This article is strange: Some parts of it are sooo right, but then some parts are sooo wrong. Yes, the Linux world should stop obsessing about mobile and tablets, since they've basically already lost that battle. (And frankly: Who gives a damn about mobile? Some of us have work to do...) And yes, they should focus on their strengths in the desktop. But their strength is not web design and multimedia. As someone pointed out, they lost that domain to Apple a long time ago. Their strength is the scientific and technical desktop users. CERN uses Linux heavily. Many of the screenshots from the Curiousity landing showed a Linux desktop. I.e. people who need the computing equivalent of a truck, not a car. That's the core audience that desktop Linux should be trying to serve. And frankly, both Unity and the Gnome 3 Shell are steps in the wrong direction. Unity is clearly aimed at netbooks and tablets, and both Unity and Gnome Shell value simplicity over power, which is not what scientific and technical users want.
For my money, the recent moves Linux Mint has been making are the most promising ones happening in the Linux ecosystem right now: Don't fix what isn't broken, and make a distribution the just works out of the box. When the Gnome 3 goofballs eliminate yet another piece of useful functionality, do a fork. (And now they've started selling Linux Mint branded hardware, another smart move. One of the real problems with Linux these days is that you just can't buy hardware with Linux pre-installed that is equal in quality to Apple's hardware.)
You're right, I was being imprecise. I should have said Gnome and Canonical should stop obsessing about mobile and tablets. I guess I see more evidence of them obsessing about tablets and netbooks than about mobile. The thing that really bothers me is when I see the "workstation" experience get worse so that the tablet/netbook experience can be made better. I spend all day in front of a workstation, and maybe an hour a day with a tablet.
Well he said Linux rather than Gnome, not sure which he actually meant.
As far as I'm aware Gnome have never really tried to get in on the phone/tablet space. I'm not aware if there is even a phone that you can purchase with Gnome installed.
I am less concerned about getting Gnome or Unity on my phone and more concerned about an experience that you get between iPhone and Mac. A seamless integration. Their is no single media player in Linux that does that. libmtp doesn't really work.
I have thought about doing that but I would like to setup a media player which automatically generate playlist based on my listening habits like most played, recently played, etc. And it to work two ways.
Playlists can be very useful.
iSyncr and iTunes can do that but I haven't found a seamless solution for Linux yet.
>Who gives a damn about mobile? Some of us have work to do
Who gives a damn about these "personal computers"? They are toys for obsessive nerd hobbyists. Some of us have work to do.
Who gives a damn about this "multimedia?" It's all toy sound cards for vidgames and silly fragile CD ROMS. Some of us have work to do.
Who gives a damn about the consumer internet? It's all AOL sex chat and Pez dispenser trading. Some of us have work to do.
Who gives a damn about these "social networks?" It's for sex-crazed drunken college students, child molesters, and shut-ins. Some of us have work to do.
Who gives a damn about this "Linux?" Its a neo-communist academic experiment for people too poor to license a real commercial Unix (TM) or even Windows NT. Some of us have work to do.
I probably should have said that _I_ don't give a damn about mobile. I spend my whole day in front of a workstation, and maybe a few minutes a day using my cell phone. But still: Do you really think that Gnome and/or Canonical should be focussing on mobile?
Open source tends to work better in mature markets where there is something relatively stable to copy and where there are a lot of players who are incentivized to cooperate and adhere to standards. That's not mobile right now, and by the time mobile turns that corner Android (currently locked down on most installs in a very non Free way) will probably be the open source contender of choice.
From what very little I know about GNOME, it sounds like a great plan. I just happen to think mobile is extraordinarily important and look forward to the day I ditch the iOS platform.
There's a lot of similarities between a good scientific platform and a good creators platform. Text processing, image/signal-processing, 3d modeling/simulation, low latency, input/output devices etc. even things like storage.
Add to CERN and JPL every ASIC (chip) design shop in the world. My team is all on Ubuntu 10.04 and I dread the upgrade to the next LTS because Unity and Gnome 3 are not engineering workstation desktops.
How will Unity hinder your work, other tan having to get used to a new interface? I've had plenty of success using the latest Ubuntu as an engineering workstation desktop. It's better than everything else I've used.
I agree. While I'm not happy with the way the Unity switch was handled, this is definitely not a "KDE4" type disaster where entire apps were ported over with new bugs, and half the feature-set of the KDE3 versions.
I've been running Ubuntu 11.04 (first release with Unity) for a while as a workstation desktop, and while there are warts that I'm sure are dealt with in 11.10 and 12.04, it's mostly annoyances. Nothing that has seriously impacted my productivity.
Android is too bad to be Linux. It's a platform dominated by closed-source application software, a cathedral-developed kernel and core, and dubiously open bootloaders. It's utterly useless for development, and it makes numerous sacrifices on many levels to try to appeal to 'regular users'. When one pines for Linux success, one means GNU/Linux [and whatever DE or graphical environment you like]. Nobody cares what kernel your phone runs.
All fair points, but it depends on how you determine success. If you measure Linux-based OS'es penetration in the mobile market today, you will find that it 's the biggest player. That's success.
Yes, you have problems with locked bootloaders and closed hardware on some phones, American carrier-phones in particular. This is bad, but no fault of Android.
That most software on Android, a Linux-distro, may be closed source may sadden a FOSS proponent, but it still doesn't mean it's not Linux.
On my Android tablet I can fire up a terminal, hook up a keyboard via USB and then hack away in a Linux userland, using either supplied binaries, or busybox, or other Linux binaries compiled for the ARM architecture. And it will all work.
For lots of tasks where in the past I would need a PC, I no longer do. Because my Linux-based, mobile platform has me covered. If I want to build my own stuff, I can actually use the normal Linux toolchain to do so. I can do all that because Android, either you appreciate it or not, is Linux. There is no debating that.
And right now Android is dominating the mobile space. I think it's fair to say that Linux, in a form you appreciate or not, has succeeded where your traditional DE based Linux-environment has not.
There's a lot of closed source software for Android and many phones are sold with locked bootloaders.But it's possible to buy those without and google don't care if you root your phone.
So unless your definition is a system 100% open source software that is of no interest to "regular users" (not even Ubuntu qualifies) then this is unrealistic unless you are Richard Stallman.
As a full-time Ubuntu user who loves Unity and thinks it has a shot at becoming a third major desktop platform over the next decade, I think this kind of soul-searching is great for the Gnome project.
I selfishly hope this leads to better and closer collaboration between Unity and Gnome developers -- and maybe a 'partnership of equals' between Red Hat and Canonical, each playing to their respective strengths, on the further development and improvement of the Free desktop.
Unity felt like a large step backward from gnome 2, adding virtually no functionality and removing most customization. It's the ez-bake oven of window mangers. A toy.
Ubuntu/Gnome simplified the interface without giving the power users a way to go back. They should have made sure the fallback/classic modes were just as robust as the old to ease transition.
Developer mindshare is scattered due to the radical interface change. These developers want the older interface and were left something that was not in a very usable for most power users. Naturally, they moved to desktop environments that resemble the older interface (LXDE,XFCE,GNOME2) or started new projects (MATE,Cinnamon).
I find myself in a nomansland. I am constantly changing my DE because of all this bs. They bet the farm. All I want is stable and familiar Gnome2 DE! What a serious waste of time.
velodrome: FWIW, the transition from Gnome 2.x to Unity was neither quick nor easy for me, as I'd been a full-time Gnome user for years prior to the switch.
However, I find that I'm more productive now, with Unity, than I ever was with Gnome. In no particular order, I love how Unity (1) gets out of my way (I've set it to hide); (2) maximizes my usable screen real estate; (3) allows me to use complex apps like GIMP and Inkscape without having to remember their menu structures (!); and (4) lets me to do everything (including window placement) very quickly with the keyboard.[1]
Yes, Unity is different, but IMO it's also much better. Give it a six-month try!
Basically, the unity bar is crowded due to apps on other workspaces. Also, switching between windows of the same app is slower (I think I can deal with this issue though).
Performance-wise, gnome3 is snappier than unity (compiz kind of sucks).
You can make the Unity launcher (i.e., the sidebar) less crowded. Just run "Appearance" (<Super><A>, "appea", <Enter>) and change Launcher Icon Size to 32. I had to do this myself for the same reason.
You can also switch quickly between all windows of the currently focused app -- just tap <Alt><`>. (The <`> key is right above the <Tab> key on US keyboards.)
Also, note that the little triangular pointers on the launcher are different for applications running in the current workspace versus those running in other workspaces.
I agree, they should leave the streamlined gimmicky "we know what you want better than you do" approach to Unity and stick to making a developer's desktop. Canonical can invest money in Unity to pay devs - it's beneficial for them to get non-programmers as users. This is not the case for Gnome which relies on users to be programmers and develop it. If they can't use it because it's no longer for them then they will not contribute to it.
I completely agree with this. As a programmer HUD is the only thing keeping me on unity. Gnome 3 is _way_ more stable on all of my computers and I prefer it aesthetically.
I used GNOME for 10+ years and switched to the Mac a few years ago. The nail in the coffin was the Unity/Gnome Shell craziness that was so poorly executed.
Is it just me or this article outlining the most natural path for any Linux desktop? Rather then chase the general consumer, focus on your core audience who is already using your OS for the server.
Not only is it a more natural customer base, but you also have an advantage by eating your dogfood. GNOME devs would have 1st hand knowledge of what works and what doesn't for engineers and developers. And this sort of domain knowledge is gold when it comes to designing good interfaces.
I would love to see this effort happen, and might even go back to Linux if it really created something great.
I tend to agree with Paul_S that while GNOME would be better served focusing on the content production side of things, Linux on the Desktop can still be viable. The issue is that you need someone to throw money at it (e.g. Ubuntu/Canonical) and hire non-developer designers, maybe commission usability studies, etc.
Maybe once enough people are actually on Linux, non-developers with professional skill sets will volunteer their time to Open Source projects, and we won't need to necessarily use money to gain access to their skills. For now, most of them are on Windows of MacOS, and are only going to devote time to their current ecosystem, if at all.
There is only one thing stopping me from using Linux, hassle free hardware support. I want sleep/wake, video card, etc, etc, etc to simply work. And please don't pretend this is true already.
I've found that, with my last two laptops anyhow, most of the hardware problems only came up at install time. After I got the right packages (drivers and stuff), things worked well enough. Getting that set up was tricky, sure, but after it was configured correctly it worked.
So the simplest solution is not to install Linux yourself! You can buy a computer that has Linux on it already. I guess you could also get a friend to install it for you :P. The advantage with getting preinstalled Linux is that not only won't you have to figure everything out yourself, but you also will have hardware chosen explicitly for Linux.
Have you tried Ubuntu Certified hardware and encountered problems? Everything works fine on my Ubuntu Certified hardware: a ThinkPad and an OptiPlex desktop that both have Intel GPUs.
I had an optiplex desktop that did not work well. It was a giant pain in the ass to get video working properly, and it would often not wake from sleep.
OptiPlex covers a large number of models and hardware, just like "ThinkPad" does. I'd imagine that the Ubuntu Certified Hardware list specifies models that are known to work well.
He is missing the point that the Gnome projects he listed are indeed Gnome projects but the projects he mentions "we" should focus on are not Gnome projects at all.
It's not up to Gnome to make LibreOffice or Blender better. It's up to the Gnome team to work on overall user experience and a solid base on which the programs he mentions run. big difference!
He should probably not talk about Gnome but about Ubuntu or RedHat focusing more other stuff or to the libreoffice teams or whatever.
Also: So far i enjoy the Gnome experience very much. It may have still some bugs but all in all, in my mind, the Gnome team succeeded with Gnome 3.x and the new design/layout/workflow.
I think many of the Linux desktop's problems can be traced to the artificial walls between projects. Everything that ships on the GNOME OS DVD is GNOME's responsibility, including apps like LibreOffice. If an improvement needs to be made, it should be made in the proper place even if that is not "your" project. And if LibreOffice doesn't accept GNOME's patches then GNOME has to fork it.
i think it became pretty clear from the latest Gnome articles that such forks are very unrealistic. In my opinion it's the responsibility of the LibreOffice team to deliver a good user experience on the supported operating systems.
Gnome develops appropriate APIs and guidelines and it would be Gnomes responsibility to make sure LibreOffice has all the APIs it needs to work as good as possible.
Also, this is no artificial wall, this is reality. MS isn't responsible to fix Adobe Photoshop and Gnome isn't responsible to fix Blender.
I think it's also important to try and forge connections between software people like and your DE. Photoshop sells Macs even though there were many years when it was better-supported on Windows.
I think its a little misguided to focus on _media_ content creation, that market is locked up not by technical barriers but the industural attitude that if your not using photoshop your an amature. The big area where you can make headway is in the software/application content creation, but that doesn't require flashy new features that get you lots of good press. It requires you to make your enviroment faster and smaller.
Environments like that already exist, Xmonad/Xfce/Whatever. Linux already has massive developer usage, apart from in.Net/Mac/iPhone developers who wouldn't be able to use it anyway.
Media creation software doesn't have to be high level pro stuff. Having Gnome ship with software that is great for amateurs who want to make youtube videos, mess around with filters on digital photos, or record themselves jamming on their guitar would be a great start.
The existing software like GIMP would certainly be powerful enough for this, the main work would be in reworking it's difficult UI.
Right most recently people moved away form GNOME to Xfce because it was more streamlined and lighterweight, but that doesn't mean that GNOME can't do that better.
If you wanted to provide a refuge for the amateur then you are playing towards one of your weaknesses. The camera that I buy at BestBuy is not going to have instructions ons on how to use it with linux. If your not targeting teh professional you have already lost.
"Streamlined" and "lighter weight" are subjective phrases. I think the Gnome team are building what they would consider a streamlined interface, just their idea of what that is differs from some.
System resources are not much of an issue anymore, the amount of memory most window managers require compared with other applications is fairly negligible even with "heavyweight" Window managers. Unity+Nautilus together are using ~400MB/8GB on my Ubuntu 12.04 PC.
The hardcore "lightweight" WM fans only need a way to tile their terminal and XMonad already provides that.
With Ubuntu 12.04 every digital camera I've tried has just worked. I plug it in and and offers to import all of the photos for me straight away, no need for instructions or drivers.
Getting professional software would require either persuading enough of the big names in the business like Adobe , Steinberg , Avid etc to port their stuff to the platform or it would require Gnome contributors to create full equivalents for all of these programs from scratch with their already stretched resources so very impractical.
Creating an iPhoto type front end for GIMP seems a more achievable goal.
That ~400M of ram comes with other penalities then just thw space it takes up. This is the problem firefox got into. they just kept getting bigger continually saying ram is cheap and desktops have tons of ram. Just because I have 8Gb of ram doesn't mean I want my WM taking up all of it. What do I get for all that ram anyway. I get to have my desktop composed using javascript. This is why people left for XFCE this isn't a small subset of XMonad users, there are actually lots of people in between who want a windowed desktop, but don't want a huge heavy enviroment. GNOME can be that desktop.
I guess I wouldn't consider Gnome a "heavy" desktop, at least not any heavier than Windows 7 or Mac OS desktops.
I don't think ~400MB for a window manager on a dekstop is a big deal in 2012 and being able to compose your desktop using javascript is a breath of fresh air vs using old arcane APIs. When you have as much excess horsepower as you do in a modern PC (for most tasks) it seems wasteful not to use some of it to make your life easier.
What I don't understand is, Canonical has a big opportunity to serve the entire Workstation market. But what do they focus on instead, jumping on the mobile/tablet bandwagon. Do they honestly believe they are going to be a major player in that field going up against, Apple, Google and Microsoft?
If instead they'd focused on the Workstation market, which currently isn't really catered to by anyone serious, they could have got them selves into a nice little niche.
We (hn) have this gnome debate often enough it has been talked out to death. Nobody likes unity except for a few. Many of us using Ubuntu (me included) don't really care much about the desktop GUI, I spend most of my time in shells. Anyone that does care about the windowing likely switched to OSX years ago.
That aside, this article seems to think that new media production is being done on Linux. My experience is that this is owned by osx. Adobe's creative suite is the lifeblood of any media development and the Linux equivalents are not as good. If you're commercially invented to create new media then you'll use the best tools for the job.
Linux has contributed to media creation in rendering farms and such, but gnome played no part in that decision.
I wish people would stop putting this line without anything backing it up. We've had this conversation recently, and it turns out that actually not everyone hates Unity [1]. Besides my comments, here's a sampling:
"Unity is the best desktop I've ever used."
"I like Unity... I rarely go into a forum and announce how much I like something."
"'it's not as bad as everyone says.' This is the camp I fall in. "
"People that are happy with their desktop environment are not particularly likely to go to some site and talk about how good it is."
"unity is what you get by default on ubuntu? if so, i really like it"
"Never used Unity but I really enjoy Gnome 3"
Point being, you can't listen to people complaining then say "oh hey, there are 50 people complaining out of millions of users, this must be pretty awful". People will always complain more than they will praise, especially for a desktop environment. You might not like it, but you can't seriously extend that out to everyone save a few. Point being, people do like Unity and Gnome 3. But the people who would rather complain about something they don't like (and aren't forced to use) tend to be loudmouths. The silent majority just don't care.
So if you're stating that nobody except a few likes Unity or Gnome 3, I'm going to have to call your bluff.
To add more anecdote, I also really like Unity, especially now that I've remapped the HUD away from my Alt key (that belongs to Emacs).
I miss Unity so badly (well Linux in general) when forced to use Windows at work. I agree that the major focus of Linux should be science, development and data analysis. That's where a lot of the users are (its why I run Linux) and focusing on that would be, if not a huge market, a very dedicated and professional one.
I think you're probably right about most media production being done on osx, but I'd still love to have something comparable on linux, since you're not blocked by apple decisions as to what you can or not
access (no apt-get for instance) also with linux you get open source, better terminals/tools, and more software/lib packages.
Projects like Blender are amazing and others mentioned by the author are really almost replacements for Adobe, i really like GIMP's scheme scripting language for example which allows more flexability than proprietary adobe plugins. So I'd personally be really exited if GNOME started going in this direction.
Imagine for a second, linux with high quality media content creation tools, there's huge holes especially in video (nothing near comparable to After Effects or Final Cut Pro - forget ffmpeg for that )
Also as for myself I do all my development on Ubuntu / Debian while switching to windows for testing out on that platform (IE for example), but I only do browser/networking stuff.
Anyway, I like where he wants to go, and would love to see GNOME head in the direction...
Erm, I don't understand where this impression there's no high quality media content creation tools available for Linux is coming from...
I guess it depends what segment of the market you're looking at: 80% of the large VFX houses in the world use Linux to do the modelling, texturing, animation, lighting, rendering, compositing, using programs like:
Maya, Mari, Mudbox, SoftImage, Houdini, Katana, PRMan, Arnold, Nuke.
All those high-end apps that are used everyday to create feature films (and tv shows and adverts) work on Linux.
sorry, I should've been more specific... I meant linux in the sense that they're open source. The only contender in my book for high end open source media creation tools is really Blender which can hang with the Max/Maya lot. Also Linux tends to be used almost purely for the processing side, ie render farms, for film, 3d, etc. The designers and content creators themselves are rarely on linux though.
As I said in my post, 80% of the big VFX studios like DD, Weta, SPI, ILM, Dreamworks, DNeg, Framestore, MPC, FuelFX, RSP are using Linux for everything. Not just renderfarms - everything - creating the original content, modelling, texturing, lookdev, etc, etc.
So I don't know where you're getting the "Also Linux tends to be used almost purely for the processing side" from.
> My experience is that this is owned by osx. Adobe's creative suite is the lifeblood of any media development and the Linux equivalents are not as good. If you're commercially invented to create new media then you'll use the best tools for the job.
Which is exactly why media production should be a target for linux development?
Linux oses suck at so many things, why pick media production? Why not become the best enterprise system, or the best browsing system, or the best home media system?
For my money, the recent moves Linux Mint has been making are the most promising ones happening in the Linux ecosystem right now: Don't fix what isn't broken, and make a distribution the just works out of the box. When the Gnome 3 goofballs eliminate yet another piece of useful functionality, do a fork. (And now they've started selling Linux Mint branded hardware, another smart move. One of the real problems with Linux these days is that you just can't buy hardware with Linux pre-installed that is equal in quality to Apple's hardware.)
Or at least that's my two cents.