Submitted by Only a few days ago, we ran an article on the future of KDE and GNOME, and which of the two had the brighter future based on their developmental processes. Barely has that discussion ended, or the GNOME engineering team comes with a pretty daunting plan to introduce a fairly massive reworking of the GNOME interface for GNOME 3.0 (2.30). Read on for the details.
The proposed plan encompasses quite a few aspects of the GNOME platform, but the most user-visible one is of course the reworking of the interface, thanks to a project called GNOME Shell, developed by Red Hat employee Owen Taylor. It’s more than just a replacement for the window manager and panel, though.
The two big parts of GNOME Shell are the panel and “overlay”. The panel is pretty much self-explanatory. It sits at the top of the screen, and houses the activities button (which activates the overlay) and the system tray, user name, and clock stuff. The overlay is a bit difficult to put into words, so I think pointing you to the screenshot is a much better idea. What you are seeing on the left is the sidebar, which displays most recently opened applications and documents, but of course you can also use the search field to locate your applications and files. On the right, you see your desktops, so you can organise your applications and desktops.
Obviously, GNOME Shell supplements the user interface with subtle effects and animations, all written in JavaScript and Clutter (it uses a special MetaCity branch for now, but will fork the MetaCity code later on). Owen Taylor also wrote an email detailing the future of GNOME Shell and MetaCity.
There’s more to this plan than just the interface. Another major pillar of GNOME 3.0 would be Zeitgeist, a new approach to managing and finding files by using tags, bookmarks, and timelines (an .ogv video shows off how it works).
There is a big advantage to working with GNOME Shell and Zeitgeist: both have already been in development for a while, and already have working code and have seen lots of development. This means that GNOME 3.0 will be less about “starting”, and more about “finishing”.
The GNOME foundation will also be cleaned up and streamlined. What this means is that deprecated libraries will be removed from the platform, so that people will stop coding against them. They will also “create a staging area in the platform for libraries that aim to be in our platform but do not offer enough guarantees at the moment (like GStreamer): this will send a clear message on what should be used”. Other aspects include more focus on new technologies (like Clutter) and more clarity on what external dependencies should be used. Applications will have to be adapted for these changes, of course.
The GNOME team is currently not sure yet whether or not to include the “old” MetaCity/panel in GNOME 3.0. It would make sense to do so to make the transition an easier one, but it might also hinder adoption of the new technologies and ideas.
I’m glad that the GNOME community finally decided on a good roadmap for the future. It’s also clear that they have chosen a very practical approach, by focussing on technologies that are already in heavy development, with code available, which gives the plan a major headstart. It’s also good to see they aren’t just copying the competition, but have clearly chosen a path of their own, one that seems to fit within the current GNOME ideology.
What about the spatial environment? I don’t know how many people use the feature (I suppose many use the default Windows/OSX-like browser interface)–I do, simply because it’s the closest you can get to an OS9-like behaviour (the right theme also helps
). Will that still be available? Or will the changes to Metacity be so dramatic that it’ll no longer be an option?
You can rest assured that no matter what it’ll be an absolute pain in the AS$ to change it. Currently there is no way to change it through nautilus itself…you need to start up gconf.
Bullshit. Edit, Preferences, Behavior, “Always open in browser windows” checkbox. Checked, it’s an explorer-like interface, uncheck and it’s spacial
But why be bothered with the facts, given your obvious KDE leanings?
I’ve always found the text for that option to be pretty vague and easily missed. I can tell you for a fact that you haven’t always been able to configure Nautilus like that (spatial was intended to be the complete de facto default), and that’s only after a few arms have been twisted up backs as to how brain damaged spatial browsing is for anyone doing a lot of file management.
To make spatial browsing actually work then you need a way of managing and organising all the open windows representing each folder that result (Expose for file management or something like that), and that’s why, errr……things like treeviews were invented. Duh.
I think I’ve actually made hundreds of dollars in consulting fees for telling people that. On systems I actually administer, I make it the default from the git go, of course.
you can too change spatial/browser mode in nautilus
I don’t like the look of it. Smells to me like the jump KDE made from 3.x to 4.0. It was a total disaster. Might happen again. Where’s that incremental spirit of GNOME?
And KDE was not in “incremental spirit” on 3.x series??? Sometimes you just need to make a move.
KDE is not a total disaster. I use KDE 4.2 and love it. The first two point releases were bad, but it’s gaining shape now. Sometimes you need to make a clear cut in your path, like Apple did, like Microsoft did with DOS->NT migration, like KDE guys did.
On the other hand, I see the word enterprise and document used too much in gnome, too much of business involved in the decisions. How boring.
Think about when jumping GTK1 to GTK2.I think that was a mess too.
Form development point of view, stick to old library means we have to deal with potential legacy/ugly code when implement new ideas. And difficult for newbie to contrubute.
Think about the delay of Xorg 7.5. X works for decades. It take afford to implement new feature based on old code. That’s why Wayland comes out.(Although it is not a replacement for X, but Wayland make it a lot easier to implement modern stuff).
KDE4 needs a bebase/rewrite to incorporate new things, so there it went.
No one knows when GNOME needs too.
GNOME and KDE are just two different ways of how FOSS evolve.
Like a Jump ???
You did NOT even read it and you dare commenting …
Apparently you don’t remember the GNOME 1.x to 2.x transition which was arguably more jarring than the recent KDE fracas.
Hey. I was using KDE long before there was a KDE 1.0. I was using Gnome long before there was a Gnome 1.0. And I went through the Gnome 1.4->2.0 transition. And you are spewing crap. There is no way in hell that the Gnome 1.4->2.0 transition was ever even in the same ballpark as the disaster that has been KDE3->KDE4.
I get the impression that some people think that if they keep claiming so, it will somehow become true. But it won’t. I know. I was there. And I will continue to remind folks who are engaging in revisionist history that they are engaging in revisionist history.
Edited 2009-04-04 18:04 UTC
Agreed. About the worst thing I found in the GNOME transition was that, for quite a while afterward, there were still a lot of apps written in GTK+ 1.2, and so didn’t integrate with some of GNOME2’s components.
This is what has made Gnome’s overwhelming success over the last few years, and have the top 3 distros make Gnome default desktop environment.
Gnome, please stick to your philosophy.
… and the desktop zooms out so the entire desktop fits within the overlay. When you select something from the activities menu, or press it again, the overlay closes and the desktop zooms back out to full screen.
This doesn’t really come across well in either the screenshot or the text.
Personally I think it is a neat way of exploiting the new graphic capabilities to do something useful.
Since this is a prototype, however, it is very unlikely to look exactly like the screenshot when finished.
Let’s hope that GNOME 3 will not repeat the KDE 4 disaster. KDE 4 was announced with colored bells and whistles as the breakthrough desktop environment and even a Windows version should be released (there was a pre-alpha Windows port of KDE two years ago). Today, despite its 2 major releases KDE 4.2 is still a mess and unusable. I have used and loved KDE 3.x for five years but now I have switched to GNOME and will not come back because every major Linux distribution uses GNOME as default DE.
Spot on, what I love about Gnome is simplicity. I don`t get that with KDE, features or not.
It is not unusable, I’m using it right now. I can write mail in the incredible (and best mail client ever) KMail, I can browse the web with Konqueror, I can chat with Kopete, read PDFs with the magnificient Okular, download movies with the almighty KTorrent, manage files with the fine Dolphin (still I agree with most people that nepomuk doesn’t solve any problem, as there’s not any problem with files), write some LaTeX with Texmaker, play movies with Dragonplayer (needs more work), etc. It’s already usable for a lot of people. Maybe that’s a matter of what distro are you using, perhaps your distro is screwed. Mine works fine (arch here).
Still I want KOffice 2 to be in good shape with some Word and Powerpoint support. What else can one ask for?
There’s lot of people criticizing Plasma and the new desktop without icons (you can enable icons) and all that shit. People that use gnome and still say KDE offers too little customizability; f–k, you’re using gnome, you can’t even edit most toolbars! So vote me down for saying this, but if you don’t like KDE now, because IT IS INCOMPLETE and IT IS BEING WORKED ON, go use gnome and shut up. Don’t just say it’s totally unusable.
It is not about writing mail with KMail – it is about writing mail with KMail without any crash. Ok, I do not know how stable KMail is but the overall stability of the KDE desktop is only comparable to Win98 (unfortunately it is not just a bad joke). It is also about responsiveness: Why must I wait almost a second until KMenu comes up? On Vista the Start menu interacts with the user instantly and without any – the same with GNOME.
Again: I WAS a big fan of KDE for 5 years, but KDE 4 is in my view the biggest disaster in OSS history and only the HURD saga (will be released next year:) can keep up with this disaster.
GNOME could be better in many ways: Almost everything is static here and you can’t customize anything, and, and, and. But hey! I get my work done using GNOME. I cannot claim the same about KDE. Basically I use only 3 apps: Eclipse, Firefox and the Terminal. Eclipse and Firefox still look odd under KDE. Ok, GNOME’s Terminal is crap but at least it remembers its settings.
KMail hasn’t crashed to me yet, but I’m not a heavy mail user.
About responsivenes, maybe you need to switch your distribution? My PC specs suck compared to today standards, and KDE reacts quickly most of the time. I’ve presenced some hangs, mostly in Konqueror with flash plugin. But otherwise the system feels very good.
I will NEVER EVER(!) forget how KDE has screwed up my Powerpoint presentation because it has not managed to connect to the beamer; and no, X has nothing to do with it. Of course I have tried if the beamer works fine with my notebook setup before (two times!). But as it came to the actual presentation nothing worked. Fortunately I had a good old paper backup of my presentation… But the nicely prepared presentation was screwed up – by KDE!
Again, it wasn’t any nerdy tinker game. I was the stuff I do to pay my rent – screwed up by KDE. So long to the reliability of KDE. An now hate me because I don’t like you favorite toy!
Dude, you’re a bona fide troll. Fair enough, connecting to a projector (if that’s what you actually mean) is difficult at the best of times, certainly with Linux, but to present yourself as knowledgable enough to know that the fault is somehow with KDE is just trolling. Plain and simple. No matter, it’s what we’ve come to expect over the past few months.
I’ve never, NEVER EVER had Mac OS X or Windows fail when it comes to connecting to a projector, and I’ve used a lot of those things the past ~9-10 years.
It will always remain a mystery to me why something so mundane as plugging in a 2nd monitor is such a big thing for Linux. It’s not like we’re asking for world peace.
Plugging in a second monitor is common, but not exactly mundane from a technical standpoint. To some extent, I think we are still safe to blame David Dawes for the lost years of development we suffered. Granted, it’s been a while since he was ejected from the scene. But there has been a lot of work to do to recover from it.
Of course, there is also the matter of our bizarre focus upon eye candy, to the exclusion of the more practical matters at hand… like plugging in a second monitor. (Does compiz *really* help our desktops significantly?)
But I do think that things are coming together in a “just works” form. It’s just been a long, rocky road, full of detours.
Edited 2009-04-02 22:06 UTC
Good point. I really think Compiz is useful (I should write an article on that one of these days *makes note in his E71*), but it is indeed a mystery why something like Compiz works pretty well, but multi-mon support is a bitch.
Well, I must confess that I’ve never had an issue either with Windows or OS X and projectors. Maybe I’ve had some trouble getting the correct resolution, but I’ve always got an image coming out.
Linux, and X specifically, is a bloody nightmare to get a second monitor connected to. There’s nothing universal in X so ATI or nVidia go off and create their own ways of doing it. It’s certainly not the fault any of the desktops though, and it’s only now that we’re getting rid of that pile of junk called the xorg.conf file. It’s just a pity that there’s nowhere to easily put settings when you need them!
How about a config file? I would suggest xorg.conf as a name, oh wait…
Seriously, what’s bad about xorg.conf? We just need a GUI that touches the file when you need to change something in xorg. Better then, that xorg touches its own config file when it’s requested to save changes.
Edited 2009-04-02 22:28 UTC
xorg.conf is not bad per se, because you always want to store some settings permanently, but it is unspeakably bad given the use cases we have today of people plugging different monitors, projectors and even TVs in as well as second displays. What gets set in xorg.conf on install cannot be easily and dynamically changed on-the-fly based on what gets plugged in, and if you have a display that is incompatible with what’s in there then you that adds up to black screens.
I have! Both OS X and Windows. Of course, all cases were with the same projector.
NEC projectors and IBM laptops using Intel video chipsets don’t play well together. Neither do some early HP laptops using Centrino chipsets. We keep three different brands of projectors available at the board office for people doing presentations. These laptops use Windows XP and Vista.
Oddly enough, Kubuntu 7.10 and later on HP laptops don’t have any issues with the NEC projector.
Just like with Windows, it all depends on the driver. I’ve been using the multi-monitor support (MergedFB) in ATI’s fglrx driver on Debian 4.0 and 5.0 without issues (no-name X1300 card). I’ve recently switched to using xrandr with the OSS ati driver. It’s a bit slower, and the few times I’ve tried desktop effects it’s crashed plasma (but I don’t use desktop effects even on systems where it doesn’t crash), but it works.
Others at work using the ATI driver have no end of issues with Kubuntu. Same with those with nVidia and Ubuntu.
It all depends on driver quality.
I have on both Windows and OS X with different projectors… the projectors didn’t seem to send any information about the resolutions supported and when I ‘forced’ 800×600, which I knew it supported, it got an out-of-sync error. Admittedly, they were pretty crappy projectors…
Stop being so dramatic. I would hardly term KDE 4.2 “a mess” or “unusable.” In fact, KDE 4.2 is the first version of KDE 4.x that IS usable and ready for the masses. Obviously the Debian developers think so as KDE 4.2 is moving from Experimental to Testing this week. I’m sorry you seem to be having trouble with KDE 4.2. Have you considered that it might be operator error?
Some of us might respecfully disagree.
It’s completely skipping the “unstable” branch?
Here is how Wikipedia describes the Debian repos:
“””
Software packages in development are either uploaded to the project distribution named unstable (also known as sid), or to the experimental repository. Software packages uploaded to unstable are normally versions stable enough to be released by the original upstream developer, but with the added Debian-specific packaging and other modifications introduced by Debian developers. These additions may be new and untested. Software not ready yet for the unstable distribution is typically placed in the experimental repository.
“””
So up to this week, Debian has considered KDE4 not to be stable enough to go into “Stable”, “Testing”, or even “Unstable”. (The Debian “Experimental” branch is one that I only learned about due to the prolonged KDE4 fiasco.) But you think KDE4 has been ready for the masses since the release of 4.2? And I’m unclear why you think that inclusion in “Testing” is some vote of confidence. It is only a vote that they think they might consider it stable in a year and a half or so.
Edited 2009-04-02 21:06 UTC
As a long time Debian user, I do understand how the Debian repositories work thank you very much. I do apologize, however, as I did misspeak: KDE 4.2 is moving into Unstable this week. (http://lists.debian.org/debian-kde/2009/03/msg00157.html) Nevertheless, I stand by my opinion that KDE 4.2 is indeed ready for use by KDE 3.5.x die hards. Except maybe for those who are narrow minded and refuse to learn new ways of doing things. (See yourself in that description?) As for votes of confidence, why not take a look at the myriad of major distributions including KDE 4.2: Fedora, OpenSUSE, Mandriva, Kubuntu. By the way, I’m not a KDE fanboy; I use Parsix which uses Gnome as its desktop. I also point out that I’ve been using Gnome for several years. I am fascinated and excited about the progress of the KDE 4 branch and my experiences with it thus far are very good and usability has improved dramatically with this release. Of course this is my opinion and I’m entitled to it. Same as you.
Edited 2009-04-02 23:20 UTC
OK. So Debian now considers KDE4, released to the world for a year and a quarter now, to be stable enough to go into their “Unstable” branch. Hooray for KDE4!
Well, that’s weird of you.
No.
Well, heading up your list with Fedora was probably not the best strategy. They’d use /dev/null as their default filesystem device if they thought it would be fun.
Naw. Just practical.
Last time we talked about it you said you used Mint. I have not used Parsix. But I have been interested in it. Their python-based boot system seems pretty cool.
I was too. But they can only cry wolf so many times without sacrificing credibility. The times I have hopefully tried KDE 4.x, after hearing that “this release is the real one” it has been crap.
Edited 2009-04-02 23:47 UTC
Mr. Bergman, we often (mostly respectfully) disagree. However, I really must say this message was funny and I had a good laugh.
I’m sorry you had bad experiences with KDE 4.2. I suspect you are more critical than me, which isn’t weird. I can take some beating from KDE, since I run SVN, and still love it 😉
Stability in 4.2 isn’t at 3.5 (nor current gnome) level, at least that’s my personal opinion. So in that regard I agree with you. However, I think it is usable for most casual use, and I also think that it is an important step for the KDE 4 series. And obviously I expect stability to improve, so you’ll probably can change your mind some day and enjoy the goodies without the bad feelings.
Meanwhile, have fun @Gnomeland. Don’t get too bored.
Uh oh. This is starting off like a dunning notice. 😛
I admin 80ish Linux desktops. Boring is good. You guys keep wanting to inject
into my life. I just want my users to be able to get their work done effectively. If and when our goals ever align I would seriously consider KDE.
Cool. I completely understand you don’t put KDE 4.x on any corporate desktop yet – I know I wouldn’t. So that’s perfectly fine. But I have high hopes for the future 😉
I guess we can agree, then. Except for the high hopes for the future. Surprise me.
Edit: Eeek! I’m not sure how I managed to get the quote tags so badly wrong in my previous post. Apologies all around. I hate it when that happens.
Edited 2009-04-05 14:36 UTC
what you need is windows, is a SO that comes with almost every pc on the world ;D
I don’t think gnome shell can be appreciated by “mere words” – when reading the explanation, I hated hated the idea, (yes, that is a double hate) but when seeing a video of it, it seemed to start making sense.
The one I saw was from fudcon (and was an opverall desktio session, but a lot of it is about gnome shell):
http://alt.fedoraproject.org/pub/alt/videos/2009/FUDConF11/fudcon-f…
(go about 25% in to see the gnome shell stuff, but I found the whole video interesting)
As for gnome panel and taskbar – it may just be me, but I would struggle without one. I generally find the top panel to be less useful than the bottom one, so it puzzles me why they are planning to keep the functionality of the top one, but not the bottom (or rather they are undecided, but they seem to have realised its importance since the original planning stages.)
Edited 2009-04-02 17:08 UTC
The man’s presentation skills are non-existent.
I know. Was the two minutes of set up and of watching people mill around in the room really necessary? And I do agree that the presentation, once it actually started, was pretty lack-luster. And the video does pretty well establish Linux as the OS *not* to use if you want your presentation to go smoothly.
In general, all the (lack of) video editing, as well as the presentation multimedia and the presentations themselves from FUDCon tend to be on the embarrassing side. And I’m a Gnome fan!
Edited 2009-04-02 21:46 UTC
So to launch an application you press Activities, the whole screen turns into some sort of dashboard, you click the application and then it morphes back into a normal desktop? If that’s true and it’s how it’s going to be, I can tell you beforehand it is doomed to fail. Too much screen movement will make your eyes hurt. It is really innecessary. It will be another idea dumped into the recycler.
Yes, I was thinking about that too. In fact, it is not very different from an advanced global menu, such as kickoff or lancelot.
I know there are MANY who are going to say “leave well enough alone”, especially after the tussle with KDE 4, but incremental changes is not a permanent strategy. It is a good maintenance strategy, but every so often “new blood” has to come in. From what I can tell, they are doing a good job of interjecting new functionality (and removing cruft!) without causing the disruption that KDE 4.0 had.
It is always a balancing act between old and new. You can never make everyone happy. That is why leadership is crucial. In the past Gnome has been criticized for lacking the leadership to make the big decisions needed for these kinds of changes. This is a good test of the Gnome leadership processes – can they create enough “buzz” to bring in the new, without creating ill will that chases out the old.
You know what is sad? Reading the page http://live.gnome.org/GnomeShell/Features you can only conclude there is some great potential there. Then moving on, you see they created some gross hacks and are building it all from scratch. While there is a fully free, LGPL framework out there, which has been in development for over 2 years now, and is build to do specifically this: recreate the desktop experience in a new way. Sure, it’s KDE technology, but who cares? They can work with it in javascript, python, ruby, C# or a mix of those – so it can’t be the C++ they hate.
So what is the sodding argument to re-invent the wheel (*poorly*) again? They could have been finished by now, damnit, and the FOSS desktop would actually be innovative NOW and have a chance against the big boys. But no, they rather waste their time building upon a pile of dogshit.
Sorry that I’m so harsh but it is frustrating. PPL with excellent ideas and design capabilities choosing crappy technology for political reasons, letting their work go to waste. Blegh. The FOSS desktop could’ve been so much better.
Yes, I know what sucks. Everyone saying “GNOME should just use KDE this, ditch that in favor of KDE framework.” Well, if they did that, we wouldn’t have much variety now would we? No chance for innovation, for new ideas. Guys, KDE is not the end all be all.
Interesting how the KDE/QT guys don’t have (m)any problems with using GNOME/GTK technologies in the name of interoperabilty (dbus in place of dcop, using the glib event loop in qt, and there was another big one I can’t recall atm), yet when someone suggests that GNOME/GTK use a KDE/QT technology it’s like they’re asking GNOME devs to sacrifice virgins to the volcano gods.
It’s a technological issue, not a religious/political issue. Integrating C++ components into a C environment is MUCH more difficult than doing it the other way around.
I’m not saying the gnome devs should stop working on Gnome technology. I’m just saying they should think about using the right tool for the right job.
Sharing lower-level technology saves time for actual innovation. Duplication isn’t good for innovation – if you need something and it’s available, use it, so you can spend time on the implementation of the original part of your idea.
If gnome would start to build upon Qt and (some) KDE libraries, they would have more time to spend on innovation, finding new ways of doing things. They still have their own HIG and their own ideas about interfaces and usability – it wouldn’t mean Gnome would become KDE or the other way around. It would just result in an accelleration of development for both and more innovation on the Free Desktop.
Something I would love to see.
Anyway, qt is much better than gtk. And real desktops are written in C++. Not in basic, pyton, javascript, C#.
Imagine Windows 7 written in python! Muahahahahah…
KDE 4 is just at beginning right now. It will develop in a fully usable desktop at some point in the future.
I hope that firefox, open office and Gimp will be built with qt instead of gtk.
why would the gimp not use gtk (the gimp toolkit)?
Because gtk is dated and Qt is much better. Even winapi is better.
You obviously never tried to remember the 25,000 parameters that go into CreateWindowEx…
– Gilboa
Most desktops are written in C. Windows is predominantly C (see the Win32 API). GNOME is in C. OS X is predominantly Objective-C with some C legacy bits that are being ported over to Objective-C.
There is a lot of C code out there that’s still getting developed.
If so, what gnome has to do with python and C# code?
Python and C# are GTK bindings, GNOME’s official languages are C and C++. C++ was added recently.
Millions believe in some kind of God or Goddess. Doesn’t mean He/She/They exist.
Many companies are using C# these days – doesn’t mean its a good choice, just means the companies selling their service and ‘advice’ to them think it will make money.
I beg to differ.
While I make a living writing pure C code (mostly low level stuff) and I enjoy the freedom it gives, one cannot escape the harsh reality that outside the OSS world (GTK/TK/XLIB), pure C has been relegated to kernel modules, OS interfaces/services and high performance / RT software. (With C++ eating slowly into the RT space)
At best, the industry is using C++ for servers and C#/JAVA/VB for applications…. And God knows I’ve seen far too many performance oriented (!!!) network facing servers written in C#. (And before the flame war begins, in all cases the servers were a complete failure [performance wise] and have been rewritten in C++ [or C]).
Never the less, as somewhat that has spent quite a long time writing GUI in Win32, I can say that UI is one of the only places that I see little reason in using C as using C# or JAVA has a number of built-in advantages: gradual learning curve means you don’t need highly experienced developers *; it’s far easier to find/buy/use 3’rd party modules meaning you don’t have to develop everything from scratch on top of Xlib and Win32. All of this means cheaper and faster software development – the holly grail of the software industry.
– Gilboa
* Try explaining void pointer and/or union to a -well- seasoned C# developer and you’ll know what I mean.
Agree about the Win32 part, programming Win32 in C is masochism at its purest.
Edited 2009-04-04 20:13 UTC
Haha. That is a funny joke. We all know real desktops are written in ASM. Can you imagine anything usable in a high level language like C++ or Java?!!?!
…. really man, competition is good between toolkits as well as languages … and if you don’t like high level languages like Python/Ruby or scripting languages like Javascript/Basic then don’t use them.
QT is good. GTK is good. Pick the one that does what you need.
Completely agree. Give it about 5 years and it will have caught up with the requirements of our 2009 desktops. Of course, by then it will be 2014. But hey, KDE is progressing.
You are missing one vital point of why Qt simply isn’t an option for GNOME, and it probably never will be.
Qt != Gtk+.
As simple as that. GNOME is a Gtk+ environment, and it will most likely always be. Gtk+ probably leaves a lot to be desired from a development point of view, but that only means there’s a lot to improve.
KDE and GNOME both have their weak points, and it’s up to the competition between the two (and others, of course) to make sure they improve upon their weak points, benefiting the whole of the community. Let;s not forget that KDE 3.x was an extremely messy environment, ten billion million widgets all over the place. KDE 4 is trying hard to fix this – would this have happened in a similar way if GNOME had not been around?
Gtk+ will improve, thanks to the competition with Qt. Your post is just mindless fanboyism that does nobody any real good. If you really want to help the FOSS desktop, as you seem to imply you want, then write Gtk+ bug reports, write a long article detailing the flaws, so that the Gtk+ developers can actually improve their code.
But that’s really not what you want, is it? You are just here to proclaim the superiority of Qt (which is most likely accurate), but you’re not doing it for the greater good of the FOSS desktop – but only to take cheap shots at the competition.
Really lame.
Edited 2009-04-02 17:49 UTC
Why reinvent the wheel, when we have Qt, it works and it is commercially grade allredy?
I don’t see the point in improving gtk, if it takes 10 years. Better use something that is allready there and use the work to improve the software, not the toolkit.
What’s also sad is that, when an article about GNOME appears, the KDE fanboys have to jump all out of the woodwork and proclaim how much better their favorite environment is. To the KDE zealots: you guys put Apple zealots to shame in your sheer fanaticism, you could really teach them a thing or two. Funny how I don’t see a lot of GNOME fanatics jump on the KDE-flaming bandwagon unless provoked by one of you.
As for the changes, I’m not sure what to think. On one hand, I like a lot of GNOME’s interface at the moment as it is, but on the other, some of these new ideas really are intriguing. I probably won’t end up forming an opinion until I’ve played with it a bit, I like to have hands-on with a new UI or program before I judge it. I am happy at the removal of obsolete libraries, and hopefully they will also take this chance to really slim down the number of abstraction layers in GNOME itself. They’ve been moving towards that goal slowly, so I imagine we’ll see a much slimmed down GNOME come version 3.0, which would be a very welcome change indeed.
Lame is name calling — like calling people with a respectable track record like Jos “fanboys”. That’s not an argument, that’s just lame.
In one respect, you are right of course: it’s still GTK against Qt — but that’s all there’s to it. If it’s not gtk/glib based, it won’t need to apply.
Technical arguments are moot as long as libraries like akonadi, which was developed to be as cross-platform as anyone could wish, but using Qt, are barred from freedesktop.org.
But that doesn’t mean it wouldn’t be the right, technically right, design to implement new parts of Gnome using Qt, if there are already good, portable libraries that almost all the heavy lifting.
Whoa, go back to the last millenium. KDE vs. GNOME is no longer about Qt vs. GTK. The last toolkit barrier fell with Qt 4.5 (GTK style included by default and license change to LGPL).
These days it’s mainly about the usability philosophy since GNOME 2.x.
No developer seriously says that GNOME should drop GTK alltogether. However it’s absolute perfectly possible to create 100% GNOME HIG conforming applications with Qt. Qt itself is rather small (at least by today’s standards). So there’s probably no technical reason to not use Qt here and there in GNOME.
There are people who say that GNOME should adopt Plasma. This case it a bit more complicated. It really depends on how much KDE code is really required for Plasma. When many KDE libraries have to be loaded into memory then memory footprint is a technical reason not to use Plasma by default.
Luckily, on an individual level it doesn’t matter, because it is possible to use Plasma under GNOME. Thanks to common back-end standards used by both projects, clicking on the Plasma Log-Out buttons brings up the GNOME Log-Out window.
Perhaps conforming to most of the HIG, however one philosophy of GNOME is in regards to accessibility, and assistive technologies. At the moment, QT is not an option for this reason if no other, due to the lack of integration with at-spi. This is eventually going to be fixed, with the at-spi-dbus project aiming for useability by the end of the year, but for now any QT application would not be considered a GNOME application on those grounds alone. It’s not just about the HIG, it’s about integration, and as long as QT doesn’t integrate with all of GNOME’s core technologies no QT application would ever be considered for GNOME. As at-spi and the relevant assistive software is considered a core GNOME technology, this effectively bars QT apps. NOt trying to put down QT, just laying out the facts.
You sure have a point. However, investing in fixing this in Qt would be far more efficient than fixing everything else lacking in GTK2+.
I think it is time to start integrating Qt and GTK2+ more closely.
I’ve long felt that there is a *huge* potential in fixing usability issues in KDE and QT. I would consider moving to KDE3 if I thought there was a real chance of that happening. I’d even consider a migration of the business desktops I administer, over the next few years. But it has become so painfully apparent that there was *no* interest, in the KDE camp, in doing so, that the only reasonable thing to do has been to support, and use, the product of a desktop project whose devs cared. Or, perhaps I should say, one whose devs even recognized that there was a problem, which KDE devs simply don’t seem to do.
If you think you can convince your KDE buddies, then fantastic. I’ve already spoken to your Aaron Seigo. But he is so immersed in his own fantasy world that there seems no possible way of getting through to him.
Edited 2009-04-04 23:27 UTC
What kind of issues are you talking about? I see a lot of thought going into usability lately, actually since the 4.0 release and even before. There is still a lot left but the KDE usability team does nice work, and so do several individual developers. And some users contribute good things.
What is lacking are coherent usability reviews of apps with clear and concrete examples and suggestions for improvements.
Most ppl complaining about usability unfortunately continue barking up the wrong tree. For example – we all know how the plasma team, aseigo first, love the cashew, so wasting time complaining about that, as many do, is useless. Others complain about stuff which is obviously very hard if not impossible to improve, so minor even talking about it is a waste of time, or so dependent on personal preference nobody will ever agree on the best solution. Or they just scream ‘usability sucks’ and leave again.
If half of those people would actually spend that time on useful issues and quality reports, KDE would’ve been better than it is already.
For one thing… learn to “just say no” when populating menus. Putting every possible option (Open In Cervesia?) into the menus makes it more annoying for people to find the thing they want than most people are willing to put up with. (Though I think you did eventually get rid of Konqueror’s infamous “Open in Cervesia”, at least. Good for you.) You guys spend year after year paying lip service to usability. You say you are working on it. But your product never seems to reflect that any usability work has actually been done…
…And then you blame the users for KDE’s poor usability, after the fact.
Yep, that’s wrong. How do you suggest to fix it? If Cervisia is installed, I guess you can assume ppl will use it, right? Maybe that app should only install these submenu items when asked to? Then some users will never find it. So how about it. Put your money where your mouth is and give a good solution to this, and while at it, all the other useless items in the submenu of Dolphin. I’m sure Peter Penz would be very interested (I’m serious, it’s a great guy and very interested in usability). Oh, for your info, in KDE 4.3 the submenu items can be turned on and of one by one from the Dolphin configuration dialog. Don’t say nobody is working on this stuff.
No, I point out that many people are willing to waste their and the developers time complaining about stuff without being willing to be polite, let alone actually contribute 5 minutes to improving the situation. I’m tempted to say a lot like you, but of course I don’t know yet if I will see you next week on the kfm-devel mailinglist trying to work something out with Peter.
I think that there is no way around a lot of hard, boring work to resolve this. Hard work in the name of usability is something that the KDE team traditionally does well. But boring work? Not so well.
Someone is going to have to bite the bullet and look at things on a case by case basis.
And don’t look at me. Another project has already shown that it is willing to take on the hard work and the boring work. And it is what I and my users have been able to depend upon every day.
The KDE guys often say that they “do it for fun”. But sometimes essential usability work is not fun.
Edited 2009-04-05 14:46 UTC
Pretty weak response, bergman. Did you even read my post or didn’t you get further than the line you quoted?
I thought we had agreed upon civility as the venue of the day. I guess I was wrong.
You asked. And I gave you my best advice regarding what it takes to achieve true usability. But since it involved working on boring stuff, you as a KDE dev instantly rejected it. No surprise there. I have been watching you guys shirk your responsibilities for years.
Edited 2009-04-06 02:58 UTC
I didn’t wanna be mean or impolite, i was simply disappointed. To repeat my earlier statement:
But your response simply boiled down to: work harder, you lazy KDE developers!
Well, considering the amount of hard work and (free) time KDE developers have already put into this, while the person complaining (you) clearly only wants to spend time on COMPLAINING, not doing something usefull, I’d say that’s disappointing.
As I said, I expected a bit more.
Read what I said, Jos. You asked what I thought could be done about KDE’s menu mess. I said that someone was probably going to have to do that boring work, and take things on a case by case basis.
If you want to have true usability, the devs are going to have to be willing to do the boring work in addition to the fun work. Looking at KDE’s history, it doesn’t surprise me that you are having trouble registering that. No doubt you want to be able to write a few lines of code that will do it all for you. Buy sometimes you can’t just do that. Your can devs can kid themselves for years end, pretending that it is possible. But users still see the truth.
Oh, come on, you can’t pretend to think I don’t know that – of course it’s gonna be a lot of work! Nothing new there. Many developers already have been doing a lot of boring work in that area, and I think it’s sad that you don’t recognize that.
My point was (and is) that many users don’t seem to see how much has already been done, and how their time wasted complaining could be spend a lot better in actually helping out a bit. Many usability issues are very hard to find a solution for – the actual coding isn’t the worst part. Yet the KDE usability team has a huge lack of resources.
And in that specific instance, I asked if YOU knew a good solution to the ‘mess in the menu’. Apparently, you have no clue (I’m assuming here that if you did, you would’ve given your input). Well, don’t be surprised if the developers don’t, either. We NEED people who are willing to help think about creative and effective solutions for such issues. Usability isn’t about removing features (like gnome did). We think usability is about finding better solutions which are both powerful and easy – like Peter did when he wrote dolphin’s breadcrump bar.
I think it’s an insult to all the developers who have spend so much of their time doing exactly what you’re saying here they never did. Luckily most users recognize that.
As I said. You guys have never had a problem with hard work. It’s the boring work that your guys avoid like the plague. KDE programmers say over and over again that they work on KDE for the fun of it. Perhaps that has its drawbacks. Perhaps it is not the route to the best desktop environment. Perhaps it gets you part of the way there, but not the whole way there.
I don’t have a step by step recipe. The more popular Gnome Desktop Environment solved the problem some years ago. Perhaps you should ask its devs.
Really? I would think that those who have would take it as a vote of confidence. If anything, I would expect them to be put off that they have never received adequate support and help from the rest of the KDE project.
Edited 2009-04-07 01:02 UTC
They ARE doing the boring work. Take the plasma team – for the 4.2 release they put aside pretty much all the cool things they had been working on (like support for small formfactors, special N810 interface etc) and focussed on re-implementing the stuff users had been asking for, like an auto-hiding panel. As I said, it’s an insult to say they don’t want to do the boring work – as they have already proven they DO.
Well, they simply don’t have a proper pluginstructure and lack the features. That’s not usability, that’s – well, lack of features. Not the way we want to handle our users – by taking things away from them. I don’t see Gnome come up with REAL solutions, only stop-gap solutions. Their marketing might make them more popular in the US, but pretty much large deployments outside the US choose KDE – for exactly that reason.
Bullshit. If you don’t know what you’re talking about, please shut up.
Enjoy further ‘discussion’, I’m out.
You’re glossing over a lot. It’s not just about whether you can create HIG compliant applications with QT. QT does a hell of a lot more than GTK. GNOME relies on several external libraries that do essentially the same thing as parts of QT. Take Cairo for example. What do you do with an application that depends on Cairo? Do you eliminate the Cairo dependency and integrate it with QT or do you leave Cairo and end up with a lot of extra and uneeded libs that only complicate things when someone decides to make a new app that doesn’t use Cairo but uses QT directly instead. If you ditch Cairo and other libs like it you end up rewriting GNOME from scratch. That doesn’t seem very helpful either. Ditching GNOME’s GObject system for a more C++ compatible object system would also probably be necessary and make the task even more arduous.
Gnome doesn’t have to be GTK, and QT != KDE. There is no doubting that it would be a large effort to rewrite applications for a different framework, but there’s not reason not to have a meaningful discussion about doing so.
I think a lot of people would appreciate standardising on one toolkit. For a start, it would give a cleaner environment, it would allow more code to be shared be shared between KDE and GNOME, and it would make lives easier for those who currently try to get their applications to work with both GTK and QT.
It doesn’t stop GNOME and KDE still going their separate ways!
I’m not saying Gnome should stop working on what they have, I’m just saying they should be more pragmatic and choose the right tool for the right job. Sharing more low-level functionality between KDE and Gnome would result in more time for both developer teams to work on actual innovation instead of rewriting each others frameworks. If you call that mindless fanboyism – well, sounds a lot like ‘blablabla I can’t hear you blablabla’ with your hands over your ears.
All the “superior” Qt technology in the world won’t fix your arrogance, And you just gave me another excuse to reject KDE:
btw, the KDE code of conduct is a joke.
Please, just dissapear.
So what is the sodding argument to re-invent the wheel (*poorly*) again? They could have been finished by now, damnit, and the FOSS desktop would actually be innovative NOW and have a chance against the big boys. But no, they rather waste their time building upon a pile of dogshit.
You know, one of the reasons I don’t use KDE at all its for the arrogant mentality of its members, and you do not contribute to it, Don’t like GNOME, find don’t post in GNOME related topics, I and Im sure many more and sick and tired of your trolling here.
You are always asking for construstive critics for KDE, do the same for GNOME or don’t post at all.
Edited 2009-04-02 17:50 UTC
Cut the crap. Until I even reached this thread, I’ve already read about the “disaster KDE 4”, the superiour incremental GNOME development model, etc. several times within these comments.
No much for arrogant mentality of only the KDE users…..
You just proved my point.
The Plasma scripting api used for writing javascript, python, ruby and c# plasmoids is not the same as the C++ api that is used to develop the heavyweight plasmoids such as the task manager. There are quite a few things that you can do in C++ that you can’t do in those languages, and conversely you can’t download and run C++ plasmoids from the internet like you can with the scripting ones.
That is a deliberate design decision, but I think the Gnome shell approach of using a GObject-Introspection based JavaScript binding as the primary implementation language is interestingly different to the Plasma’s approach of using C++ as a foundation.
No, it’s not a political reason. It’s a personal reason — “too large ego” / “not invented here”. Just like the Konqueror developers who refuse to adopt QtWebKit and instead stick to KHTML until the bitter end.
A political reason would be “I don’t use this software, because its license is not free and I support democracy and freedom”.
There are LOTS of cool, cooperative people in both projects. They are the people who cooperate within FreeDesktop.org, work on shared back-end technologies like dbus, and jointly organize the Gran Canaria Desktop Summit (Akademy+GUADEC) this summer.
Those people know that the main difference between GNOME and KDE is the HIG philosophy, but that GNOME and KDE have the same political (!) goal: Freedom for the computer desktop.
Being real, I don’t care the toolkit they use, GTK being inferior or not to Qt always agve what I wanted, GNOME having less options or less customisable still gives what I want in the way I want it.
Sure. My point however was that if developers would be more pragmatic you would have even better software. Gnome is more than GTK2+ – you clearly show that in your comment. So wouldn’t it be good if they would accellerate development?
Guess most people around here are too young to remember this, but Zeitgeist has some resemblance to NeXTstep’s old file management concept. I really enjoyed using it, it’s concept really made sense. Glad to see the idea revived nowadays.
Actually, it’s not so similar because the columns in Zeitgeist are dates, and not related to the directory structure like the NeXT file viewer.
Furthermore, it’s only the UI part and it will probably change.
wow , what a great idea !!!!
wow … dunno why nobody ever thought of having a desktop with activities where you change your “overlay”.
kde4 is dead , it doesnt have activities….
ohhh wait…
When doing radical changes like this, why take such a large and dangerous leap? Many new ideas that sound cool in theory just don’t work out in practice so why not just create a gnome fork and if it is awesome then it will just take over naturally and if it sucks then it will just die. That way users aren’t forced to adopt something that may not work and the transition isn’t as violent.
That’s just the way it is planned: please, RTFA.
I found some screenshots of Gnome Shell at this site for those, like me, that had never heard of it and wanted to see some pics:
http://blog.fishsoup.net/2009/02/10/gnome-shell-status-2/
(hope it was okay to post that…I have nothing to do with that blog, just found it doing a search)
From the announcement:
– include new exciting technologies that we’re starting to see used in
our desktop. Some obvious examples are 3D effects (with Clutter) and
geolocalization (with GeoClue and libchamplain).
Cool! I’ve had a look at some Clutter videos on YouTube and it seems like it gives people a way of creating some really fantastic interfaces. However, I am interested as to how well it will perform on machines with little or no support for hardware-accelerated rendering, which are unfortunately quite common in the world of FOSS operating systems. Is it possible to integrate this technology into the desktop environment without leaving a bunch of people out in the cold?
GNOME Shell doesn’t really have a fallback mode for older systems. They can probably continue using metacity + old gnome-panel I guess.
It is linked from the post itself but read the thread at
http://mail.gnome.org/archives/desktop-devel-list/2009-March/msg001…
Just one question, When?
Should be like 3, 6 Months? or Like 1 or two Years from Now?.
GNOME 2.30 is going to be GNOME 3.0 so atleast a year or so away.
A little under 1 year away. 2.28 is scheduled for the Fall of this year; and, 2.30 (== 3.0) should be in the Spring of 2010.
You can have the best technology in the world but without happy users you have zilch, and soon you’ll have no users anyway. I think we are spoiled for choice on Linux with the amount of new development and new ideas coming on stream, but as always the key is going to be presenting an environment that users can get on with. They want to concentrate on running their apps and not the computer (which should take care of its own stuff) and they don’t want to sacrifice functionality just because it’s Linux or Gnome or Windows or KDE or whatever. Given the slow but steady progress of online apps which bypass such restrictions, any platform which tries the old “Oh, we don’t do that one here” is surely doomed.
If Gnome hits this with version 3 then great, and ditto KDE with the 4.x series. We’ll see. The important thing, I think, lies in never forgetting that appealing primarily to developers or Linux cognoscenti is appealing to a niche within a niche and it simply won’t fly. Users want all their stuff – from cameras or mobiles to spreadsheets or Facebook – and they want it working fully and with no Byzantine config required. It’ll be interesting to see what their respective user numbers are in, say, 2-3 years.
Took one look at the screenshot, borders, menus, most recently used, active documents and all kinds of stuff eating up screen space, and thought, thank Heaven for Fluxbox. Going to be time to change soon. And get away from HIG fascism. Are we who just want the environment to get away and stay away while we are working in a vanishing minority?
Its like being unable to listen to a concert because the engineers have decided to fill it up with a whole bunch of artificial ambience noises and historical explanations of the composer’s life that you know already…and so on and so on. Till you want to scream at them to just shut up and let you listen to the music.
So what I feel is, get out of the way will you, and let me work!
The overlay is designed to only be open for relatively short periods of time, and most typically a few seconds if you find what you want through search.
I also definitely think the role of the desktop is to be out of the way as much as possible, until you need it.
I had a much better idea. I switched to XFCE.
About time too, after months on my hard drive being thrashed I finally realized it was GNOME’s fault.
Now I just need to switch to Opera, and that’ll be the end of bloatware for me.
I’ll wait to see how it feels when I’m using it. If it turns out to be not to my liking, there are plenty of other WM/DE options out there from which to choose.
On the subject of the actual article, I congratulate the Gnome guys on trying to do something different that helps people to organise their desktops beyond what we’ve had for a couple of decades. We need to move forwards and come up with challenging new ideas. Stagnation can’t be an option.
I like this here:
“move all of the deprecated libraries out of the platform, so people stop using them in new code”
What they need is a 2 release cycle where they say that a api is deprecated and it is totally removed in 2 releases (deprecated for example in 2.24 and totally removed in 2.28) that gives the relevant project over a year to upgrade their code, which imho is more than enough time.
If you’re going to break API compatibility, then that should be a major-version upgrade. You can add to them, but API’s shouldn’t break in point upgrades.
For better or worse, the blogosphere was outraged that KDE released a developer preview as a final version with 4.0. That would be a tempest in a teapot compared to the outcry if Gnome starts breaking compatibility with incremental upgrades and starts confusing the hell out of everyone as to what libraries and APIs should be used with which point version.
Just my 2c.
You need to re-read what I wrote because in no place did I advocate a KDE 4.x solution. I stated that deprecated parts should be gradually removed not all at once – and not until a superior and stable replacement has taken its place. KDE 4.x was a Bennite solution to something that could have been done via evolution rather than a ‘great leap forward’ with the body count of many behind the change.
Not just the blogsphere. Pretty much everyone noticed that they were so anxious to get out their Gnome killer that they shitted all over themselves in the process.
I think you are completely wrong about the motivation there. The developers had spent YEARS of time working on the project, without any release. The way I remember, the whole internet was talking about how KDE4 was vapourware and never going to come out. In that light, following the old release often and early strategy doesn’t seem so stupid.
The mistake was all the hype that got generated around the 4.0 launch. All the talk was about what they wanted KDE4 to become, and people missed the fact that they weren’t anywhere near making that happen yet.
Well, the way I remember it KDE fans were saying that as soon as KDE4 was out Gnome was as good as dead. A year and a quarter after the premature KDE 4.0 release, Gnome is still the #1 Linux desktop, KDE4 is still vaporware, and the KDE devs are still cleaning the shit off themselves.
Yeah, that was part of the hype I was talking about.
Always debatable. Gnome is clearly #1 in commercial desktops, and in the US. KDE dominates in South America and Asia from what I understand, and obviously that doesn’t bring the press that the US market does. Most of those are probably still running KDE3.5 actually, in long term support situations.
Disagree strongly. 4.2 is here and it’s usable. Still definitely rough around the edges, but the releases so far have followed the same path that OSX did. Is 4.2 as good as Gnome? No. But it’s getting close.
Ok, you seem to be having some sort of emotional reaction here. Did monkeys throw some poo on you as a child, because you seem to be awfully fixated on the word “shit”.
Relax and enjoy the ride. KDE and Gnome are both going to be around for a long time, and I think they both benefit from the competition and the ability to explore different ideas.
Edited 2009-04-04 04:05 UTC
Reading it keeps me hopeful but I was put off by the screen shot a little bit. No matter, too soon for me to have any real opinion yet.
what about a dock???, like gnome-do or awn.
i can see alot of gnome/ubuntu users crying ;./
Please don’t!
This is all getting so frustrating for me. I’m a Windows refugee and lazy OSX user. But the GDI font rendering / anti-aliasing of OSX drives me nuts after a few ours in front of the screen. So I’ve recently tested a few Linux and BSD distros and found a GNOME based Linux to be the most easiest to adapt to. GNOME is simple and clean. And with a “good” theme engine it looks as nice as OSX… 😉
Having no super duper futuristic stylish desktop concept is not bad. Some people chose GNOME exactly for its simplicity and because it’s easy to customize.
Edited 2009-04-03 08:17 UTC
Gnome seems to me to be becoming a desktop environment with a split personality. I think the only way not to end up with an over sized and over complex behemoth, is to get back to basics. Why try making a desktop do too much?
Let’s play a mind game; Imagine if gnome had no applications. None at all. What are the bare minimum set of functions expected of the desktop environment?
– eye candy (window decorations, alpha blends, themes)
– an effective way to open external apps
– default font & color management
– provide ways for apps to communicate; dbus, clipboard, IPC interface wrappers
– provide one or more workspaces
– provide interfaces to operating system functions
– file system interface(s) incl. search, bookmarks, thumbnails etc.
– networking configuration
– printer configuration
– display hardware configuration
– keyboard configuration
– USB or other removable device mount/unmount
– RPC & initd services config
– user and general security management
– sound hardware config & control
– software package management & installation
– crontab management
– mail subsystem configuration
– remote display config & management
– basic command line interface
– desktop management (icons, desktop folders and intuitive user interface)
The latest idea of using Clutter seems to be addressing only the last issue, that of the Desktop paradigm. Even if 70% of users actually like the Clutter idea, this would be alienating 30% of Gnome users.
Gnome would be better off defining a standard interface for desktop functions, and allow the choice of Nautilus, Compiz, Clutter or any others that follow to remain in the hands of the user.
I strongly recommend watching the following videos:
http://live.gnome.org/GnomeShell/Screencasts
Very… very… very… cool!!!
After seeing those screencasts, I think it looks like a total nightmare.
Cool != Useful
Gnome is great.
Gnome Desktop is just 11 years old and GNOME Mobile and Embedded Initiative Was reading here:
http://www.techunits.com/content/list_all/72/gnome“