The Future of GNOME

Having just returned from GUADEC and now having the desktop as my job focus I been thinking about where we are going and what to do next. Benjamin managed to get quite a bit of press for his abysmally negative post, but I think his mis-stated the situation quite badly. As Dave Mason said during the history of GNOME talk, we have more or less always feel threatned and felt at times that we had an insurmountable task competing against systems with a lot more developers available than we do. There are some dark clouds in the skies, no doubt about that, but when hasn’t there been in the 15 years of GNOMEs existence? For instance the current tug of war between the GNOME shell and Unity? We have been there before, with Ximian and Eazel pushing competing visions for the GNOME desktop back in the day. Major corporate backers leaving (Nokia) or being in trouble(Novell)? Eazel and Ximian again. GNOME being perceived as being Red Hat only? That was the exact story that was being pushed before Sun and Ximian came on the scene. GNOME 3 turning away users? Hey, I can tell you that the amount of flames we got for GNOME 2 easily beats the GNOME 3 flames.

But for each of those events in the past we ended up bouncing back stronger afterwards, and I suspect we will come back with a vengance this time too. The thing is a lot of things are actually going very well. First of all the are a lot more shared projects with a healthy amount of resources behind them these days. WebKit is a great example of a project of crucial importance to GNOME, but which we share with a lot of other projects and companies. Take LibreOffice for example, yes LibreOffice do not contribute a lot of git commits to the GNOME git respository statistics, but thanks to LibreOffice we have the best and most full featured Office suite ever available to our users. And thanks to Firefox and Chrome we have browsers available with world known brands. And thanks to GStreamer, which is on freedesktop, we have a world class multimedia framework available. Instead of having our own sound server like we did with ESD, we now share a top notch sound server with all linux systems in the form of Pulse Audio.

Back during the GNOME 2 development cycle the GNOME project took an active decision that we would try to reach out to more of these external projects and try to work with them to integrate them into the GNOME experience. And we have succeeded at that, which of course have the side effect that instead of having all that development happen inside GNOME, we have bigger more active GNOME friendly projects doing these things outside GNOME.

We also made a conscious decision to focus on draining the swamp, which meant that instead of trying to work around issues and come up with funky configuration options on the user interface level to deal with missing features and bugs in the underlaying system, we instead have tried working with and contributing to fixing things in the underlaying levels. This of course moves attention away from ‘GNOME’, because people then commit their new fixes to those underlaying systems instead of implementing 100 000 lines of workaround code to ‘GNOME’, I mean I still consider Lennart Poettering a GNOME hacker even though his current work isn’t on anything in the GNOME git repository. But that does not mean that GNOME does not benefit greatly from these improvements.

So as we have succeeded in all these goals, GNOME should be a leaner setup than it used to be, as we don’t have to do all the heavy lifting inside the GNOME project anymore and that is a good thing.

As for projects such as Unity, well personally I don’t mind them at all. I mean I always saw XFCE as being a strenght and benefit to GNOME as it brought more users and developers to the shared library stack, and there is nothing different with Unity. And yes, there are some unfortunate duplications which seems redundant even in the context of differentiation, like the GNOME Online Accounts and Ubuntu Online accounts split, but hopefully we can work together to resolve such things over time.

And as for taking the desktop to the next level I think the main challenge for GNOME is not inside the GNOME 3 desktop, but instead it is a question of if the open source community can come together around pushing things like the Linux OpenGL support forward and make sure we have a strong foundation for supporting games and improving visual effects in the desktop. I for one really hope that Valve’s decision to support Linux with Steam, and the success of the Humble Bundle games on Linux, will help rally support behind the efforts to get Mesa to support newer iterations of OpenGL and get the open source Nvidia and AMD drivers in shape. Microsoft has long ago abanndoned PC gaming in favour of the Xbox, lets see if we can make Linux fill the gap.

In general I am very optimistic about the future, the strenght of the open source ecosystem is not that we have more developers to assign to a given project at a given time than Windows or Mac, but no matter what they do we keep moving forward on all fronts, so while Apple and Microsoft focus on beating each other in the tablet and phone space, maybe the time is ripe for us to strenghten our positions in the server and desktop markets? I mean tablets and phones might be all the rage atm, and maybe they are bigger markets than the traditional server and desktop market, but the server and desktop markets are still huge and ripe with opportunity if we play our cards right. And if we can deprive Microsoft of things like their Office cash cow, then that levels the playing field all the more when we later take them on again in other areas.


#1 thinkabout on 07.31.12 at 11:57

I also think that it’s better to continue concentrating on the desktop as it won’t go away soon. Let’s try to integrate all the new Mobile Options already based on linux or even windows or ios as good as possible into the gnome ecosystem. Syncing Notes, bookmarks etc.

I don’t think GNOME needs to be everywhere, I just needs to work with everything.

#2 patrick_g on 07.31.12 at 12:12

Hello Christian,

Could you please comment on this specific extract of Benjamin’s rant :

“GTK has 1 person working full-time on it (me). Glib doesn’t even have that.”

If it’s true I find this frightening.

#3 uraeus on 08.01.12 at 08:09

I don’t follow GTK+ development very closely, but it could be the case that Benjamin is the only person dedicated 100% to GTK+ atm. But that doesn’t mean he is the only one working on GTK+ or that it is a problem. A lot of people move around in the GNOME project based on what is perceived as the immediate needs and requirements, and GTK+ might not be where the major challenges are currently. That said there where quite a few talks at this years GUADEC about GTK+ related development, so to me it seems like development is moving forward. For instance Emanuelle Bassi (who was quite surprised to learn from Benjamins blog that he had left :) did a talk on GTK4 and Clutter and Owen Taylor did a talk about smooth animations in Applications.

Benjamin has to defend his own statements, but the fact is that both the people he claimed had quit GNOME was at GUADEC this year, while he wasn’t….

#4 aklapper on 08.01.12 at 18:21

The recent commit log (and the names of authors) can be found here, if you want to get your own impression:

#5 ebassi on 08.02.12 at 08:06

Benjamin is currently the only one that is working full time on gtk+; this doesn’t mean that he’s the only one working on it, and neither that if he stopped working full time on it then gtk+ would be doomed.

the truth of the matter is that glib and gtk+ have never seen more than a dozen people working on it, and rarely full time. people manage to work full time on aspects of it, for some time – because nobody can be expected to work on the entirety of the code base. in the end, though, we managed to get a 2.x -> 3.x transition with people working part time on gtk+, and that was the latest in the major achievements of the people participating in the project.

honestly: gtk+ existed before Benjamin was allowed to work full time on it, and it will continue to exist after he gets reassigned or decides to change his day-to-day work schedule.

#6 Felipe Contreras on 07.31.12 at 12:20

In other words; everyone that works in the Linux ecosystem is a GNOME developer? And if everyone stops using GNOME and switches to Xfce that’s a success of the GNOME project?

I am equally positive about the Linux ecosystem, but not GNOME.

Did the flames over GNOME 2 continue after more than a year? (I certainly don’t recall that) The GNOME 3 controversy even has it’s own article in Wikipedia. The big problem in my opinion is that GNOME developers have absolutely no idea if GNOME is doing well, or bad, is it gaining, or loosing users. That’s one of the problems of breaking user experience totally and completely. There’s only the subjective opinions of different people, no objective measurement. So, I would’t be betting either way.

#7 uraeus on 08.01.12 at 17:35

To some degree yes, if we end up in a world where everyone uses XFCE that is a success of the GNOME project, because GNOME is so much more than just the visual layer on top. 90% of GNOME is its library stack, which XFCE uses.

Take my own little application, Transmageddon, it will look and feel equally at home in GNOME, Unity or XFCE, which of course means that any application written for those 3 desktops has the potential of benefiting each 3 equally as they share most of the same infrastructure.

Personally I like GNOME 3 and think its our best shot at more widespread adoption beyond the traditional linux core crowd, which is why it is my desktop of choice and what I support. But that doesn’t mean I see anyone using something else as a failure or detrimental to the goals of GNOME 3.

#8 Jan Jokela on 07.31.12 at 12:27

I’de have to disagree with you regarding the bigger picture.

What drove (and still drives) GNOME is its mindshare. Years ago, it was all about the desktop.. a bunch of great people creating an alternative to Microsoft, and that’s something that created a certain kind of passion for the project.

Today we have a completely different landscape and GNOME doesn’t even compete in that. The Desktop isn’t going away, but it has become a little bit like a grandparent in a high-school party.

There was a time window for GNOME to enter the game. It passed by. And even on the Desktop the product isn’t a great one. Now, of course there is amazing work being done on stuff like WebKit and the Vala language, but we have a sub-par product. Aspects like whether LibreOffice moves to GTK3 or not is quite irrelevant, you can’t do magical user experiences with GTK3 and the LibreOffice team has never done anything more exciting than copying a boring 90’s era office suite. And this is true across the board: none of the new GNOME3 Apps are best in class, and I personally believe it’s a waste of time to deliberately create sub-par products.

As with Benjamin’s post, this might sound harsh, specially considering all the effort involved, but I care enough to feel like pointing out these structural issues.

#9 Alexander van Loon on 08.02.12 at 19:17

I’m tempted to say ‘don’t fix it if it isn’t broke’ because I think LibreOffice is a good product, but I can also imagine that some innovation would be good. But you should probably be more patient with the LibreOffice developers. AFAIK they spend (or spent?) a lot of time on cleaning up the messy OpenOffice code, better fix that first and let the innovation come later.

Can you elaborate a bit more on why the GNOME 3 apps aren’t best in class?

#10 Ferry on 07.31.12 at 14:06

Good post.

However, GNOME developers – while having improved GNOME massively – also have dumbed down GNOME way too far on some points IMHO.

The one thing that _really_ bugs me is the multi-monitor approach of GNOME3 and the fundamentalist/extreme approach of the GNOME developers in this (ANY form of fundamentalism is dangerous because it closes the mind)

The whole point of having multiple monitors (for me) is to have a larger surface to work with and _not_ to have those developers tell me I’m sick in the head.

GNOME needs a utility that allows the user to setup the screens. Have 2 or more linked together for a lerger surface, have a third as a (single) separate screen beside that, etc.
Designing the UI for this is not a difficult task.

#11 Michael on 08.03.12 at 07:35

Well, if you think the UI for that is not difficult to design, I am sure you can propose something that would cope with the various criticism that others will do.

#12 istok on 07.31.12 at 14:27

oh the future of linux is pretty safe. but i would not confuse that with the future of gnome.
i saw a slide the other day that in no uncertain terms suggested that gnome hopes to reinvent itself as a DE for mobile devices (“more focus on mobile”, while also listing “focus on the desktop” as one of the “bad things” about gnome). that is fine as mobile’s a big and expanding market. in doing so, gnome is losing the desktop (and by that i mean 15”+ notebooks as well), and that’s fine too, because there are many other good and usable DEs. but which mobile devices are running or will be running gnome shell, nay, “gnome OS”? to put it bluntly, it takes a google to wrestle away OEMs from other giants, and keep them. that’s the gist of the problem. it would be great if red hat stepped up and started forging these partnerships. will it? in the meantime it’s silly to run gnome shell on the desktop and normal size laptops – not impossible, but silly, because the work-flow is cluttered with steps and clicks that did not exist before. people feel like fools and reject it. the two most widely used distros, ubuntu and mint, have replaced it for their own shells and they are not coming back. but at the same time, gnome shell is not on any tablets/phones and there are no concrete or vague announcements that it might get there any time soon. so you got a project that has painted itself into a corner.
it’s nice to be optimistic, but perhaps not wise, considering.
and all that said… i’d very much like to be proven wrong. the failure of any big linux-related project cannot be good under any circumstances.

#13 Jeremy Newton on 07.31.12 at 14:37

Good to see someone else is as optimistic as I am :)

#14 AmbientRevolution on 07.31.12 at 14:55

I hope this gets as much press as the “Staring into the Abyss” post did. Somehow I doubt it will though. Press is good at sensationalizing the negative and ignoring the response.

#15 Michael on 08.03.12 at 07:35

of course, since that bring attention to their own post, and so values in the form of ads.

#16 orclev on 07.31.12 at 15:11

I really think there are two different issues at work here. One of them, is the amount of support the various libraries that make up the Gnome project as a whole receive. To me, this has never really been much of a topic of concern, people come, and people go, but any real deficiency I feel would naturally balance itself out because of the sheer number of projects depending on things like GTK+. The other topic, is the relevance and support for the Gnome 3 desktop, which I think is at an all time low. The primary reason the other article you mentioned received so much attention was because a lot of people used it as an opportunity to voice our dissatisfaction with the direction Gnome 3 has gone down.

You shouldn’t comfort yourself too much with the thought that you received more flaming during the Gnome 2 days because in a perverse sort of way, that was a sign of the greatness that Gnome 2 displayed. Enough people used and enjoyed the desktop to spend the time to gripe about what was wrong with it. Gnome 3 on the other hand, no one even cares enough to really complain about. I know I abandoned it long ago and have absolutely no intention to go back so long as it continues down its current path, and I suspect a lot of people feel the same. Why should I bother to flame about something I’m not even using anymore?

I’ve noticed a trend lately with regard to Desktop interfaces, and I can’t say I like it. Everybody seems so quick to jump on tablet and touch interfaces, and is so determined that they need to converge on one interface for both tablets and desktops, but I really feel that’s the wrong approach. Keep desktop UI mouse and keyboard centric, and keep tablet UI touch centric, there’s no reason at all to merge the two (same underlying code base is fine, but don’t unify the UI!), and plenty of reason not to. Windows 8 is making this mistake, there’s rumblings that OS X is heading down this path as well, please Linux, let’s not follow the idiots over the cliff!

TL;DR; GTK+ is awesome, keep up the good work (OK, it has some issues, but it’s still one of the better frameworks out there). Gnome 3 blows chunks, please go back to Gnome 2.

#17 Alson Kemp on 07.31.12 at 15:40

I’m a happy user of Gnome 3 (installed it when it first appeared in Debian Experimental… perhaps not my smartest move ever…) and I was a bit surprised by the unhappy facts in Benjamin’s post. I also took everything he said at face value, so was dismayed by his portrayal of the state of Gnome. I don’t doubt there are challenges for Gnome, but I’m glad that you’re aware of and excited about addressing them.

Thanks for the post.

#18 Anderson Cardoso on 07.31.12 at 15:49

great post, thanks.

I’m happy to know that the Gnome project is active and foward thinking.
I love the gnome 3 desktop. It’s by far, on my opinion, the best dektop experience on the market right now (I even sold my mac, because the osx interface was too boring and restrictive after I started using gnome-shell on linux).
Many of the complaints over gnome-shell its more about people being commodated on their bubbles, rather than the gnome concept itself.

Keep the great work, and know that the Gnome Team have true supporters who truly believes on this project.


Anderson Pierre Cardoso
São Paulo – Brazil

#19 Sandro Mani on 07.31.12 at 16:25

“so while Apple and Microsoft focus on beating each other in the tablet and phone space, maybe the time is ripe for us to strenghten our positions in the server and desktop markets?”

How true. Lots of people like to do one thing: working with a familiar user interface and get their job done. I still wonder how on earth Microsoft thinks any person will be able to work with their Metro interface, considering the number of office-pc’s which run windows. So this is exactly the opportunity Gnome needs to take advantage of: focus on the familiar traditional everyone has learned to use and appreachiates, and deliver a rock solid desktop for getting one’s job done. So, let’s stop this tablet nonesense. Tablets are not here for working, but for leisure. Stop forcing all the social media nonesense into the desktop, leave it optional. Stop trying to be masters of our disorder by introducing “tracker” and similar programs which hog our resources by indexing everything. And mainly, this touch interface nonsense.

#20 Michael on 08.03.12 at 07:42

Working with a fammiliar interface is the exact reason why people try to reuse ideas from phone, because the mobile phone interface is what is familiar to most people ( only because there is several people who have a phone and no PC, especially in some country ).

#21 Nathan Fiedler on 07.31.12 at 16:27

Keep up the good work. I’ve used GNOME for what seems to be an epoch and have been thoroughly happy with it. Ubuntu and Unity are nice looking, but I more often use OpenIndiana and prefer the desktop the way GNOME has done it for years. Thanks.

#22 l.kot on 07.31.12 at 16:34

As long as there is a goal to “improving visual effects in the desktop”, GNOME will not have me as a user.

GNOME will only regain me as a user when it focus on allowing me as a user to adapt the environment to my needs, my work process and my habits.

Currently GNOME is more like a tablet bastardization mishmash of Android and OSX, which for sure fits my grandma maybe. Or illiterate brother.

So far, XFCE and LXDE is far more usable and gives me a productive environment on my machines.

#23 anes on 08.09.12 at 05:48

Well, if it fits your grandma then that’s a sign of GNOME doing the right thing :) I’m just curious, what is this ‘productive environment’ people talking about?

I think the problem is that GNOME 2 is so great that people refuse to learn new habit with GNOME 3.

Not all things going perfectly well, but I still use GNOME 3 and happy (and, I can say, productive too)! I still have high hope that there’s great things to come in GNOME 3.

#24 Jimbo Smiths on 07.31.12 at 16:37

So it’s ok that GNOME is perceived as RedHat-only, because that’s always been the case?

So it’s ok that the GNOME brand is losing credibility, since other unrelated brands with healthier communities are doing better?

So Steam’s incredible popularity is evidence that Windows gaming has been abandoned?

Don’t say things like “Microsoft has long ago abanndoned PC gaming”, when they’re putting in more work than everyone in Linux world combined ever has. They haven’t abandoned the PC for the XBox, because game programming for both is one and the same. Compare to linux, where game development isn’t even compatible between distros, and the only system where gaming actually works is Android.

#25 Tiao on 07.31.12 at 16:43

Ok. You just changed the point of discussion.

It’s not about doing a rock solid environment, its about HAVE enough people to make the project go on.

And on the same way, GNOME is disapointing a lot of people that still believes the solution would be go back to Gnome 2.x, or use one of aaaaaall forks that intent to bring back the good experiences of Gnome2.

All the gaps that you’ve talked about exists on KDE, and they move forward on the same way. It’s not a “Gnome excluse dificulty”.

#26 rune01 on 07.31.12 at 17:39

> And thanks to Firefox and Chrome we have browsers available with world known brands

So what is the reasoning behind pushing the Empathy/Web web browser? Couldn’t these resources go to ensuring tight integration between Firefox and GNOME? Isn’t that somehow duplicating efforts?

#27 uraeus on 08.01.12 at 08:12

Could be, but you know GNOME is a volunteer project and thus people often work on things that you or I doesn’t consider priorities :)

#28 Michael on 08.03.12 at 07:45

Maybe because firefox release too fast to unsure tight integration, or because they are not interested into such integration ? ( because that’s a work for the 2 groups, and you cannot force them )

#29 Mike Linksvayer on 07.31.12 at 17:57

there are a lot more shared projects with a healthy amount of resources behind them these days.

we instead have tried working with and contributing to fixing things in the underlaying levels.

Thank you!

I’m someone who really doesn’t care about precisely what the desktop looks like or how I interact with it (and I switch fairly regularly among GNOME, KDE, LXDE, Unity and Xmonad on a few different computers). I want the desktop to facilitate my control and use of my hardware and OS — if sound, suspend, monitors, networking, software installation, etc, work flawlessly, fantastic; if not, total fail. Second, I want to use applications that others use too so that my work is not hindered; Firefox and LibreOffice have been the most important there. It makes me a little sad that there’s so much gnashing of teeth over the rest.

Thanks again for focusing on the right things, and please see how this focus can be taken further! :-)

#30 anon on 07.31.12 at 17:59

i’m pretty sure gnome will go on (worst case turn around) and get to the point where it was usefull as gnome2 (or maybe gnome2+compiz) was, with all the added benefits of the gnome3 architecture. but as it stands, gnome3 is pre-alpha. both gnome-shell and unity are bad jokes when forced to the end user that is not wishing to contribute to the project.

I know because i contribute some JS code and even knowing all my work is just to get me what i had with gnome2 i understand the new added flexibility for the longer term (btw, install via the browser… baaaaad idea) and also see my wife using it for mundane tasks that have to be completed in place of opening a bug report.

funny how the best office suite is very useless thanks to gnome shell… you can’t alt-tab. try to copy several lines from one document to another in a notebook monitor. it’s HELL.

or how all the usability options simply vanished so it’s now a mac clone. focus follow mouse? sorry, steve jobs didn’t liked that. …thankfully most developers are annoyed by those and so they are coming back quickly.

So yes, the other post was pessimistic, but this was bit on the delusional side. As it ignore the only bad thing talked on the other article. user base complaining of the actual product. not about misconceptions. nobody care if gnome3 is ubuntu only or not. the fact is that people exposed to it, complains. the switch was turned too soon.

#31 Taryn Fox on 07.31.12 at 19:33

I very much appreciate the historical context! It makes me more optimistic.

I don’t feel our goal should be to “beat” anyone, though. We’re filling a need for a Free Software OS, and creating a useful thing of beauty that is meaningful to us as contributors. It’s useful to take into account what “competitors” are doing so we will know how to differentiate ourselves (or what features our users will expect), but our success does not have to mean anyone else’s loss. We can’t necessarily predict whose business model will be disrupted by what we do.

#32 Jack on 07.31.12 at 22:14

I couldn’t agree more. Thanks for sharing the other side of the coin.

#33 Ernest on 08.01.12 at 13:04

XFCE can be configured to look and behave like Gnome 2. It takes some time to do it right, but it can be done. Also, it loads noticeable faster and the panel applets don’t seem to have any memory leaks, unlike some others I could mention (Gnome’s hardware monitor :-). All in all I’m pretty satisfied, after almost a year of using it daily, although of course there’s room for improvement. The problem with Gnome 3 in my opinion has been a total lack of product pre-testing. Either that or they are targeting a demographic totally different from that of the majority of GNU/Linux users. To Gnome UI designers I would say before designing anything you should have spent some time with various long-time Linux users, study them, watch how they use computers. Unfortunately they didn’t do that and the result is a UI that does nothing but alienate those users and make their lives more complicated without providing any real advantage at all.

#34 DMJC on 08.01.12 at 20:25

Microsoft is going to implode itself trying to fight Google and Apple for mobile dominance. The time is perfect right now to attack their main revenue streams. Samba 4, combined with decent OpenGL/games support strikes at both their desktop and Server markets. If Linux can score some big wins here it will deny Microsoft a lot of money and push Unix a lot further forward.

#35 Bokal on 08.01.12 at 21:05

I love Gnome 3, and it’s the DE that makes me the most productive. And every time I hear someone saying it added clicks, I feel like “Did you even try it? stop clicking and use your keybord! it’s damn fast!”

(Just so you know not everybody shares the rant)

#36 Horace on 08.02.12 at 10:42

I use Gnome 3 everyday, and I believe it’s the best Linux desktop at the moment. I hope GNOME shall not let all the naysayers who never really gave it a chance dictate its future direction.

#37 Alexander van Loon on 08.02.12 at 19:10

With all this criticism I feel that I should I stand up in support of GNOME. The different approach of GNOME 3 didn’t bother me, as a user of GNOME 2, KDE 4 and Windows 7 I was able to adapt easily. And I’ve stuck with GNOME 3 for more than a year now. My mother uses it too and is satisfied.

I can understand that some people are disappointed with the disappearance of the window list, using Alt-Tab and the Windows key did take some time to get used to. Regarding the so-called ‘dumbing down’, the only thing I’m missing is the ability to change font settings, that’s the only thing I use GNOME Tweak Tool for.

Recently I finished writing a master thesis which cost me a tremendous amount of time. GNOME 3 has served me tremendously well. I’ve been using Epiphany, Evolution, gedit and Evince day in, day out over the last weeks. Just like many other pieces of free software such as Firefox, LibreOffice, R and LaTeX. Sure they’re not perfect, but they got the job done and I’m satisfied.

I don’t see myself writing my master thesis on a tablet or a smartphone. Heck, I don’t see myself using anything which doesn’t have a keyboard for any writing work. Sure, mobile devices are growth markets, but saying the desktop is going to disappear is utter bullshit. In my opinion there’s nothing wrong with steadily improving the desktop experience and not entering the mobile or tablet market at all, those will hopefully be served by Mozilla’s Boot 2 Gecko.

GNOME developers, you are all my heroes and you’re doing the right thing. Keep up the good work.

#38 Ale Abdo on 08.02.12 at 20:00


In my experience as a user, GNOME is doing well and has delivered great things in recent times.

Still, questioning and reflecting about it does good. The thunder passes, the rainwater stays and feeds the land.

I also agree with Mike L about the importance of strengthening the foundations, as the post suggests.