So Unity 8 continues to be a no-show as production desktop environment, despite the claim that adopting Mir instead of Wayland would accelerate progress past Wayland. Meanwhile, Fedora 25 is shipping soon with Gnome on Wayland as default, and KDE-based distros are not too far behind anymore. And Mir would be even less far along without infrastructure created for and by the Wayland community, such as libinput, which it came to rely on. (That's without talking about mobile - Wayland shipped on Jolla's SailfishOS phone to endusers more than a year before the first Mir-powered Ubuntu Phone.)
I'd say history has proven the critics right on this one.
Writing a new backend for an existing DE (Gnome3/KDE) is not the same as writing a completely new DE from scratch. A fair comparison would have been porting Gnome3 to Wayland vs porting Unity7 (based on Gnome3) to Mir.
Also, Unity8 did land much much before on the phone which was the first target.
> Writing a new backend for an existing DE (Gnome3/KDE) is not the same as writing a completely new DE from scratch.
Having written substantial parts of the Plasma 5 shell (also nearly a full rewrite, which may just ship stable on Wayland before Unity 8, despite having been started later, and the Wayland conversion even later), I don't agree - "writing a new backend" is a substantial amount of work when the frontend doesn't have a concept of "backend"; a lot of X11 shell code traditionally made a lot of assumptions about being on X11. We had to rewrite our dock[1] as well for example, and that's before you get to talking about turning a X11 window manager into a Wayland display server and its many additional responsibilities.
And as mentioned, Wayland handily beat them to the phone, too. I don't see any evidence that Mir hasn't just been an extra expense in resources for Canonical, and that's with freeloading on activity of the Wayland community. How did it in any way accelerate their progress to production to be the odd man out?
Thank you so much for your work! plasma 5 is imo the first de for Linux that can compete with Windows and Mac OS visually. And as much as I would like to be a cli keyboard only user that's just never going to happen
From the first announcement, Canonical said that the reason for developing Mir was for their specific vision of touchscreens and convergence. FLOSS is so good because people can scratch an itch out in the open where everyone can benefit from what they do, even the mistakes.
When we tell people they're wrong for pursuing their own projects and instead should contribute to existing efforts, we're just encouraging them to develop in secret where nobody will critizie them. FLOSS is about choice, not everyone doing the same thing. Nobody has an obligation to develop for any project they don't want to.
> When we tell people they're wrong for pursuing their own projects and instead should contribute to existing efforts, we're just encouraging them to develop in secret where nobody will critizie them.
I think most people criticizing think that:
FOSS and contributing to existing efforts > FOSS and pursuing your own project > FOSS and developing in secret > proprietary
So you're basically saying "it could be worse, so please stop criticizing at all" ;)
Maybe some existing efforts need help, but I think Wayland is doing fine without Canonical. Either way, shaming developers for choosing to go their own way is toxic, just as bad as harrassing people for not taking every request on Github. Developing FLOSS can be so stressful because everyone has an idea about what everyone else is supposed to do.
Imagine what would have happened if more people had that attitude about Linux developers for not following the GNU Hurd!
The criticism is because they didn't just go their way.
I have problems with: 1) start of Mir resulted in Canonical not doing some Wayland commitments 2) start of Mir was kept secret for at least 6 months 3) loads of incorrect statements about Wayland that (despite retracting their claims) loads other people still assume is true 4) the reason for doing this has never been clarified 5) it other software to behave buggy. E.g. Wayland on this Ubuntu is problematic (see Phoronix); all kind of problems you won't find on any other distribution. It shouldn't be affected but it is. Wastes time of loads of people by a decision Canonical made without any clear benefit or reason.
In brief: People are _not_ being toxic. There are various very good reasons to question the decision.
It can be depressing at times, I agree with that. But I think that the majority of "criticism" wasn't shaming Canonical for starting Mir, but rather just asking "Why?". Canonicals answer consisted of many set phrases and things that just weren't true (e. g. "Wayland doesn't work on mobile" - Sailfish?). Don't you think that such a behavior can also be considered "toxic"?
IMHO Canonical's reason to start Mir were mostly to be in control. For example via CLAs. I understand it, but they should also understand that many people criticize that.
Any word on what's happening with "Snappy Personal"? That's their desktop build composed entirely out of snaps with an A/B system partition format. They have been really quiet about it for the last few months. They started talking about it for 16.04, but then planned to defer that until 16.10 (https://ubuntuforums.org/showthread.php?t=2312045).
We have a sprint next week to discuss some of that. The recent focus around snaps has been on the feature set leading to Ubuntu Core 16, which is itself in freeze at this point. The features are mostly general, though, and have been landing on 16.04 and 16.10 on the way.
I miss the days when new Ubuntu releases included things desktop/laptop users could get excited about. Nowadays it's all stuff for servers and phones. I run Ubuntu on both my workstations, so it's sad to see that use case feel more and more like an afterthought.
The reason is they make little to no revenue from desktop, and no profit. As a cost center they subsidize it to bolster their server and hopefully mobile offerings by a network effect - people using Ubuntu server may want to develop on Ubuntu desktop and vice versa. [1]
It would be a loss if Canonical drops support for desktop as in my experience Ubuntu desktop is the easiest and most practical desktop Linux available. I like that Shuttleworth is committed to it but hope that that lasts.
To me this is refreshing. The Linux desktop has been, for me, a very mature product for the past couple of years. I no longer fear that having to update will mean tweaking everything to get things working.
The issue that Red Hat realized long time ago, and many of the other GNU/Linux vendors as well, is that there isn't hardly any money to be made on the FOSS desktops.
Developers of desktop software usually live from unit sales of their software and consumers don't pay for subscriptions, trainings or support.
So this leaves enterprise customers as the only source of income for many distributions.
you have inadvertently indicated exactly why the desktop, up front cash drain as it might seem, is actually a huge cash gusher. That is because it is the piece of the ecosystem which users bond to most intimately, and if they don't like your endpoint UI, they sure aren't going to be axed to use your (highly profitable) backend services. Witness how Ubuntu, arguably the original "pretty" Linux, dominates amongst startups doing cloud. And why Windows lost so much ground. And why AAPL was so stupid to dump the Xserve. Can you imagine what might have happened if Apple took just 1 or 2 billion out of its 250 billion cash horde and "did BSD" on the cloud? Properly?
I have a nexus 5 that I just play with no sim card (currently on the Nexus experience ROM). I wonder if anyone here has tried installing yakkety yak on a nexus 5. If you have, will you please share your experience? Thank you.
I believe the parent means installing Ubuntu Touch on your android phone. The best tool to do this with is MultiROM[1], an app in the play store that allows you to select the ROM you want at boot time.
I tried this a couple years ago on a N6 and was very unimpressed as it was incredibly slow and unresponsive, had no apps you could install, and just barely met the definition of "phone OS". It was more of a PoC and didn't even have a shell prompt. Things have probably improved a bit since then[2], but I would expect you'll get bored of it. I follow the android ROM scene and (unlike iOS) I've never seen people say "#$%$% all this! I'm gonna run Ubuntu Touch from now on!"
Still, it's fun to install just so you can say you did.
I installed Ubuntu Touch on my Nexus 4 a few months ago to compare it to Sailfish OS. It ran reasonably smoothly, but seemed to be missing a bunch of functionality. I installed a Spotify client only to learn that because Ubuntu Touch pauses background apps, you could only listen to music with the screen on and the app in the foreground. Similarly, if you started a Terminal session, and, say, went to the browser to look up a command, you would return to a completely new Terminal. I uninstalled it a few hours later - I'm willing to put up with a lot in a mobile OS, but a lack of multitasking puts it in 2007 territory.
I know exactly what you mean. If you still want to get excited for UI changes...have a look at Ubuntu Gnome. They generally have nice tweaks/changes every update.
I used to generally update version after version, but recently, I have not seen a good enough reason to switch from 16.04 to 16.10 (I intend to stick to 16.04 till the next LTS version is released).
If we speaking in terms of upgrading Ubuntu to 16.10, it makes sense whether you desire to have latest version of DE-specific applications e.g. Nautilus (updated to 3.20), xfce-terminal (was ported to GTK3 [1]) or obviously a whole DE like GNOME [2] or XFCE [3]. Personally, I am going to stick with 16.04 too.
I will be downgrading to 14.04 due to bad hardware issues with the newer releases; there are big reports with many angry users, but ubuntu doesn't appear to find AMD mobile graphics and Broadcom to be a priority.
What he want is a solution that updates the packages he happens to want while backporting fixes to everything else (until he wants an update to the latest version). Of course, that's what everybody wants!
Unfortunaly it's impossible to deliver unless you pay people to do the work or do it yourself.
Yes, people are using in production. Last community meeting we had some folks reported a large server deployment in the UK (in the order of thousands of servers) that was being partly transitioned to snaps. I think that's the biggest use I've personally heard about so far.
There are some people interested in having them working on 14.04, btw. It'd be great to see that working well.
I'm always intrigued by this desire for integration. Even if you want to have one device that is mainly a phone but also a PC when you connect it to an external 24" screen and keyboard - why can't it have totally different UX experiences optimized for each state? Is that really so bad?
Windows 8 was a UX disaster - so maybe that's just coloring my judgement. I haven't used Ubuntu on mobile yet so guess I need to try it.
Windows 8 was a separate UI for desktop and tablet. Windows 10, for comparison, more closely combines the tablet and desktop modes. Windows 8 was a disaster and Windows 10, at least for desktop/tablets, is more of a success.
I think the problem with completely different UX experiences is exactly that. People want a similar experience with the same software. Microsoft assumed with Windows 8 that the tablet interface would work on the desktop but people don't want that watered down of an experience. Windows 10 at least lets you use tablet-designed applications on the desktop in a more integrated way.
> Windows 8 was a separate UI for desktop and tablet
Wait what? The "Metro" was the combining of the GUI for touch and mouse. I actually prefer the Win 8 UI since it is more of a tiled window manager like I use in Linux. Windows 10 is back to click with a mouse. Windows 8 I never touched the mouse unless I needed it in a program.
> Windows 10, for comparison, more closely combines the tablet and desktop modes
It was a move away from Tablets and back to laptop desktop focus.
> The "Metro" was the combining of the GUI for touch and mouse... I never touched the mouse unless I needed it in a program.
The Metro interface was not optimized for mouse operations; it was a touch UI you could "touch" with a mouse. All apps were full screen. UI elements were hidden and meant to be swiped from the sides of the screen. Nothing about Metro seemed to be optimized for the desktop.
But thankfully you could completely switch modes and have a regular Windows 7-like desktop experience. When I ran Windows 8.1 it booted straight to the desktop, I used ClassicShell for the start menu, and I never saw a single metro app.
With Windows 10 can run metro applications in regular desktop windows as well as switch to tablet mode and run regular desktop applications and metro apps full screen or tiled.
Metro supported keyboards (to an extend) but I wouldn't say it was "optimized" for it. Even some Metro apps, like Netflix, couldn't be controlled by the keyboard. That one kept me from using the Netflix app on Windows 8 as a HTPC solution.
Windows 7 was not a tablet environment. Windows 8 was a full tablet environment and Windows 10 is a mixture of both and tablet took a tiny hit.
I'm weird I know but I preferred Windows 8.1 and more so on touch devices. I have about 20 touch screens at work and I miss the charms and the swipe in from the right and the easy short cuts. Open a program in Windows 8 was win-enter. Windows 10 it is win-arrow up. silly example but these arguments are usually silly when talking preferences.
You aren't that weird. My daughter had a touch screen laptop and she also preferred Windows 8 to Windows 10 and missed the charms and the swipe in. I, however, could barely use her computer because of all the fingerprints!
Windows 8 on the "desktop" was a carrot for all the computer manufacturers who decided, that year, to put touch screens on everything.
A better strategy for Microsoft would have been to call it Surface OS and release it just for tablets (and phones) and focus on Metro and have the desktop as a cool add-on for their tablet OS. For the desktop, they should have done a Windows 8 that was more like 7 yet still have Metro but more as an add-on. This wouldn't have helped you as much but it would have been better for them overall.
> A better strategy for Microsoft would have been to call it Surface OS and release it just for tablets
I say open up the ability to run optional desktop environments like we have in Linux. I can have my tiled window manager (I am sure 2% of users would LOVE it and 98% would HATE it) and then others could use one of the official desktop environments. No need for a whole OS version. Its bad enough we have Pro and non-Pro Windows 10.
Seems like a recipe for confusion and fragmentation. It certainly hasn't helped Linux.
Although my suggestion sounds like two different operating systems, it's really just about how it's marketed. Fundamentally the technology doesn't have to be any different.
They shouldn't be totally different. Common look and feel is welcome in my opinion. Then obviously they should be customized to adjust for different input devices and screen sizes.
It can be done without repeating the Windows 8 fiasco - which seemed to be tailored for touch-input devices, leading to clunky experience on a regular PC.
It's Android, but from the look of it, it's adapted nicely: https://youtu.be/TjAxCWO84L4 (I admit I haven't used it myself, but even if there are some quirks not visible at first glance, my point is not that Remix OS hit the sweet spot, only that it can be done in principle).
Yes, I think Windows 10 is a more mature and sane take on this concept, where the "apps" can now run windowed as usual. Makes the transition _much_ less jarring, and is more like how RemixOS also mentioned in the comments here does it with Android.
Right now what I am seeing is the majority of tablets with keyboards are being sold with Windows 10 on them, lets see if OEMs decide to sell Android 7 as well.
I used to have the same opinion until I found an example of UI continuity which actually helped me learn a different UI quicker and didn't feel like it held me back once I had learned it.
But well, that wasn't continuity between phone and desktop. It was continuity between GUI and CLI (openSUSE's YaST).
Whether it really makes sense trying to find a compromise between [mouse&keyboard + big screen] and [touch + small screen] is then again an entirely different question. And in my opinion, this isn't happening until we have a good generic solution for projecting the vast number of different screens that you need on a phone UI to one big, densely-settled screen as is necessary for a good desktop UI.
Windows 8 was not a UX disaster. It just wasn't finished. How long did you use it for?
I personally used it the whole time it was available. Despite the fact that it was unfinished, it was still vastly superior in many ways to other desktop systems. So, I'm quite certain that I don't know what you mean at all when you say "UX disaster".
I prefer Window 8 since I use a tiled Window Manager (i3) on Linux. Windows 8 was a much better productivity UI for me since I never had to touch a mouse unless I was in a program that specifically worked better with a mouse.
Since you seem to want actual evidence that Windows 8's UX was a disaster, well, what's the biggest difference between Windows 7 and Windows 8 for a user? It's the UX, I assume you can agree.
And if you look at the marketshare of Windows 7 vs. Windows 8/8.1, you see a pretty big difference.
StatCounter (includes mobile): 20.36% vs. 5.64%
netmarketshare.com: 48.27% vs. 9.61%
Steam Hardware Survey: 36.38% vs. 10,78%
Now, you might respond with Windows 7 simply having more adoption, because it has been around longer, which is a valid point. I can only really refute that with the anecdotal evidence, that Windows 7 is currently seeing more growth off of the people moving away from XP and Windows 10, than Windows 8 is seeing.
To be fair, I work as a sysadmin / helpdesk monkey / sometimes-developer at a Windows shop, and the amount of complaints about Windows 8 was a lot lower than I had expected. A couple of users did ask me, however, to install ClassicShell on their machines.
(Windows 8 has given me more trouble sysadmin-wise than its predecessors, but most of that was due to third-party software. Especially, Windows 8 will regularly hang on startup say "Please wait...", at least on some machines - but that has nothing to do with the GUI.)
Personally, I still think the lack of a start menu is very disruptive - and since Microsoft re-introduced it in Windows 10 makes me think I was not alone in thinking so.
The basic problem with Windows 8 was that Metro apps and Win32 programs didn't mingle. The whole desktop was basically treated, within the UI at least, as a single big app.
Come 8.1 and this was partially reversed, with Metro apps being possible to manage much like Win32 programs (showing up on taskbar, could be closed or minimized via icons top right).
And i must say i kinda like using the start page concept for launching programs once i have things set so i go right to the "list" rather than dealing with the live tiles.
I have been contemplating getting a smaller Windows based tablet as an alternative to my existing Android one.
The basic problem with Windows 8 was that Metro apps and Win32 programs didn't mingle. The whole desktop was basically treated, within the UI at least, as a single big app.
Hah. The first time I played a music file in Win8, the application took over my screen, and I had 50 Cent glaring at me (almost larger than life) while my song was playing. I guess he though my music was inadequate, and thought I should buy his stuff instead.
Also, just going to the program tiles screen was just a wall of visual noise. All these tiles were animated, it was hard to focus on what was there with all that going on. You can (and I did) turn off animations for them (or just delete them from the shortcut list), which helped a lot. So yeah, that was not a great first experience.
I'm totally fine with Windows 10 these days, at least as far as UI goes.
> The basic problem with Windows 8 was that Metro apps and Win32 programs didn't mingle.
True! OTOH, I have never seen one of our users use a Metro App intentionally. At most they would run into the builtin PDF viewer if no other viewer was installed or registered to open PDFs.
One of my coworkers has a Windows phone, and he tells me on the phone the interface is very nice. But I still cannot imagine what the people at Microsoft must have been thinking...
In the sense of Win32 the API, I can understand their desire well. ;-) I worked on an application written in C for Win32 for about two years, and while I learned a lot about C in that time, I came to despise the Win32 API.
On the other hand, I don't think they will ever get completely rid of that thanks to the huge, huge body of existing third-party apps.
"Most" isn't probably true, but the number was quite significant, indirectly confirmed by Microsoft by backing down from some design decisions in version 8.1 (bringing the Start button back) and 10 (restoring the more traditional-looking Start menu). Userbase pressure was an obvious factor in these choices / concessions. I can't recollect anything similar in the history of Windows UX.
I consider myself experienced and unbiased, but the visual inconsistencies (animation glitches, awkward scales and proportions) drove me away from trying Win8 longer. I'd be very curious to have more details about your experience with it considering your above message.
Meh, let them try! At least on ubuntu we can choose between a lot of DE and WM, and we are not obliged to use unity anyway, so nothing is really forced on us (Still wishing to have a win7 ui on w10)
It depends how it's done. Microsoft's approach was pretty wonky as it basically gutted the desktop. If Ubuntu's one is more respectful for the desktop, why not try it?
Seems kind of like aiming for a convergent experience with a pocketknife and a table saw, but hey, maybe my pessimism is unwarranted. I use Xubuntu anyway. I'm extremely grateful for all the effort people have put into Ubuntu over the years.
I think this is a good thing in the long run. I have no problem being the beta tester for a unifying ui. This can consolidate development focus which arguably will reduce bugs. As people get more tired of one ui, there will likely eventually be enough pushback to cause more custom options.
A major difference between ubuntu and every other os is ubuntu is open source with a rich enough history to show it has staying power. Worse comes to worse another fork happens. But the OS UX paradigm has a huge opportunity to evolve with this kind of focus
There's a reason I've not used plain Ubuntu in years, and it's Unity. Pretty much all canonical variations of Ubuntu ship with a better UI than Ubuntu anymore.
I don't agree with this. I have done Gnome 3.18 for 6 months, did i3 for another 12, and tbh, I found myself with a bog-standard Unity interface on a new machine about 3 months ago and I haven't bothered to change it. Why? Because it doesn't try to be clever, it doesn't get in the way, sure I have to use the mouse a tiny bit more, but it just seems nice and vanilla. It works well by being inconspicuous and straightforward, while not trying extra hard to be boring (Cinnamon). And the fonts look crisp like OS/x which, to my detail-oriented eye, is a big deal. Actually the only UI I consider truly interesting, believe it or not, is e17. That feels like being on a Star Trek transporter beaming onto a Romulan Warbird. Real deal no compromise "this ain't Kansas", if that's your thing. Try Terminology for a taster.
All the others (KDE) are trying too hard while simultaneously not being truly interesting. Or they're stultifyingly boring (cinnamon - why?) without being resource cheap. On this front, xfce on armbian is a beaut. Cinnamon without the weight. I also give i3 a big thumbs up btw but I just can't live without the tray icon conveniences for bluetooth, dropbox, power down etc.
Keep in mind, iOS apps send your devices' serial number & MAC address via a simple API command, and this can all be done without asking for permission. If you bought an iDevice in a store, you're also on CCTV buying that device and it can absolutely be traced to your canonical name.
The idea behind Ubuntu tablets is that it imports the security model of the Debian Linux trust chain to the user, and so worries of serial numbers and the typical 'link-ability' one finds in proprietary iOS devices are vanished.
They're not forcing it on anybody any time soon. Until Unity 8 has parity with Unity 7, it's not even going to be considered as the default. Funny thing is, their efforts to avoid the issues with and backlash against the original Unity are the reason Unity 8 has been delayed for so long.
Whatever else you might say about Canonical, at least they're doing the whole 'convergence' thing in a pretty open and responsible manner.
Is there a meaningful difference to responsive websites? Because those have generally evolved to a point where the interfaces work quite well across devices while sharing meaningful amounts of code to make it easier than build x individual UIs.
It is a terrible idea. No matter their popularity phones have tiny screens and limited, slow input methods compared to desktop PCs. And mobile OSes have to be provided with special workarounds to deal with the absolutely terrible mobile IP networking that cannot even keep a TCP session open reliably.
Convergence to a user interface that provides the crutches for these limits of mobile platforms neccessarily gimps the desktop UI too.
I'd say it depends. I'm a rather happy owner of a Microsoft Surface 4 Pro - and I do see the need for the ability to switch between modes (and I can't say MS has exactly succeeded here, but also not entirely failed). Granted, I don't really see why I'd want UI/UX adapted to a screen 1/8th the size - but to a certain extent "tablet mode" and "tiny tablet mode" are probably more similar than "tablet mode" and "laptop/desktop" mode.
I'd have to test it of course. One thing that MS has not managed is make a UI that works well for high resolution screens (for legacy applications, that is). But I guess one of the reasons I'm positive to the use of "convergence" buzz word, is that it could mean that all GUI-frameworks are expected to scale in a sane manner. I absolutely understand the skepticism that it will end up being "a worst of all worlds" UI.
Nope I think it is a horrible idea also. Microsoft showed us that mixing the UX between a 5" screen and a 25" screen sucks. I can't imagine Ubuntu working it out any better.
Why the hell do they worry about UX convergence yet I can't put the dock on the right side of the screen?! (and only just got the ability to put it on the bottom earlier this year!)
Working everywhere with the same tools on one device would be awesome and I can imagine Unity and Linux being much better at this than Microsoft's pathetic attempt.
Having used Windows 10 and the ascendance of responsive webpages, I can see its potential. Obviously Microsoft completely bungled it in Windows 8, but otherwise:
This is something that can be done and done well, and if so I'd love to have a phone with the power and software of a full desktop device, expressed in a phone-oriented UI.
The trick is that there's a matrix of several features that have to be considered:
1) Tiny screens means that multiple visible applications is not useful, and that all OS UI needs to be auto-hidden (something users want to be able to do anyways). Font sizes may need to be scaled up (that's something you need to be supporting anyways for older folks).
2) Touch means adding new ways to interact with existing widgets (drag scroll, for ex) and buttons need to be scaled up to have a larger hit-area (and you alreayd have to deal with scaling up fonts, so fixed-size UI elements already went out the window).
To me, the important thing is that 1 and 2 are not duct-taped together. If I have a phone and a stylus or I plug in a keyboard and mouse, I want 1 but not 2.
On the other hand, if I have a 20" touch-screen, I want 2 but not 1.
Unified UI is a terrible idea: the idea boils down to this: instead of having several UIs that serve the interface they operate on well, have one UI that serves all the interfaces poorly.
I'm not sure if it's a terrible idea in general, but a Ubuntu phone would be many years behind iOS/Android when it comes to apps. I think only Apple and Google have the marketshare to pull off a convergent OS at this point.
Admittedly, I have been told by several people that the new interface worked very nicely on tablets. That's what it was made for, after all.
On desktops without a touch screen, I found it annoying and still do. Of course, one can still get work done using the new interface, but I still don't understand how they got the crazy idea of throwing out the start menu.
It's been tried over and over again for a decade now, and keeps running very hard onto the rocks of "users really, really don't want that".
I get why devs want it: it's a single codebase for everything. But nobody, and I mean nobody, wants to use a desktop computer the way they use a tablet or a phone. There's a reason people tend to disable computer touchscreens.
>But nobody, and I mean nobody, wants to use a desktop computer the way they use a tablet or a phone.
That's not quite what's on offer. It's more like the ability to plug your phone into a monitor and carry on working. No longer carrying a laptop around (once tablets/phones are at CPU horsepower parity for office work).
So yeah, it's not working backwards like you suggest, it's making the phone/tablet more like a desktop (add mice, keyboards, screens etc.) One mobile device is quite likely all we'll need in the near future, not three separates.
2-in-1 laptops are already selling quite well as MS gets its own convergence efforts underway. Apple are actually lagging in this field and kudos to Ubuntu for steering the ship in an innovative/inevitable direction.
> It's more like the ability to plug your phone into a monitor and carry on working.
Do I really need a phone to do this ? As long as I can reach the same data I don't really care if I'm on the same device where I started working on it.
In fact the software I want to use when I am in front of a monitor is much different than the one I have on my phone, and I am not talking only about user interface, but also features.
Cpu power is getting cheaper and in smaller form factors.
I suspect that instead of connecting our phones to the display devices, those devices will be powerful enough that we won't need any integration. And if I really want something portable, I will be able to "add a cpu" with an external stick (like Intel Compute Stick, but cheaper e/o smaller).
And if I really want something portable, I will be able to "add a cpu" with an external stick
I'd say it's even more likely that anything computationally expensive won't be done in the same room with you to begin with 10 years from now, but will be done by an anonymous compute node in a datacenter that you're leasing for the 5 seconds you're using it.
The problem is that model doesn't work very well for probably the most common computationally intensive task consumers do: games.
It also doesn't work well with tasks involving non-trivial datasets. If you've got say 20GB of video from a Gopro you want to transcode, shipping it out to a datacenter is going to be really painful.
> There's a reason people tend to disable computer touchscreens.
They do? I've disabled the touchpad on my surface pro 4 (and have been known to toggle it and the nub on my thinkpad) - but the touchscreen actually works great, I think. So much so, that I'm now touching all screens everywhere, like a three year old that's been given a tablet as her first computer.
I've actually thought that a good touch screen and perhaps stylus, would be the perfect environment for something like the ACME editor: https://www.youtube.com/watch?v=dP1xVpMPn8M as a new power-user editor.
Tried 16.10 with Unity 8 in a VirtualBox VM, and after typing in my password and hitting enter, all I get is a black screen for a few seconds before being put back on the login screen.
Ubuntu has done that a long time if there's a problem with graphics drivers. Don't know more specifics for your issue, since I haven't run Ubuntu in a VM in years.
I am currently using Ubuntu 16.04 LTS. There are lot of changes since 14.04 LTS. GUI change is easy to notice at the first time. You will see different look in GUI. But, I don't know the GUI somewhat buggy to me sometimes. The screen appears to be black and it will repaint when you are moving your cursor around the black screen.
My friend told me that Ubuntu's GUI system was so heavy. He suggested to use Ubuntu Mate or Kubuntu as alternative distro.
Ubuntu 16.04 and newer have a feature to auto detect if there is no hardware acceleration and can work with that.
If you want fast fast, you can force to work without the nice effects.
IIRC, 16.10 has low-resource usage mode that cuts down on the visual effects.
If you're seeing that kind of artifacting in Unity, you should consider reporting it as it's a bug with details about your graphics hardware. Personally, the only times I have issues are when exiting games, but that's complicated by the fact that my laptop has a 4K screen. Still, everybody's machine is different, and what works fine on one configuration can be buggy on another.
You need not use Unity. Things are much better with Openbox; the usual problems of compiz and friends eating all the available RAM pretty much goes away.
I didn't get that sort of problems with compiz, but I'm on a 16 GB laptop so maybe I didn't notice.
I agree with not needing Unity. I'm on Gnome flashback and 16.04 finally let me get about the same DE I had with 10.04. It's Gnome 3 but it looks like the Gnome 2 desktop, with only one panel to the bottom and the compiz cube. I use the 3D rotation effect to remember which side I came from. A 2D slide doesn't hint me enough and I get lost in the virtual desktops. I don't use the cube for anything else. All considered, I'm happy with 16.04 and I'm staying here at least until 18.04.
All considered, I'm happy with 16.04 and I'm staying here at
least until 18.04.
Having moved from 14.04 to 15.x, I cannot agree more.
However, many packages aren't updated in the LTS releases which prompt one to move over to the intermediate releases. I wish Canonical were more agile about updating compilers and such on LTS versions.
You can find a PPA (apt repository) for almost any app or tool these days. I find that just adding a few of these for specific cases to a LTS release gives you the best of both worlds.
My current ones: git, Google chrome, flux, node, nylas, spotify, tarsnap, blender
I went from 12.04 to 16.04. I skipped 14.04 because of the DE which I couldn't bend to my will. I remember that I had similar problems in the last months on 12.04, with some new versions of some programs unavailable on 12.04. I guess that I could have installed them in a 14.04 docker container (would the kernel version have been ok?) Maybe you can do the same for a compiler. I don't know if it would be convenient. I'm more into scripting languages and I use gcc directly only every other year or more, so no big deal for me.
Current Kubuntu with Plasma is anything but lightweight.
There's really lightweight alternatives like Lubuntu and Xubuntu. You can give them a spin in a VM. If Linux for you mostly means a browser window and a few terminals, they'll do fine.
That undersells Xubuntu which is a full DE and very similar to GNOME 2, if you just want a classic stays out the way desktop then Xububtu is excellent.
To be fair, in comparison to most other DEs, it is still rather heavyweight. Definitely more heavyweight than LXDE and Xfce, and Unity is pretty well-optimised at this point, too.
But yeah, for what KDE is capable off, especially in comparison to GNOME or the Windows-world, it's actually incredibly lightweight.
I run it on a 4 year old mid-range laptop with 4 GB of RAM, Intel HD Graphics 4000, and as long as my system isn't under heavy load, it's buttery smooth, even with many of the bells and whistles enabled.
> I run it on a 4 year old mid-range laptop with 4 GB of RAM, Intel HD Graphics 4000, and as long as my system isn't under heavy load, it's buttery smooth, even with many of the bells and whistles enabled.
It is a desktop environment, not a 3D game. It should ALWAYS be buttery smooth. You can run a desktop environment on a 20 year old machine.
I used to run Kubuntu (with KDE 4) on an old Yakumo laptop, with 384 megs of RAM and a 600 MHz Celeron. Ran very smoothly. Coudn't have more than 2 apps open at a time, though. (And even that was sometimes pushing it.)
I don't know what they do, but Kubuntu does seem to be a bit heavy compared to Plasma 5 on other distributions. I seem to recall KDE devs calling out Kubuntu for its poor KDE integration.
> Why do I need to disable a filesystem indexer by default that slows the system to a crawl?
As far as I'm aware, that's considered a thing of the past. I haven't seen it recommended for any recent version anywhere and at least for me, it runs perfectly fine without disabling it.
I also only have one kwallet daemon running right now.
My rule of thumb: don't use Kubuntu, use some other KDE distro ( used to be mandrake/mandriva, now possible OpenSuse or Arch or something, I don't follow it too closely ATM.)
you are almost certainly using some laptop with optimus graphics, given your symptoms. On a standard PC, and I have tried dozens of permutations, the upside of Unity 7 being so "old hat" is that it is absolutely 100% rock solid.
Can anybody say how well will Unity 8 work with convertible and detachable machines that run with Windows 10 out of the box? I have an HP X2 210 lying around that runs 16.04 well but without PM, Sound, detached mode and multitouch touchpad.
Not sure what about systemd you're interested in, but I was hoping for improvements in boot time regressions since the switch from upstart. Searching bugs it sounds like it might have instead gotten worse in 16.10.
Does anybody actually use Ubuntu in production? I've forsaken it on the desktop (I got sick of my system breaking every 6 months, and having out of date packages), as have many others, but it's still quite popular there. On the server, though?
Really, basically every server I see is either Ubuntu or Centos. Very few Debian. With containers is different, who packaged the container for a given service wins. Example: Elastic uses Debian for their docker images on docker hub.
In part yes, but I'm a developer so I'm probably seeing those new Ubuntus that made Ubuntu overtake Debian. If I were an op I'll probably have a more balanced view.
Furthermore the sample is relatively small and my own sample is at least 100x smaller.
CentOS is in a tricky position with regard to ALPN/http2 support--there's absolutely nothing on the horizon. A continuing shift to Ubuntu in the webserver space is almost certain in the mid-term.
I've seen it used in a big financial firm that processes card-holder data, but they were phasing it out in favor of Redhat. Still, Ubuntu is a solid choice for servers I believe, it seems to have fallen out of favor in some places though.
Second your desktop comment, Unity is unusable for me.
Yeah, it wasn't even Unity. I prefer i3 anyways. It's more that I like the bleeding edge (I need it for a lot of stuff), and I hate having to re-install every six months.
I've used Linux for a very long time. Started with Slackware on floppies. Used Red Hat and derivatives for a long time, including at some large companies.
Nowadays I'm mostly with small, early stage companies in the Silicon Valley. From what I've seen in this environment, the OS of choice in production (cloud instances) is almost overwhelmingly Ubuntu.
I started out by downloading a very minimal Slackware distro on a 2400baud dial up Internet connection. A few months later I bought Slackware (full) on about 50 floppies - I felt like a rich man having all of Slackware available. Good times.
Been using Ubuntu LTS (NEVER the October releases) in production for years. As smooth as Debian for upgrading.
If you're breaking your system every six months, that's probably because you're using the October releases instead of sticking to the April ones (which are usually the LTS ones and have fewer experimental packages).
My rule of thumb is to only set up October releases for playing around and getting a feel for what's coming down the pipe, and NEVER use them for production. Only the .04 ones.
Really? Hmm, I've been upgrading the same installation since 10.04 or something, it's been working well. However, I did create a list of scripts at some point to make reinstallation a one-command affair:
How would it "stop working"? I've done re-installations here and there, but I can't think of a single time that it just "stopped working". And yes, after a couple months I usually do an in place upgrade for my laptops, desktop, and servers that have all been running Ubuntu for a few years.
Well, usually everything breaks, rendering the system unusable, requiring a reinstall. And than there was that time where the new kernel (somewhere in 3.x, IIRC) refused to boot on my setup (I'm running Intel x86-64, so nothing uncommon). But that only happened once.
I'd say history has proven the critics right on this one.