2. If Gnome really wants to win back the hearts of potential contributors (i.e. power users), they'd better make programs that appeal to that demographic. People who have the skill and motivation to make significant contributions to a free software project often want a lot of room for configuration, including the option to use the desktop in a traditional manner. Taking away those little checkboxes and toolbar buttons is like slamming the door on power users. You might win a billion non-technical users, but none of them will ever submit a single patch.

3. Gnome is too big for its own good. Why does a desktop environment project need to maintain a complete stack of apps and libraries, from GTK to Gnome Shell to a text editor to a bundle of games to a web browser to an email client to a media player to a full-blown spreadsheet app? Why can't they just tell people to get a third-party browser? They should spin off the rest and focus on GTK, the Shell, and a small number of essential utilities. If Epiphany or Gnumeric died a slow and lonely death, how many people would really care? Heck, if you don't have the manpower to maintain anything else, just give me GTK so I can install xfce or lxde on top of it. It's really just Firefox and LibreOffice and VLC that I want, and I don't need Gnome to run them.

Edit: some rephrasing.

Anecdote: I work at a small actuarial firm that uses Linux desktops, and when I migrated everyone (ten people) over to Ubuntu 12.04, they all loved GNOME 3.

I am still involved in the community, I am a director of the foundation's board, and I'm still working on Gnome projects in my spare time - which is actually easier these days since I moved from intel to mozilla.

I'm typing this from GUADEC 2012, in A Coruña; the conference is absolutely delightful, there's a lot of talks about direction and future involvement, and everyone here is really excited about moving Gnome forward, as well as regaining the enthusiasts market.

not everything is bleak and bad.

So... after 2-3 hours of Gnome3, I had had more than enough to prompt me to bite the bullet and switch to KDE. There is nothing good I can say about Gnome3... trying to use it was painful in about every way I could imagine. Nothing works the way I expect, and nothing was intuitive at all.

KDE, on the other hand, has been a pleasant surprise. I'd dabbled with it 10+ years ago, but never made the permanent switch... and given that they had gone through their own "change everything and piss off all the users" thing a while back, I wasn't sure what to expect. But after using it a couple of days, I couldn't be happier. It took a few minutes to figure out some of the new approaches they've adopted but - by and large - a little trail and error, some exploration, and intuition, and I was back to productive work almost immediately.

I have no hard feelings towards the Gnome team or anything, but they're just trying to go in a direction that I'm not interested in. Best of luck to Gnome, but KDE is a clearly superior choice for me right now, and I'm thrilled to have made the switch.

While Gnome 2 was gradually approaching ultimate goodness with its essential configurability (not too many knobs but an explicit set of gconf properties that you could tune if you wanted to), consistency, ease of use, and ten years of GTK2 providing a platform for applications that look and behave uniformly, it was certainly lacking in the integration side (networking, messaging, etc.) for which Gnome 3 is a response.

However, Gnome 3 broke so many little things that it doesn't matter what the new features do. This is one of the cases where Microsoft has been right: when you're big enough, don't muck with backwards-compatibility.

Gnome 2 was an excellent basis for a desktop because there were so many ways it could be used. It didn't limit you; you were free to trim out things you didn't want, or add things you wanted. Aside from that, it got out of the way.

Gnome 3 aims for the same demographic, but tries to force too many opinionated decisions on its users from the start. Distros that care about branding hate that because it makes it hard for them to differentiate (see Ubuntu). Many users hate that because it's too many things that they can't change without learning a new Javascript platform. It's certainly fair to say that typical Linux users (their target market, whether they like it or not) aren't used to that.

This post is wrong to blame tablets and smartphones for the decline. Linux users aren't abandoning desktops; they're abandoning Gnome 3, because it isn't giving them what they want. It's that simple.

In my view Gnome should try to emulate the android model. Build the whole stack up to the widget level (they are really good at this) and then publish a couple of apps and an app store.

http://rhel7.net/news/98-rhel-7-roadmap

which means we might see RHEL/CentOS/Scientific Linux/PUIAS users on Gnome 3.6

http://fedoraproject.org/wiki/Releases/18

Personally, I've worked out I spend little time actually using the overall desktop GUI, so I'm not too sensitive to changes in UI logic. Typical end user I suppose.

I suspect the change to systemd will cause rather more fur to fly than the GUI in Enterprise circles.

I hope the author of the original article gets a bit of support and finds a direction for his labours.

I personally am a fan of Gnome, and hope that they become relevent again. Unity is too fancy for my linking, and Unity 2D, which seems an option, is now deprecated! :|

meh, other than bus factor this seems overblown to me. I use Gnome3 on Fedora and it is great. It takes advantage of Fitts law and there is just less fuck-around-ability with it. (I thought even the Alt-to-PowerOff controversy was overblown; I suspend far more often then I shut down, this is a welcome simplification) If anything I felt Gnome3 had ushered in a renaissance in Gnome. What "new" goals does Gnome have to have other than creating a great desktop?

Please don't drive more developers away with more gratuitous Hacker News "X-is-dying" bitching.

I remember when KDE first came out -- I heard the first press releases and thought they were on drugs, but when I downloaded and built, I was like "wow! this is so close to being a commercial desktop"

Now, in 2012, we have KDE and Gnome and a few off-brand desktops and it's still like "this is so close to being a commercial desktop" -- but there isn't any Wow anymore.

Red Hat didn't like the license of the QT toolkit, so they had to go out and build their own desktop, which was probably the most disasterous decision in the history of Linux -- it's like Windows Vista without Windows 7.

For a long term all of the major linux distros have been wasting time and resources trying to make linux something nobody cares about. There's an obsession, for instance, about office suites that are so bad they make Microsoft Office look like a paragon of reliability and ease-of-use.

On the other hand, there's been a complete disregard for the people who ~really~ use Linux such as sysadmins and developers.

I've recently set up two laptops that run Windows as a host and Ubuntu linux inside Virtualbox. I use "putty" as my *term program and Cygwin/X to run the occasional GUI app I need from Linux. It's a sign of what a disgrace the Linux desktop is that putty has the same ease-of-use and reliability that xterm had 15 years ago, whereas the "terminal" program that Ubuntu tries to push on you is a bloated disaster in which cut-and-paste is as miserable as it was in Win 3.1.

I've been using it exclusively for months and I feel that it is doing what Gnome3 wanted to achieve. It is simple, intuitive and rock-solid. Never froze or crashed and diagnostic tools are great.

LinuxMint is the old Ubuntu which you used to recommend to your friends and family.

1. When I last used GTK (about 2 years ago) it felt too big, old and bloated. If GTK were simplified and followed Qt's lead into scripting and easier interfaces (eg. Qt Quick), plus a MIT or LGPL license, this would encourage a new culture of apps.

2. I hated Gnome 3 and Unity for that matter. Gnome 4 needs to take a step back and get out of the way. You don't use the computer just for Gnome, but you use Gnome as a stepping stone. All common apps should be one click away and everything should be as customizable and flexible as possible.

As kljin said, some Gnome apps are redundant and the workforce could do a much better job focusing on the core issues, that could bring more people back to Gnome and hence possibly continue these projects again in the future.

I found Awesome, and though there are some things I don't like about it (or namely, some programs that don't work nicely in a tiling environment), every time I try to use another WM or a DE, I miss the speed and keyboard accessibility of Awesome. It's a blast to use.

I installed KDE for my wife to use. It takes KDE an order of magnitude longer to start than Awesome, for all those services and stuff that it's running--you know, all that stuff I'm probably not even using.

We have Qt that really works, much much better, and with LGPL license.

You can run stellarium, VLC or Marble in Mac or Windows without problems.

If you try to use gimp or inkscape in mac it opens x11, copy and paste does not work(in inkscape it copies pixmaps instead of vectors!!), what a botched job.

In windows you will have a lot of problems too.

GTK support for OpenGL, OpenCL was terrible, having to low code everything, while in Qt works as well as with cocoa.

Let GTK die and improve(or fork) Qt.

After installing I had to go back to GNOME 2 (on a client - installation problems) and if felt old and boring - like going back to Windows XP (from Windows 7)

I "just" did a default install with the expectation that Unity would drive me crazy and that I'd end up replacing everything, but in the end I found it very suitable for how I work (browser + terminal with tons of tabs maximized; leaving one workspace free for the rare other apps I end up using)

That said, Cinnamon is a really nice fork of it, that has the features most users are missing, and while it does bring back some of the win98-ess, it's good to see it evolve and become more stable, I'm still undecided between the two and switching back and forth every once in a while.

If you ask me, I think GNOME should drop most of its peripheral software, and stick with just making a really really stable and sleek desktop. Most people don't need email and calendar clients anymore. I just want a simple yet configurable desktop, a great file manager, a good text editor, and a console. The rest should be external IMHO, and the desktop should provide easy integration points.

One important function of the Start menu is discoverability of apps. If I want to see what junk I have installed, I can look through the menus to see what I have. With Unity you have to bring up the whatever and try typing search terms. If you don't know what you're searching for, it can be difficult -- and lots of times, you don't know the name of the application, because they have clever branded marketing names (Evolution, Firefox, GIMP, Inkscape, Shotwell, Chromium, etc.) which don't have their primary function ("image," "photo," "web," "email") as a substring.

Discoverability is particularly important for "control panel" type system administration applets, which are often vital to making your system function acceptably, and whose name, number, hierarchy, and division of functionality seem to mutate with every release.

Also, there are switching costs. The Unity interface is so foreign, I'd need several days -- possibly weeks -- to get as proficient with Unity as I am with Windows, Gnome 2 or Cinnamon. That's definitely a cost in time and frustration, and the benefits aren't clear.

Some features -- like the Mac-like "there's only one instance of each application" -- seem designed to cater to n00bs who need hand-holding because they don't understand the concept of multiple application instances, or the difference between launching an application and switching to an instance of that application.

I want multiple instances of certain applications -- terminals particularly -- and it's a major pain point with Unity. So not only do I have reduced productivity during the transition period, it seems like Unity is actually going to decrease my productivity once I do learn it, due to lack or hiding of core features.

Add to that the fact that Unity would crash regularly within the first hour of use when 11.10, the first Unity-only version of Ubuntu, was released.

I gave it a fair shot on two or three different occasions -- I think once when it was still called Ubuntu Netbook Remix, again when the beta was released, and finally with the official release of 11.10. (And a few incidental times when I've booted the Ubuntu CD for various reasons.)

In each case, within an hour of use I've concluded that Unity is a nightmare.

    GTK has 1 person working full-time on it (me). Glib
    doesn’t even have that. I think evolution is in a 
    similar situation (a complete email client).

> You don't use the computer just for Gnome, but you use Gnome as a stepping stone.

> All common apps should be one click away and everything should be as customizable and flexible as possible.

This is exactly what they chose not to do.

https://mail.gnome.org/archives/usability/2005-December/msg0...

In a reply to Linus Torvalds:

> Sorry to snip mid-sentence, but this is an important point: We're not aiming

> for "powerfully extensible". We're aiming for "Just Works". Some people will

> hate that. Some will love it. Personally, I'd rather have passionate users,

> lovers and haters, than be than average and ignored, and I think you'll find

> most GNOME developers feel the same way.

So they made their choice, ignoring most of their own base were against it. Guess what? they lost users left and right.

They had a sweet spot in Gnome 2 and a strong user base. They thought they could somehow get Windows and Mac users, but that wasn't just happening. At least not at that particular moment in history.

So that was their decision and they didn't backtrack despite all the warnings (plenty during the last 7 years). Stubbornness gets you to obsolescence.

Gnome feels un-unified. I guess this has always been the way under Linux, what with QT, GTK and other toolkits, but when one app is slightly at odds with the rest of the Desktop - it feels, odd.

Example being non GTK3 apps, like LibreOffice (though someone here suggests that that is being rectified.) Even VLC, Opera, Firefox and Chrome feel a little out of place. Each behaves differently. You can't quit VLC with CTRL+W for example. Each are designed on different toolkits. Menus are inconsistant. Tabbing behaves differently in each app (can't we relegate this to the Window Manager or Desktop?) Keyboard shortcut unification doesn't exist. These are the edges I'd like to see addressed across the Linux desktop as a user.

Perhaps unification is a lofty target. And we should just be happy with the fragmented cottage patchwork.

I don't even know the difference between GTK, GTK2 and GTK3 and QT! My desktop is such a pain to theme it's a nightmare, I certainly notice that. What's new in layman's terms in GTK3?

As for some core apps, Evolution looked promising. But even that feels a little rough around the edges (I can crash it quite easily.) Thunderbird doesn't integrate with Gnome brilliantly. One flagship email client would be nice.

Focus on the core, the desktop design guidelines and some intrinisically needed apps. Most desktops on Linux seem to suffer in much the same way. Unity still appears ad-hoc.

I guess a good aspiration would be to make it as simple as possible for people to create applications as well as use them under Gnome.

Could there be some kind of CSSification of an app's controls? Present them as interfaces that could be styled differently according to the platform you are on. Leaving Window Managers to take on the role of innovative desktops. Perhaps apps are designed like this already? At least it would be easier to port an application across different form factors.

GTK is LGPL licensed.

OpenGL support sucks in GTK and QT equally. Especially if you want a more recent version of GL.

OpenCL does not depend on your widget toolkit in any way.

> Let GTK die and improve(or fork) Qt.

Qt is not really as great as you seem to suggest. It's a bloated "batteries included" framework that has everything from it's own string type to wrappers for things like threads and sockets. GTK is a widget library that is a lot leaner and meaner in comparison.

These massive platforms that try to wrap everything in the underlying operating systems to a common API use only the least common denominator of the systems that it runs on. The easy 80% works quite well but things tend to fail miserably when you enter the 20% realm. Things like memory mapped files or asynchronous socket i/o tend to be missing from these wrapper platforms.

And Qt is still using an old non-standard conforming version of C++ to achieve portability to things like Symbian. And their own C++ extensions that require you to use their build system and pre-processing tools. So you can't really use Qt without going all the way to Qt land.

If someone did for Qt what XFCE/LXDE did for GTK+, I think it could catch on.