gnome in the age of decadence

7 June 2008 11:25 AM (gnome | decadence | guadec)

It is with some trepidation that I go to buy my ticket to this year's european GNOME conference, GUADEC.

Decadence \De*ca"dence\, Decadency \De*ca"den*cy\, n.
[LL. decadentia; L. de- + cadere to fall: cf. F. décadence. See Decay.]

A falling away; decay; deterioration; declension. "The old castle, where the family lived in their decadence." --Sir W. Scott.
[1913 Webster]

Take a look at the list of slated talks. What is your general impression?

Mine is of a large project in a state of marginal returns, in which a larger and larger part of the effort goes to maintenance. On the one hand you have the large deployments, the integration with other software projects. On the other hand the new developments that we have are very careful not to bite off too much: a printing dialog; another revision of ekiga; a new image library. Ed: Ekiga was probably not a good example of this.

The problem, as I see it, is that GNOME is in a state of decadence -- we largely achieved what we set out to achieve, insofar as it was possible. Now our hands are full with dealing with entropic decay. Take, for example, Evolution's random walk to improvement. In most releases it's better, in a few it's worse, but basically it still works fine, and has been that way since about 3 or 4 years ago.

It's like, welcome back to 1984's Macintosh plus interweb. We did it!

Seriously though, it does not seem to me that GNOME is on a healthy evolutionary track. By that I mean to say that there is no way there from here, if "there" is universal use of free software, and "here" is our existing GNOME software stack. The evolutionary thing to do would be to do something web-like, because that's where all of the programmers are these days. But that's not part of our culture. Until recently, with WebKit/GTK, it wasn't part of our software stack either -- all of the new web platform bits were dribbled to us over the wall from Mozilla, or embedded within GNOME as Firefox.

The other side of that is that while the web will be a core part of the computing future, it's not clear that swallowing it wholesale is the best strategy for client development. There are too many things that local computing offers: other software paradigms (emacs, unix, independence), unmediated input (sound, video, alternate input devices), direct access to powerful output devices (control part or the entirety of a screen, access the underused GPU).

But even if we eschew "going with the flow"-style evolution, it's not like GNOME is on a revolutionary track which will win in the end with its compelling UI or programming-linguistic metaphors. The screen is still constructed as a static landing strip on which the mouse pointer might alight, an array of possibilities necessarily constricted by decontextualized space. The metaphors are the same: file, folder, desktop, even as these things cease to exist for many people. And techologically, we don't even have a way of considering how the visual elements of space might be anything other than static, much less have any way of interacting with those elements other than the impoverished point and click. What we're left with is the GUI equivalent of chartjunk.

There are exceptions to this story. There's Clutter, which I expect will be the "way out" both for GTK+ and for GNOME. (I know there are technological differences with other canvas models, but at least they have the hackers and the maintenance resources.) There's Moonlight, which is interesting and hacked by very smart folk, but whose fortunes are too bound to Microsoft. There are heroic retrofix efforts like MPX, not really a part of GNOME. But other than that, we have the decay of slavish adherence to the HIG, the logout dialog, the wallpaper chooser, the last-percent efforts of refining an increasingly irrelevant stack of software.

The GTK+ maintainers are well aware of the decadent state of GTK+, and are moving as much as possible to plug the leaks. But it is no longer a nimble codebase, and will take at least 6 and possibly 12 months before a 3.0 release can come out. And that's just stopping retrograde motion; actual construction must take place outside of the "core" until the core is ready for it.

This disenchantment is personal as well: among other things, I've spent thousands of hours on bindings to GNOME libraries, and just now when I am ready to make an API and ABI stable release, I just don't feel like packing another button into another hbox.

Anyway, I'm buying my ticket, but mostly for the hallway track -- GNOME folks are smart and kind, and I want to see what's going on, what people are really thinking about. Istanbul ho!

48 responses

  1. Mårten Woxberg says:

    Good post. I started using Evolution when it was still Ximian handling the reins. But just about (a bit before) when Evolution got moved to GNOME everything went downhill. I stopped using it and went to Thunderbird and now from Thunderbird to Gmail. I see no need for a mail application anymore. What I need is a good calendar application that integrates with google calendar so I can have the calendar with me. Evolution hasn't evolved in the last 3 years.

    It seems to me that there's no plan on where to go next. The KDE people seem to have a vision on where they want KDE4 to be.

  2. Juanjo says:

    I agree with you, partially.

    I agree that in some tools we don't see a clear evolution. Sometimes a tool get unmaintained and outdated, and another tool with less features remplaces it (with it's own problems an leaks). The Evolution example is perfect, IMHO.

    Seems that Gnome misses 'a big picture', where all together work forward.

    In the other hand I like a platform to work that remains uniform in time and doesn't shake my world in each release.

    Perhaps exists a middle point.

  3. Karl Lattimer says:


  4. Ronald says:

    I agree. It's not something of this year alone, GUADEC has always had more, let's say, free-form random thinking rather than big-picture work. The work that stands out tends to either be accepted and be uninteresting for such talks because it is already mainstream, or be developed in silence while others speak about their projected nirvana at GUADEC. Talk or walk, so to say.

    GNOME has great parts, and GNOME now needs integration. Integration of N components is (theoretically) an N^2 task, which is probably why there's so little interest in it.

  5. sottix says:

    Which is the so called "new next great vision" behind KDE4? Rewrite all with QT4 and expose it's new features? They are doing a great job but I really fail to see where "new great vision" is....
    I think "vision" has a more powerfull meanings beyond rewrite all the platform in another way....
    So who has the "new next great vision"?

  6. Mike says:

    Mostly agree, Evolution is certainly the Elephant in the room. GNOME/GTK+ is slowly being reworked from the inside out, but radical ideas like implementing a document-centric UI and/or a more naturally interacting desktop are continually put on the backburner because of the idea that "10 yr old hardware can't support those features". From others theres a real apprehension to change, people sounding like my grandfather who still loves sitting in front of his 10in 50 yr old tv ("What's with all this glitz? I want something that works". Well gramps the glitz does work or will if its tested properly and offers a better model of human interaction). I think a balance needs to be found where inclusion of new ideas ignites the excitement and creativity of the project again while also keeping some old stuff around for the grandpas of the project and old hardware we insist on supporting indefinitely

  7. ooi says:

    I don't know about you people, but writing web apps was a crappy experience for me (HTML/CSS/Javascript/IE bugs/...). I'd take coding native GUIs anyday over webcrap. Even if it means GTK+ in raw c.

  8. Dean says:

    Looking at the schedule I see plenty of evolution:
    Mobile stuff with maemo and openmoko
    Multiple discussions on canvases
    Lots of stuff around online content - "people browsers", integrating telepathy, synchronisation with various web apps via conduit and desktop-data-model

    There's a couple of lightning talks on document centered desktops too.

  9. Mike says:

    if you look at old GUADEC schedules ago you'll see about all the same talks are listed and unfortunately they haven't changed much... From my perspective it seems like a lot of talk without much action.

    To they're credit, some projects have been working tirelessly and making steady progress and (fingers-crossed) they should become more usable soon (Telepathy, Clutter/Pigment, Webkit). They'll be an asset to the future of GNOME.

  10. Havoc says:

    I still like this post about how to create a GNOME direction:

    The bottom line is that to innovate you have to start from what you want to do and who you want to do it for, not from some existing pile of code.

    GNOME should explicitly refocus the current codebase on the Linux audiences (sysadmins, programmers, etc.) - and then think about what new codebases to create, to provide something exciting for other audiences.

  11. troll says:

    Good post. Nice that you mentioned Evolution. It is not the only application on default Gnome desktop that should have been removed completely as the time and development simply went past it. It actually reveals the main point: It's not the Gnome/GTK+ that is decaying. It is the developers, who are unwilling to move on.

    At this moment, Gnome should be integrating something like Mozilla's Prism, removing legacy stuff, going for Clutter perhaps, changing the main language used for development (99% of the new programmers on the field nowadays really do not do C) to something like Vala (it's still far away and it's not default or well documented) or C# or Java, bringing Telepathy into the core, removing the boxes (see , good example is Pidgin which is entirely fucked up solution to anything), the first EVER multimedia player for Linux (Banshee 1.0, the first to handle audio AND video properly, ever, on Linux!), ...

    Instead of that some idiots keep building TABs on Nautilus (file manager? on this century? wtf?), optimizing terminal applications (wtf again), integrating 5 different VM's to default desktops (just see those horrible startups! the reason is that if gnome/gtk did things well, they would not HAVE TO do that!), integrating VNC servers (even though we had already pretty nice ssh+X11 from the 90s or so already), totem (what a piece of *, that's not a player, that's a player COMPONENT, and even that barely), virtualization clients (even though virtualization should be really transparent), ... The list of plain brainfarts on the direction of Gnome/GTK development is nearly endless.

  12. Jeff Bailey says:

    The echos something I've been wrestling with a bit - I used to hack because there were things I wanted that I couldn't have.

    Now, aside from consistently working suspend and resume, and the ability to hook up to projectors when I give talks (The reason I have one laptop running Windows right now. All the rest run Linux and this one is mostly in a Linux VM) It's hard to find the burning need to hack on it to make things better.

    BTW, your "what's your favorite number?" thing has too narrow a definition of "number"

  13. troll says:

    I have to comment here for Julien, as his blog does not allow comments. Yes, decadence. You see, it's not the codecs or other improvements. The whole idea of having separate "box"/application is decadent. Especially as it requires tendering. Ekiga is one of those applications that should just die, no matter how many fancy new codecs you add. Telepathy and being transparent is the way.

  14. Yannick Defais says:

    Hello Andy and commentators,

    Raising concerns about the Gnome future is something useful, even if I'm not sure shocking the reader is the good way to get a decent discussion. But it's more or less just a formal remark, and I do not pay much attention to this.

    Let's get deep in the concerns. First, I fully agree having new and fresh coders blood in the game is vital. Some projects are now mature as you noticed, having them evolved or bring new solutions to old tasks is right and legitimate. Surely the interface between the user and the systemmay improve and your probably right mentioning Clutter; we will probably see major stuff coming from the easy use of 3D on the desktop (other that just fancy eye candy like Compiz is mostly for now). But this not means the unix way should break: we have many components doing one task and doing it right, one the top of another and I strongly believe in this pratice. Now we are in the path to center more and more our desktop on "people" (nice name for an application too :) and less on task (like evolution is mainly about reading emails). This evolution to focus on people rather than task is something lot of us feel. For instance, in Ekiga the Call menu as turned in chat, the GUI focus now on contacts, telepathy framework is backing a structure to help focusing on people throughout the whole desktop etc.

    I see this as a right move, still, as you noticed, many services appears on the "web 2.0". The major specificty of those services is this: they all break what internet is to go back in a client-central server model. Please, take the time to think on it. Client-server is *not* internet. When Microsoft started its own network "MSN" in the way back, it miserably failed. What is hapenning now? What is google doing? You tube? MSN/WLM (the im-chat client)? Centralisation! The intelligence, which is peripheral by design in the internet, is going back to some big servers. Indeed, that's not the Gnome philosophy. Gnome build intelligence on the peripheral, Gnome is internet friendly. And IMHO it should stay like that. Not only because of technical values, mostly because I identify free software with internet (they grown up almost at the same time and they benefit strongly each other).

    Backing a client-server model, is just a way to kill free software. At least, it is killing the "free" part of it. Centralisation means giving up control on our contents. As more those contents focus on our own social network, it means loosing more and more control on *our lifes*! That is "decadence". e.g. see what Facebook intented selling private information to companies...

    Healthy ecosystem for free software means a real network without center. We must fight to take the web back to the people. That means support the intelligence in peripheral: intelligence in Gnome.

    Frankly, what's the point having a new and more natural metaphor for the desktop if we give up on freedom in favour of centralisation ?

    That being said, let's talk about Ekiga, VoIP and neutrality. An essential "feature" of internet is its neutrality regarding contents. We are not tied to some formats issue, obscure protocol promoted over another etc. Aren't we? Is the network really neutral? Can you share files as efficiently as youtube using p2p? Does VoIP today require proxy forwarding to "just work"? Are all the routers really able to fit p2p?

    What is peer to peer? Is this different from internet? Or is peer to peer identical to internet? I strongly believe it is.

    Having peer to peer in the VoIP field to "just work" transparently like, for instance, Skype does really means cheating for the way internet is today. As a side effect, choosing this path for Skype, made this software to be banned from some subnetwork (like universities) because it is a big hole in security. Telepathy, Ekiga, Wengophone, Twinkle, etc. do all have the very same issue with this: network neutrality. Why does the network *always* works from a server-client model and is hard to comply with the internet model where each point is both a client and a server? And why are most of the technologies involved in the server-client model proprietary? e.g. flash to see video, MSN/yahoo messenger etc. are backed by servers etc.

    Please, make no mistake: there is no war between Ekiga and telepathy, there is a war to free the internet or to tied it to some central points. At least, for the Ekiga team which I think I can talk on their behalf on this: we do not have interest to promote Ekiga over another VoIP client. We promote interoperability, choice for our users in an ecosystem we believe will be more and more neutral to what internet really is about: sharing free contents safely. Now, we are the contents, we will not allow anyone messing with us.

    Best regards,

  15. Andy Wingo says:

    Sorry for not getting back to this until now. I think that Ekiga was a poorly-chosen example, and besides that not central to my argument. I've edited the article.

  16. Andy Wingo says:

    Also, thanks for taking time to comment, Yannick. Your points are all valid. And besides Unix, the last thing that I want is a world without emacs ;-)

  17. Felipe Contreras says:

    I also see GNOME as decaying, and I think there's more people with that impression.

    I think integrating WebKit would provide huge advantages to UI development. Some applications might use the rendering capabilities to create fancy themeable UI's (html, css). Others might use the network protocols (http, proxy, soap) to communicate with web services.

    Web 2.0 doesn't mean centrilized information. For exapmple, there's the dataportaility movement that looks very promising.

    Is there some kind of aversion to web stuff in GNOME?

  18. rodney says:

    Thanks for the good article about the decelerating rate of GNOME development. I just wanted to point this out:

    Decadence implies losing wealth. GNOME 2.22 is released and will never be lost. Thus GNOME's wealth can only increase. Windows XP is in the age of decadence. Soon you will only be able to buy second-hand copies until it stops working on your hardware.

  19. Henri Bergius says:


    Interesting post. I had quite similar feelings after last GUADEC where I saw all promising new ideas and technologies being mostly ignored.

    Hopefully GNOME can get its energy back!

  20. Philippe Kalaf says:

    Dear Yannick,

    "Why does the network *always* works from a server-client model and is hard to comply with the internet model where each point is both a client and a server? And why are most of the technologies involved in the server-client model proprietary?"

    It's not some conspiracy about network neutrality or freedom of networks or anything like that. It's just because we have NATs, and NATs are hard to traverse, therefore making point-to-point connections more difficult than client-server. This problem is mostly solved with ICE, this is also why we are working hard on libnice. SIP is mostly client-server anyways because of this problem, but it's not proprietary. XMPP uses servers to route signaling, but it's not proprietary... I don't quite understand how you tie net neutrality and client-server architectures, and in that logic how you associate Ekiga with net neutrality when it's mostly a SIP client (server-client arch)?

  21. fool says:


  22. Dustbin says:

    NOOOooooo. . . Don't use KDE4 as an example of what to do in order to move Gnome forward. The web is awash with complaint after complaint of KDE4. Sometimes, "good enough" and "just works" are high platitudes. On the other hand, you could ask for Linus' opinion. . . oh, wait. . . I think we know that!!! --dB

  23. WalterCool says:

    Evolution... dont have evolution?

  24. Anmar Oueja says:

    I think this posting sheds the light on a GNOME angle that may have been forgotten or simply ignored: Application and integration.

    I think GNOME has the necessary foundation to offer a somewhat useable environment. Unfortunately, as users, we don't use GNOME, we use the applications. GNOME becomes a glue that ties in all the applications we all need. Mail, Web, File Manager, chat/VoIP, and an office suite is what most of us, users not hackers, use.

    Evolution is a perfect example of an application that is simply too old and too bloated. I hate it and find it quite difficult and slow to use. Instead, you can have a Calendar, Mail, address book, chat, VoIP, file manager and web browser applications that integrated very tightly with GNOME and each other. If you have used the latest incarnation of Mac OS X, you will understand what I mean. Add to that telepathy, WebKit and the likes and you have your self a solid user experience that is both relevant and modern enough for the masses.

    Drag and drop, share contacts in the address book across the other applications or any new GNOME application, etc. This what will make GNOME experience move to the next level as an environment for the masses. Clutter is great and should be explored, GTK+ should be further improved, but that alone is not enough. We need the apps and most importantly we need the integration.

    If I had my way, I would push the applications/integration goal forward and when I am 50% of the way, I would start considering enhancements like Clutter.

  25. Menno says:

    I don't use Evolution, I don't use Epiphany, I don't use any of the Gnome Office (as it is called in Debian) programmes.

    What does that mean for me?
    It means that Gnome should do wat it does best, that is to make sure that what I want to run, not only runs fluently, but also communicates with whatever other program I want for a particular function.

    Elegance is in the eye of the beholder!
    A positive note, Gnome does a far better job than KDE3 in integrating 3rd party stuff like Firefox/Thunderbird (once again: in Debian).

    Further more, the introduction of Telepathy is a step in the right direction. I am not that technical, but making sure that a base process handles the integration of 'the same kind of' function (Telepathy:communication) with a seperate frontend is a good thing.

    The term Service Oriented Architecture comes to mind and is IMHO something that can be borrowed from to build on.

    The main problem is, are developers willing to rewrite instead of developing new things.

  26. perpetualrabbit says:

    Some random thoughts:

    * the gnome taskbar is very old hat. Clearly there is need for something better, hence efforts like Avant Window Navigator, gimme, Cairo-dock etc. Not that those imitations of the OS X dock are much of an improvement, or that I think the OS X dock is so great. In fact, I find Leopard lame. Similarly, the desktop icons + nautilus combination is old too (also I don't like it that icons are drawn in the root window, instead of loose things like in KDE. But KDE I find ugly). I think that a new gnome desktop should not need a bar at all.

    I think the gnome desktop should go 3D, integrated into compiz.
    Replace the taskbar with a 3D object that is out of sight normally, but pops up when right-clicking the background or using a hotkey (maybe ESC-ESC). Icons should just be windows, but zoomed out and clustered in space according to a common task.
    Also we need no desktop anymore, we just have Space. The challenge is how to easily navigate that space, and how to meaningfully place objects like directories, files, running applications and other things like an avatar of an AI in it.
    Mentioning AI, I don't mean something like "clippy". I mean an expert system that is genuinely useful for human computer interaction. Two examples: mention "color correction" to it, and it should be aware of if and how much your display/scanner/workflow is ICC color corrected. It should be able to report the current situation (this and that is missing) and be helpful in filling in the missing parts (you should use LProf for this part, here's how. Your scanner is not 100% Sane compatible, these are the limitations). An expert system is tought by experts, and the AI system should be able to learn. The computer should regularly upload things the AI has learned, and download things that others have learned.
    Second example: your desktop has a RAID1 device, and one disk breaks down. Your AI alerts you, and steps you through the process of replacing the disk, repartition the new disk and integrate it again in your degraded array, making it whole again.
    In short: no _real_ AI (which as far as I know does not exist) but an expert system around all kinds of common or uncommon tasks for people of all walks of the (computer) life. The knowledge of artists, sysadmins, programmers, latex-ers, accountants etc. combined. Tought by the same people who use it.

    I have only the vaguest idea of how it should work, but would be genuinely useful, and something really new. (at least compared to current Linux, OS X and Vista, which are, face it, basically the same).

  27. Thomas Thurman says:

    I don't want to live in a world of arbitrary change for the sake of change any more than I want to live in a world of arbitrary change for the sake of Microsoft's shareholders. Change for the better is good. Change away from the worse is good. But change in itself is a nasty, disturbing, uncomfortable thing, makes you late for dinner.

    I want to live in a world where each program asymptotically approaches perfection and fits its ecological niche. I want to live in a world where programs have irrational version numbers and get a digit closer to them each release, where bug lists truly do shrink to nothing.

    I want to live in a world of decadence.

  28. Zach says:

    Yannick, you should take a look at the semantic web, rdf, microformats, and how they are decentralizing data. Centralized servers may still play a role but the data is becoming open. Web 2.0 is on the way out already. (There is also a wikipedia article about the semantic desktop which applies similar ideas to the desktop environment)

  29. Max says:

    Though I am not currently deeply involved in Gnome development, I'd like to share my thoughts on this issue. From my point of view (I work as a Linux consultant and work with Linux desktops a lot), the focus in this debate is wrong. I understand your point, but I do not think the road to mass-adoption of the Gnome desktop ("there") lies in new toolkits, OpenGL accelerated widgets, document-centric desktops or client-server architecture. Of course, having any or all of these help, to some extent, to build a good desktop environment. But, assuming we really want to go "there" as our ultimate goal, we need to work with the existing infrastructure in the "here"-and-now.

    In one of the previous posts, someone said 'I don't see the need for an email client anymore'. I think this illustrates my point fairly well: though he is right from the perspective of the individual user, he is utterly wrong from the perspective of the larger organization. This is where the original post makes a very good point: the seemingly steerless development model of Evolution over the past few years is a good example of how we shouldn't work. Evolution is a good mailclient, if you use POP or IMAP. Most large organizations I have professional experience with use Exchange however and Linux migration projects often occur in such environments. Evolution and Exchange do not work very well together. Improvement here would help Gnome further from "here" to "there". I mention Evolution because it is a point easily made, but there are more than enough other examples.

    If Gnome is bound to succeed on a larger scale, Gnome should think big, think big in interoperability, think big in multi-user environment options, think big in user profiling and so on. Toolkits and such can help to get there, but they are no more than what they're names imply: tools, not a goal by itself.

    Just my thoughts here, I am not a developer, so maybe I miss some very important point. If so, sorry, just trying to add to the discussion :-)

    PS. Your blog doesn't seem to like ampersands in posts. Something with bad XHTML.

  30. Luca Cappelletti says:

    Hi Andy,

    may you provide,please, a better list that sustain your theory?
    Seems just a little bit flaming.

  31. Janne says:

    "NOOOooooo. . . Don't use KDE4 as an example of what to do in order to move Gnome forward. The web is awash with complaint after complaint of KDE4. Sometimes, "good enough" and "just works" are high platitudes."

    We should separate the complaints about KDE4 and KDE4.0 ;).The complaints have been that KDE4.0 doesn't have all the functionality that KDE3.5.9 has. But that lack of functionality is not the "vision" that was mentioned here. Those absent features are absent because KDE4 is a big effort, and the initial release was more about preparing the groundwork for upcoming versions of KDE4. They are not absent because KDE-devels decided that "the future KDE will have less functionality than 3.5 has". It was due to the fact that implementing all those features would have been too much, considering the manpower, resources and time. But they will make their comeback.

    Now, as to the subject at hand... I must admit that I have seen this "lack of innovation" in GNOME. I see the version-number progress, but I don't really see that much changes. No, I'm not talking about changes for the sake of having changes. Sure, there are small changes here and there, but every time I'm left wondering "are these changes big enough to warrant a whole new version-number?". And usually my answer is "no".

    Now, what is this "vision of KDE4" that was referred? Well, there might not be a clear roadmap that spells out where KDE4 is heading, but you can see bits and pieces of it from various KDE-related blogs. Things I have gathered so far:

    a) Semantic desktop. KDE is actively moving towards this. Well, what is it? Well, basically the system would know how different things are connected. You could ask the system "show me all the attachments sent by this person". Or "show me all IM-discussion I have had about this particular model of car". And so forth. It's like Spotlight/Beagle taken to the next level.

    b) Zooming user interface. This is an active project in KDE, and it has the potential to bring big benefits in the future.

    c) Krunner. Now, Krunner might seem like a launcher. But it seems that they have big ambitions for it. It could become a central tool of working with your system. It could be seen as a tool to have a dialogue with your computer. Think of it like Quicksilver (from OS X) taken to the next level. The potential is there.

    d) Automation. Take the Folderview-plasmoid for example. So it's an applet that displays contents of a folder on the desktop, big deal. But the thing is that you can use filters, so it will only show certain type of filters. And, in the future, the displayed contents could be automated so that it only displays content that is relevant to the task you are doing. Suppose you launch a video-editor. The applet could automatically change to display your video-files. When you launch an image-editor, it could automatically change to display all your images. And so forth.

    I don't know about you, but having those four features alone would make for a pretty compelling desktop. Not to mention all the "normal" improvements that time would bring.

    Note: please don't turn this in to a KDE/GNOME-flamewar :).

    Point I'm trying to make here is that they seem to have long-term goals and projects that work towards those goals. Does GNOME have those? To be honest, I don't know. But what it seems to me is that GNOME just moves from one version to the next, with incremental updates here and there. But there doesn't seem to be any kind of grand vision.

  32. Koen Wybo says:

    Just a litte notice as a 'normal' gnome-user. As remarked: it's a question of integration (yes it can be more efficient) not in the first place of more shiny 3D-blingbling. Therefore you need more standards on desktopbehavior and documentation: example: popping up the 'add-contact' should be achieved by.... Why not have an gnome-standard that deals with contacts and which can be read by other programs.
    A second thought: adding 3d can be fine if it improves overall productivity (no compiz for me: it only gives just a few improvements for my work, but in general those are not compelling enough to activate it.)
    The problem with 3D-interaction is not the 3D but the tools we use to interact with the 3d: mouse and keyboard. These input-devices are just not sufficient enough. What is a clean way: look at apple's iphone. Apple made it as if you can move windows, buttons, scrolls, ... with your fingers (not keyboard or mouse). It's a 'natural' way to interact with 3D.
    Another way of putting it: the decadence of gnome does not lay within GNOME itself but with the tools which people use to interact with the GNOME-interface. If you stay keyboard-mouse centered then GNOME is on the right way ahead, adding functionality every release. If you want to improve GNOME: seek for other input-devices (and keep in mind that most people are still going to use K-M).

    Third thought: don't forget the businessmarket. Working with windows is just fine for these K-M-users.

    Personal opinion: if you want to try something new: remove the titlebar. Not functional at all. Buttons to maximize, etc. should pop-up if you hover over the right upper-corner of an application. Or: extend the functionality of the titlebar: add the icons of the programs which are beneath the opened window: people who never use alt-tab will be gratefull for it (and so you can exclude the opened-programs-panel on the bottom). Or: put the two suggestions together: make a thick line where the titlebar is now: it opens when you hover over it and brings you extra functionality.

  33. Rafael Jaques says:

    I do preffer Gnome instead of KDE.
    But now is the time for Gnome to improve features that are missing, like some Treeviews or a "Finder" clone (like the new Dolphin of KDE4) in Nautilus...

    There are so much work to do... We have to stay on! :)

  34. Yannick Defais says:


    Philippe Kalaf said:
    "It's not some conspiracy about network neutrality or freedom of networks or anything like that. It's just because we have NATs, and NATs are hard to traverse, therefore making point-to-point connections more difficult than client-server. This problem is mostly solved with ICE, this is also why we are working hard on libnice."

    I do agree. It is in no way "conspiracy", it is more a misery. I mean, in one hand, obviously NAT designers did not agree to build this situation. That's why there is no conspiracy. But, in the other hand, they obviously did not try to get a consensus to standardise their devices. Which is, in my mind, just misery.

    Especially, in the free software movement (FOSS, etc. no matter how you name it), we all try to standardise formats, we push for interoperability. In the field I know best, VoIP, most clients use open standards (XMPP or SIP), but, as some companies achieved to tied up most of the population -their "consumers"- to their own private, obscure, "moving target" protocols, some tools do exist to be "compatible" with those protocols (e.g. aMSN, emesene, gyachi-sidetrack, SippySkype, or as a service GTalk-To-VoIP)

    My point is: free software do care about "point-to-point connections", not for some technical reasons (like we do not have enough resources to have big servers like Microsoft or Google have), but because the goal is to spread the intelligence fruits, to spread our free softwares, not to build a center where you could find those fruits. This is a consequence of our view upon freedom. Sure, we all have our jokes about "world domination", but truth is free software do care more about interoperability using the network, than having, e.g. firefox dominate the market. Our goal is to protect users from dependency over an actor. On a network point of view, this means point-to-point connection, or intelligence in peripheral.

    The ICE technology is a workaround to a bad situation where NAT behaviors are not standardised. ICE try its best to have a working point-to-point connection, if it fails to do so, it will relay on a proxy to forward the datas. The default is point-to-point connection, i.e. to comply internet neutrality; there is no reason to grant privileges to big servers on the network as it is now. This neutrality is so important the IETF works to solve this issue with NATs through the BEHAVE working group: . If this working group achieve its goal, ICE will surely be deprecated (at least the proxy part of it).

    ICE and any of its implementation, will not avoid the need of big server to relay the streams until the NAT issue will be fixed (I hope soon based on the BEHAVE wg). This mean, libnice will not be enough in itself to get a working point-to-point connection for VoIP (or any other stream): actualy you still need big servers to proxy the stream in some cases, no matter how much intelligence you'll put in the libnice library. This is clearly against internet and libnice is *not* the solution, it is just a workaround. Standard, interoperability, "idiot proof" NATs is the solution.

    "SIP is mostly client-server anyways because of this problem, but it's not proprietary. XMPP uses servers to route signaling, but it's not proprietary... I don't quite understand how you tie net neutrality and client-server architectures, and in that logic how you associate Ekiga with net neutrality when it's mostly a SIP client (server-client arch)?"

    SIP is not server-client arch. e.g. I do have a static IP and I run my computer 24h/24h. You can join me using a SIP address at any moment using my IP: sip:my_IP It will work, no matter which SIP client you are using. The "server" part of SIP is just like BIND is to the internet DNS service: server binds the human readable address (for instance my sip address is ) to the IP I use to connect internet. Servers can also be used as proxy to relay streams, but it is optional and less efficient (e.g. this induce more delay in VoIP). Simply put, the signalling part may relay on servers to ease the connection (using human readable address, chose the cheapest route -see ENUM-, etc.) but default established communications are point-to-point (the RTP part).

    My point about network neutrality is: as the current internet is broken because of some NAT devices, any VoIP system that "just work" will need either big servers to relay tstrea

  35. Yannick Defais says:

    Argh! I pushed the Enter key...

    I quickly finish here:
    My point about network neutrality is: as the current internet is broken because of some NAT devices, any VoIP system that "just work" will need either big servers to relay streams (like MSN) or use bad tricks to fool those NATs (like Skype). Being tied to big servers breaks neutrality, giving more power to big companies, or breaks network security (like skype does).


  36. Robbins says:

    I is excesive to have more that one main desktop system. Look at windows, it is great and is only one. GNU/Linux should achieve that to be a competitive desktop environment.

  37. Cole says:

    I'd work with incrementally fixing things that are known to be bad rather than one big change that breaks everything. One example is theming and custom widgets etc. How long has this change been touted around for?

  38. Philippe Kalaf says:


    Ok the way you run SIP isn't standard :) I guess you need to setup your firewall to open that port, and most people don't do this sort of setup, specially since they have dynamic IPs inside their own networks and everytime that changes the port forward breaks :)

    About ICE, with libjingle (which uses an old draft 6 version of the spec), they have a success rate of 96% for traversing NATs without a relay :) A relay takes it up to 100%, so even though in the really bad cases it needs a relay, 96% of the time it's fine :) GTalk actually works nearly as well as Skype, without the evil proprietary protocols. We really believe it's possible to have a working P2P non-proxied Voip solution with 100% open standards, and we are working on "mingle" that will provide non-"server mixed" multi-user audio and video conferencing :) It's going to be fun :)

  39. Mitch 74 says:

    There are valid points to the articles, as well as to several posts. However, some are not too good:
    - Mono/Moonlight: do you really want to make use of a free implementation of a proprietary technology? Look where Sun got with its Java desktop: nowhere. And the GNOME developers derided KDE for using "non-free" Qt! Pot, kettle...
    - using other languages than C: for interfaces, yes maybe; however building a whole app upon a scripting language isn't exactly performance-friendly. GNOME is already slow on older machines
    - GNOME isn't stable. Period. Even using only core elements and a feature-deprived minimal desktop install, I get strange crashes or even unrecoverable sessions.
    - GNOME isn't fast. The compositing WM is available only if you compile Metacity with a special parameter, and I haven't seen anything to change said setting at run-time. So, not only is it big in RAM (come on guys! KDE4 is lighter than GNOME!), it's a CPU hog.

    So, just to add insult to injury, you should know that Xfce gets ever so tempting:
    - nice file manager: sorry, but GNOME drops me to the command line as soon as I need to perform somewhat advanced operations
    - its compositing works, and is a breeze to enable; nice boost even on a mach64 card
    - still Gtk+ based

    My guess is, GNOME is lucky that it's default on the 3 biggest distros; if it weren't, it would have disappeared already.

  40. Bernhard says:

    Hi guys,
    don't start discussing warp-speed before we have mastered slow walking pace.

    Example 1 - stopping an app:
    Iceweasel alt-d-b
    Liferea ctrl-q
    gnome-terminal shift-ctrl-q
    Evince ctrl-w
    ad infinitum ad nauseam...

    In a truly consistent desktop environment the user would only have to memorize one shortcut.

    Example 2 - convoluted senseless dependencies
    simply try "apt-get -s purge evolution" and savour the trail of destruction this would leave in your install of Gnome under Debian.

  41. Mark Nowiasz says:

    I don't want to start another KDE vs Gnome war, but let me tell you why I've switched to KDE after using Gnome for about 10 years - I think it might tell something about the state of Gnome.

    You've mentioned evolution, and precisely because of evolution I changed my desktop environment: I was using Gnome pilot/evolution to sync my Palms/iQue. For reasons unknown to me, the syncing started to grow worse and worse - gpilotd crashing for unknown reasons, much more irritating: syncing produced strange effects, like duplicate entries (addresses/calendar), changing an entry in evolution didn't result in having it changed on the palm, etc.

    (And don't let me start about gaim/pidgeon using evolution's addressbook - it was a nightmare trying to add new contacs).

    So, since I already had a lot of KDE stuff installed (due to programs like scribus oder kdissert - there are no similar gnome/gtk programs available) I installed the rest of KDE - and was suprised. Not only did KPilot work flawlessly, all(!) programs were running much faster and efficient - plus had usually much more functionality build in (not to mention: more stable).

    Kopete worked flawlessly with kaddressbook, which synced perfectly which my iQue, there's no monolithic monster like evolution - instead there's a kind of wrapper called kontact where seperate programs run (kmail, kaddressbook, etc.).

    So I think there are three problems with Gnome right now:

    1.) Performance. Gnome programs tend to start slower and consume much more resources than their KDE cousins.
    2.) Integration. The integration of the addressbook (usually evolution's) in other programs is poor or non-existant.
    3.) Community. When comparing the communities, it seems to be that the KDE community is much more vibrant than the Gnome one - I guess that "decaying" is a good word for it.

    Well, I think that - if Gnome really wants to survive - it will be a herculean effort to even fix the problems mentioned above (performance, integration, etc) - after *that* you can talk about new ideas/concepts.

    (Oh, and please remove mono - it's a huge risk using it with little gain).

  42. Mariano says:

    @bernhard: gnome-terminal adds Shift to all the standard keyboard shortcuts because it cannot trap the standard ones: they are needed for the apps that run *in* the terminal. You need to be able to send Control-C, for example...

  43. Petar says:

    I believe that GNOME is in a state of decadence for several reasons.

    The first is somewhat political. Many Gnome users used to bash KDE for it's strong Windows resemblance only to discover that from release to release their favorite desktop has somehow fully embraced many of the many times strongly criticized Windows technologies like .NET with mono, Windows Registry with gconf (which I have impression that never got fully implemented, and now we have proposals of inclusion of Moonlight ("free??" implementation of Microsoft's Silverlight).

    The second reason is it's total lack of media coverage. I'm dying to read what's going on in between two releases of Gnome, but's main page is updated once(or twice) every 6 months, has rarely something new to report. Is Gnome so bad and uninteresting to deserve this kind of treatment?

    The third reason is the problem of bad integration as many already mentioned here. As far as I understand it, this can be solved by defining a list of frameworks that should become part of standard Gnome desktop like a multimedia framework, PIM/calendar framework, IM framework or whatever other framework type you can think of. This way it will be this frameworks back ends job to provide common services to many different programs like for instance centrally store all our contacts, schedules, specific types/categories of data. Not that I don't see this kind of tendencies already (like with Telepathy) but they should be stronger.

    As for the radically new desktop idea, I believe it is not that easy to accomplish that so soon. Serious research, investigation and testing is needed, in order to give definite answer to what the future desktop should really look and function like. For instance I have seen many proposals about circular menus on other sites. It is obvious that people proposing or trying to implement such ideas are totally ignorant of human vision/perception characteristics. As for the document centered desktop, I believe we already have our first experimental platform, the Sugar desktop built into the OLPC laptops. And since it uses current GTK+ plus some external libraries, it proves that there is no need for breaking API/ABI compatibility. But API/ABI compatibility will probably have to be broken if we want 3D animated menus etc. because it cannot be done easily with the current version of GTK+ (but it can be done with QT).

  44. Allen Dulles says:

    I quit using Gnome years ago after Nautilus deleted some of my research files (a critical bug that was never tested). KDE is easily the best Linux desktop for stability and usability.

    One of the reasons Windows is the standard OS for 95% of users is because of its usability. In contrast, the Gnome designers are total retards. When they reversed the standard OK and CANCEL buttons (from Windows), I said enough is enough.

    Like most Linux projects, Gnome is a complete mess with zero project management. Most Linux developers are clueless idiots, with no formal education in programming. Typical Linux software is untested, broken, forked, or abandoned garbage.

    Gnome is the definition of retarded, and it's getting worse not better. So, put me on your "community" banned list and delete this posting. I know Linux people don't like feedback.


  45. Iuga Marian Silviu says:

    Gnome has advantages:
    -it is more stable than kde
    -it stops sounds faster than kde
    -it works on more simplistic platforms (mobile) than kde
    -it can be fixed by rewriting its source
    Gnome has disadvantages:
    -it has to be carefully rewritten and recompiled
    to work well (Debian)
    -all other applications except Totem do not support surround sound (kde's Amarok supports surround - say you make Amarog)
    -if an application goes full screen and crashes, the only option is to reset computer the hardware way (also in kde)

  46. memo says:

    Yeah, the state of decadence ? Why ? Because you can not see candies on the GNOME ? What improvements can you see to the other desktops ? What else than bells and candies ?

    GNOME is ok, it gives maximum of usability for generic users. The old Linus reverted to GNOME after KDE experience - was KDE remembering him of another well known operating system ?

  47. Juan Carlos says:

    @Allen Dulles: Normaly i dont answer this posts, but, you look so ... I dont know the word. Looks like Gnome kills your sister or Linux ruined your live. If not I dont understand this hate. Anyway one thing you mentioned its true Freesoftware people doesnt like criticism but its necesary.

    In my humble opinion Gnome its a great piece of software and its evolving slowly but in the right direction. I like very much the WebKit integration and the introspection work.

    If you ask When Microsoft Windows grow so much to be the number 1. I bet that the day MS developed Visual Basic. Probably the bigger lack of linux is a good integrated RAD.

  48. Vamp898 says:

    I think the point is that GNOME really _needs_ Vala

    They fall in love with C# but they really shouldn´t start writing every new app in C# beceause they dont like there own great love C

    Using GObject can be compared to water boarding. Vala is the _only_ usefull way for GObject so they just should pump a lot of work into Vala insteat of "a bit of every".

    A bit C a bit C++ a bit Python a bit C# a bit Vala

    with the time it seems that GTK+s main Code are more bindings to other languages than really code beceause C was the baddest choose for the situation in which GNOME is.

    If they fall in Love with C# than just make Vala usable and use it!

    and btw. Linus reverted to GNOME after KDE 4.0 was invented in his Fedora beceause KDE 4.0 was no stable version.

    Until today i did not understand why every stupid distro out there put KDE 4.0 (what was masked as a 100% unstable version) into there repo and told everyone its the greatest thing in the world.

    @ Allen

    you´re opinion is like were in 1980. Go and live in the future xD

    Everybody knows if Windows would come out right now it would have the same future than ReactOS and everybody knows ReactOS have no future.

Comments are closed.