explore alternatives is an advantage for research projects that are never intended to ever be used as an actual product. For actual products you want to focus your efforts on the best alternative only.
– Brendan
Microsoft (and many other companies) does have many technologies that are not well known to the public.
In fact nowadays there’s an ethic of just trying out every little thing someone might think of and then see how the market reacts to the idea.
With Linux the fact still is that most of the desktop oriented flavors are build by individuals in their free time.
Most paid alternatives backed by RedHat, Oracle, Canonical simply target the enterprise market which has been slowly but surely migrated to cloud environments for some time now.
It’s much cheaper to not care about your own data center and having now the ability to deploy your core applications on whichever platform you desire is for most managers and young entrepreneurs a very good decision to take.
For Canonical it’s a very nice position to be in right now.
I’ll agree that if you are trying to make a product of something, you are right.
But those reasons for a unified system are counter to what makes Linux and alternative OSes appealing to people like me.
Additionally, I work for a company that was previously a part of Microsoft. I think that MS is a lot less concerted than you think it is, same as any company of that size.
The last thing to set as competition is a display manager, a much complex project that takes time to mature.
Virtually nobody outside Canonical are interested to contribute to the Mir project because it is considered as a waste of time and another attempt get control via Clause Licensing Agreement. It will be a matter of time Mir will either be yet another abandonware or converted into Wayland compositor.
A reminder is without Wayland developers doing the heavy lifting, Mir would not exist.
Yet Ubuntu is the only GNU/Linux experience able to match Mac OS X and Windows, specially on laptops.
Using systems with GNU/Linux since Slackware 2.0……..
Edited 2016-10-14 06:35 UTC
I don’t understand when people claim this… have you actually used any other distribution?
There is no reason Ubuntu would work better on any machine, especially since they aren’t working on anything related to hardware compatibility at all… all the code comes from the community at large.
Indeed, Ubuntu has more reliably introduced issues than they have solved them… the whole pulseaudio issue was because of the configurations they shipped it with, for instance.
It really angers me how much Ubuntu gets credit for other peoples stuff.
Sure:
– Slackware
– Red-Hat
– Mandrake
– Mandriva
– SuSE
– Yggdrasil
– Debian
– Ubuntu
Dual booting desktops since 1995 and laptops since 1999.
Only Mandrake and Ubuntu provided me a pleasant out of the box experience on laptops with full support for the hardware.
Sure, but have you actually used one in the last 10 years or so? Seriously now…
After installing Debian on my current laptop about two years ago, the only thing I had to install manually was the firmware blob for the Bluetooth adapter and BOOM! I was up and running.
In fact, I was even surprised to see that my networked printer was automatically detected and fully set up by Avahi! No need to hunt for drivers nor anything like that… Take that, Windows!
And that’s Debian, supposedly one of the hard-to-use distros. And except for slightly less battery life compared to Windows 10 on the same machine, everything Just Work (TM).
Honestly, I don’t think that people claiming that Linux hasn’t made great strides in usability in the last decade or so have actually looked at the damn thing in the last 10 or so years…
And yeah, it also bugs the hell outta me seeing Ubuntu getting credit for the heavy lifting done by RedHat and others in the Linux community…
Edited 2016-10-14 19:12 UTC
Eee PC 1215B recently updated to Ubuntu 16.04 LTS.
Bought in 2012 via the Amazon Germany with official support for Ubuntu 12.04 LTS.
The only laptop I still use with GNU/Linux natively, all the others have GNU/Linux VMs on them.
Edited 2016-10-14 19:25 UTC
Thank you, this explains your post.
You can tell he hasn’t seriously looked at other distros by that list.
Notice RedHat is on it, and not Fedora.
Mandriva and Mandrake … Dying and dead.
Slackware: not anyone’s idea of an easy to use system…
Missing: Arch/Gentoo/Mint/Puppy Dog
No, one can tell that I have better things to do with my life than trying every single Linux distribution listed on Distrowatch since 1995.
No one is trying to do so. But by that same token, one should refrain from issuing blanket statements about a subject that he/she apparently is purposely avoiding at all costs.
I mean Yggdrasil, really? You should have thrown SLS in as well. ^_^
And as correctly pointed out by other posters, Ubuntu not only does little in terms of actual contributions to the Linux kernel – i.e. hardware compatibility improvements you keep praising it for – but it is currently so set on its own ways due to the worst case of NIH syndrome I have ever seen (Unity, Mir, upstart, indicators, snap, etc.) that it has practically become a different beast altogether.
And it keeps breaking functionality/compatibility with typical Linux stuff where you least expect all the damn time!
If you wish to stay in an Ubuntu-centric environment for whatever reason, I’d posit that Linux Mint does a better job at “being Ubuntu” than Ubuntu itself.
But the last thing I want is to tell you how you should live your life. Different strokes for different folks and all that jazz.
Ubuntu and Red-Hat are the best supported distributions by ISV vendors, sorry if I don’t want to waste time figuring out hardware support, and also why software doesn’t install properly or runs on distribution flavor of the day.
Nothing wrong with that at all. Just those were the distros you listed as evidence that you’d tried other distros. Its just not an up to date list. I think there was a time when Ubuntu first came out that it was heads and shoulders above the others from a usability stand point. I think that’s kind of out of date now. I think there are many in the same realm of usability now.
I rather stick with distributions that companies care to support, which isn’t the case of those that you listed.
It is not only having to waste time trying all sorts of distributions, it is fixing issues because software packages from ISVs weren’t tested on them.
Hardly. Unity is quite inflexible. If anything outmatches OS X and Windows, it’s KDE.
A desktop OS experience is much more than just the UI.
It was an OK experience when Mandrake and SuSE were the best distributions for the desktop, that was around 10 years ago.
Nowadays KDE developers seem to be busy making the KDE Platform run on Windows instead or rewriting everything from C++ into QML.
Edited 2016-10-16 06:50 UTC
That has two problems. One is graphic cards´drivers support and this is a huge problem for Linux. The other is that you have to port the toolkits to it.
Is….was…..and always has been. Not that it’s the fault of the OS. It was the closed architecture, no support model from hardware manufacturers. Half written or undocumented APIs on proprietary hardware. It’s gotten a somewhat better in the past 10 years but it’s still not on par with windows and/or mac driver support.
What problems are you referring to?
Torvalds complaining a few years ago has led to stronger support from NVIDIA.
http://www.geek.com/games/torvalds-rant-against-nvidia-works-new-li…
Proprietary drivers work well but opensource ones fall behind in features and performance. The problem is that Mir has a different driver model than Wayland and X. That means that the companies need to port their drivers but they aren’t willing.
Glad I’m not the only one that noticed. I feel like there is an inflation of NIH reactionism with Canonical.
Oh well, no worse than Red Hat’s relentless systemd campaign. Maybe even a reaction to Red Hat’s tactics for systemd, Weyland, etc.
Not crossing fingers. At least Canonical as a “Company For Profit” is obliged to deliver what they’ve been promising and now we have MIR on Ubuntu Phones.
KDE???
KDE as a desktop is really beautiful seeing the screenshots, but the Experience?
Not to mention the promises!
Did you accidentally miss off “Developer preview of Unity 8 includes desktop, tablet and phone UX convergence”?
Have you tried using it? It aint pretty…
I use it on my phone, where it works well. I tried it a while back on my laptop and it was awful, but lets see where it goes
I have used it, briefly, on desktops. And Unity 8 still has a way to go for desktop use, which is probably why it’s still considered a tech preview. Unity 8 on Mir is a lot better than it was a year ago, but does need some polish on the desktop.
However, on mobile devices Unity 8 is quite good. It tends to be smooth and stable and I’d definitely recommend it for phones/tablets.
No, he’s just smart enough to know that these buzz-words mean… absolutely nothing. Microsoft calls it continuum and Windows universal apps. Ubuntu calls it convergence. Both mean jack crap, unless you count a piss poor user experience.
of course this isn’t really about the desktop anymore, there are no money to be made from the Linux desktop. the only way to get some profit is the RHEL way.
“…this isn’t really about the desktop anymore, is it?”
They closed ubuntu bug #1 with bunch of “well since Linux is on smart phones…” warm and fuzzy public relations stuff. They’ve given up on that desktop thing. With no official SDK for Linux to guarantee that programs just work, Linux will never have a chance. However ReactOS is getting surprisingly compatible, just need some rich guy to start funding development to get in done anytime soon.
Only Difference is YOUR documents and YOUR flows on THEIR ‘cloud’ desktops [Those have your name in it, just in case of any problem with Law & Order].
… there is no cloud,
– its just someone elses computer …
(As could be read on a T-shirt.)
And Id rather use my own computer ..
No thank you Canonical. You can keep your buggy software. I don’t know why anyone bothers upgrading to releases other than the LTS versions. Mint has the right idea by sticking to an Ubuntu LTS base, improving its desktop experience for users over the next two years and only upgrading the base when the next LTS version of Ubuntu is available.
Because not everyone is like you?
To be honest, I don’t see why desktop users ever use anything other than rolling release distros.
It’s true that things can break at any time with a rolling release. But the same is true of any software that receives any patches at all. And opening yourself up to the changes being a torrential and insurmountable issue someday later just seems like a lazy route to me.
Edited 2016-10-18 18:54 UTC
It hasn’t been about the desktop for several years now. The desktop is a saturated market (soon the mobile market will be, too). There is no “growth” in the desktop market which is what shareholders and analysts push for. No, what this is all about and will be for the foreseeable future is “cloud services” and the ad revenue datamining that cloud data can generate. If you’re not willing to use the “cloud” you better believe the major vendors of software and hardware are going to try to push you into it anyway. They will probably succeed, too.
Kubernetes = A Google project for docker container orchestration, f.e.;
App is seeing increase in requests, current container count not sufficient to handle load. Based on obtained data Kubernetes will spin up additional containers and/or shuffle existing containers around.
TL,DR: Now Docker can do what VMWare has been doing for years
divided they fall. here’s how desktop linux happens: the big players voluntarily standardize and then create value for for-profit companies in morally-okay ways, like mozilla taking money for default web search