GNOME vs. KDE: the Latest Round
jammag writes "The debate about whether KDE or GNOME is the better Linux desktop is longstanding. Yet as Linux pundit Bruce Byfield discusses, it has entered a fresh chapter now that both desktop environments have versions that are radically different from their incarnations just a few years back. Moreover, 'the differences in KDE 4.6 and GNOME 3 (the latest releases) are greater than they have ever been,' he writes. Casting aside his usual diplomacy, Byfield acknowledges that he's heard rave reviews about GNOME 3, but disagrees: 'I suspect that the majority of users are more likely to be satisfied with KDE 4.6 than GNOME 3.'"
I've preferred to use Gnome over recent years as I just found KDE to be not right - couldn't get on with it. With the way both are now going, I can see myself having to switch again. Given my recent hunting round, I really hope that the Enlightenment crew actually get their shit together and get a stable, solid release that can be used as it is simple, clean, easy to use, easy to configure and add gadgets to.
The whole idea of linux is choice. I run xfce4; once in a fit of stupidy I suggested my wife log in using KDE as it was closer to Windows and not as sparse as XFCE. Bad idea.... Turns out some people (4 for 4 in my family) prefer the sparseness of XFCE to any complicated desktop. I know this will bring forth an avalanche of "What about Ratpoison, Windowmaker, etc, etc, etc?"
Exactly. Run what you like and let the pundits amuse themselves.
Having been in both communities, I characterize the Gnome community as very MS-like in these more modern times.
While working at MS, I saw a lot of the same "Not invented here" crap that I see in the gnome community on a daily basis. I also saw the same political maneuvering, the same tribal fears, and in general the same of what is in my own personal opinion a great lack of regard for others over their own projects/groups/goals.
I see the KDE group as entirely different. They work as a team, have the same common goals (in general) and let good ideas thrive even if it violates somebodies pet project or personal goal.
Posting AC an not mentioning the company by name for obvious legal reasons, but consider your here I figure your smart enough to get what I'm trying to say.
true infidels go with Emacs/BSD
Funny! All these desktops look the same from inside a command prompt.
Chas - The one, the only.
THANK GOD!!!
We better figure this one quick, seeing as how this is going to be the the year of the Linux desktop...
XFCE, LXDE, EDE, Enlightenment, ...
plus all of the alternative window managers like Openbox, Fluxbox, IceWM, FVWM, twm ...
The whole desktop thing is overblown. I have very little use for widgets or what what ever your desktop calls program updated icons. As far as customization that can also go too far. I want a nice clean UI elements and wall paper. The big weakness for the desktop right now are notifications. What it really comes down to is the API as far as I am concerned. Your desktop environment is used to launch apps and maybe manage files. Everything else is just fluff. The API that it offers the developer is the key IMHO. Yes having complete scripting control is cute but who cares? I use a computer to do thing.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
I run an architecture firm entirely on Linux. All our workstations have two reasonably big screens and use Gnome. I have used Gnome since its earliest inception in various flavers of Redhat, Fedora and Ubuntu.
I have to say that as much as I don't want to, we are going to have to change to Xfce or some other alternative. Gnome shell is a disaster for the way we work. I can't believe that the developers and UI designers have completely failed to take into account those of us that are actually using our workstations to do heavy duty computational, graphic and design work.
We have spent the last 20 years moving to ever larger and multiple screens because we need the real estate. Now we are supposed to work as if we were using a cell phone? What a joke.
The developers need a good whack will a clue stick. As does Redhat. The least they could do is have a fall back to the Gnome 2 series.
We don't want to be the subject of an experimentet about how we "should be working."
This is serious business to us and has a big effect on our bottom line.
Kurt
O'really? Jolicloud doesn't have anything to switch to; Its a web page that stores your browser bookmarks and thats pretty much it. I can do the same thing in firefox, without using that proprietary software, and get the same functionality without needing to worry about not having the software I want to run. Jolicloud doen't offer anything special, and it provides no value whatsoever for most people.
- d
One was too austere, they other over-eye-candied. Neither had any significant impovemnets in functionality over earlier versions.
Now, Gnome and KDE just get in the way of using my desktop environment to complete actual work.
Hello to IceWm and LXDE.
"I believe in Karma. That means I can do bad things to people all day long and I assume they deserve it." : Dogbert
That would be hard to do, seeing as it's already GNU/Emacs. GNU/BSD just seems unnatural. (oh wait)
Bio questions? Ask me to start a Q&A journal. Computer analogies available for most topics!
However this is a problem that the entire software field faces. java tried to solve this problem with the jvm, but then Microsoft released .net and created another choice for developers and companies.
i think choice is more good than bad. the alternative means you are stuck with what is there.
Byfield acknowledges that he's heard rave reviews about GNOME 3, but disagrees: 'I suspect that the majority of users are more likely to be satisfied with KDE 4.6 than GNOME 3.'"
I've actively sought out reviews and have yet to read a single positive review of Gnome 3. Not one. In fact, they are as universally bad as they are universally duplicates of each other. They all seem to very quickly identify and cite the same core problems with Gnome 3's usability, the specific and seemingly broken process which yielded Gnome 3, but also touch on Gnome's process failures and general lack of specification and healthy process.
I'm personally excited to see what all the brouhaha is about with Gnome 3 (hell, can always revert to Gnome 2 or KDE), and I say that as a current Gnome 2 user, but frankly, based on a wide number of reviews, I have exceptionally low expectations of Gnome 3.
Seriously, if you know of some good, unbiased Gnome 3 reviews, please post them here. Thus far, I've never read a single one.
KDE doesn't emphasize on putting Ks into the application names anymore. The new file manager is Dolphin, other K-less apps include Marble, Gwenview, and there's the whole Plasma interface. That said, you still have apps like Konsole and Kmail, but there are several that don't have the K.
As for the feel of the apps, that's entirely up to you. You'll have to give KDE 4.6 a run to see for yourself.
'I suspect that the majority of users are more likely to be satisfied with KDE 4.6 than GNOME 3.'
I'm certain that the majority of users are likely to wish developers would stop fucking with the interface they're already comfortable and familiar with and find something more useful to do with their time.
Proud member of the Weirdo-American community.
Around ten years ago, I was interested in building some GUI apps for Linux, but there was no clear path as to which of the two GUI APIs I should learn. I found the lack of a clear path to be enough of a discouragement that I ended up losing interest. I doubt that I'm the only one who has felt that way about it.
You're doing it wrong. Go with whatever API / toolkit you prefer. I'll use your software if its good even if it isn't 100% with my desktop environment of choice. In fact, I'm more likely to continue using your excellent software no matter how much taste might change and motivate me to move to a different environment.
I understand that this seems strange to someone from a different environment. But this is Linux. The chaos is a feature.
What others?
There's a billion window managers, but very few desktop environments in the sense that GNOME and KDE are.
A few of them:
I'm probably missing one or two, but that's pretty much it. Running some window manager with a few KDE or GNOME programs doesn't give you the full experience of the desktop environment. That's fine for some, like me and you, but a lot of people really want the integration and whatnot.
The argument is important not so much to the Linux world, where most distros give you the flexibility to run either, but to the commercial Unix world and companies who use commercial Unix software or inhouse software. For example, Sun went with GNOME starting with Solaris 10 (I think). That was a big blow for KDE at the time, because anyone writing commercial apps for Solaris pretty much had to switch to GNOME. Sure, you could run KDE on Solaris, but try convincing your customers to switch desktop environment just for your little program.
Those who can't do, teach. Those who can't teach either, do tech support.
It used to be differences in compiler technology (C vs C++) made Gtk+ based applications and frameworks much faster in start up and also a slight edge in run time. Add to the fact KDE traditionally tried to be an extremely crappy Windows wanna-be, most naturally gravitated toward Gtk+ (meaning Gnome).
These days, compiler improvements have come a long ways and KDE (Qt) applications no longer have performance penalties. Furthermore, KDE has grown considerable beyond their windows wanna-be days. By all accounts, they are an excellent framework/desktop in their own right. Both have strong offerings both in features and applications. In fact, despite me being a Gnome user, as a developer, Gtk+ absolutely sucks compared to Qt - although Qt has some real kludges and warts. Though I've not recently looked and more recent versions of Qt may address some or all of these - really not sure.
Which means, now, the appeal is largely based on user entrenchment and application preference. It wasn't so long ago memory was still a deciding factor and running two frameworks was not a satisfactory solution which further forced users into one camp over the other - again, based on application preferences. These days, with 16G become more and more common, the overhead of mixing and matching doesn't pose anywhere near the downside it once did. As such, running Gnome desktop and some KDE apps, or the inverse, is far more likely to be much more palatable. I predict this to become more and more common over the next couple of years.
Why bother fighting about it?
I agree with you. Historically there were good reasons to be in one camp over there other. These days, IMOHO, it large boils down to available memory and the preferred application mix. I strongly suspect Gnome 3's ability to hit or miss at time of release will potentially mean a massive influx of KDE users. And based on all reviews I've read to date, I strongly suspect KDE will prove the real winner once Gnome 3 is finally released. Of course, I'm hoping that's not the case. But either way, as you point out, ultimately, it may not matter one way or the other.
Green!
Purple!
You can install Gnome & KDE apps side-by-side and they just work.
This is not the case, you probably missed the slashdot story on the drama between GNOME and Canonical. In particular see Aaron Seigo's rant on how GNOME ignored "status notifiers", a cross desktop specification submitted to Freedesktop.org and with an existing implementation by Canonical.
Not to pick on you in particular but I am sooooooo tired of hearing the claim that "choice is a good thing". It's not. In fact, a good way to frustrate people is to give them too many choices. Moreover, the wide choice of windows managers is an example of Linux market failure. People don't use computers to run various windows managers, they use computers to run applications that perform tasks. The fragmentation of low-level libraries for sound, graphics, UI, packaging, etc., means that developers don't have a clear target for Linux apps. For open source efforts, this means wasted efforts on ports, plugins, and duplicate projects. For commercial ventures, it means that additional money must be invested to reach a more restricted market segment.
-1, Too Many Layers Of Abstraction
Oops... here's a non-paywall'ed article about the tyranny of choice.
-1, Too Many Layers Of Abstraction
IMHO, and from a "power user's" perspective, there is just way too much clutter in KDE (from the break from 3.5 and onwards), and all of the background services, databases, yadda yadda yadda - just make for a more complex journey through the daily working. KMail and all that that entails are dreadfully slow, dreadfully NON-standard (especially with HTML and RTF mail); as well, I have a tendency to utilise ancient hardware for my own purposes - not the fancy dancey brand-new stuff, therefore, KDE moves like a herd of turtles across a plain of peanut butter. GNOME, OTOH, allows me to streamline it and sterilise it enough to make it daily usable and stable - especially having Compiz functionality, Cairo-Dock, Screenlets and what-have-you. It's not prone to having "hissy fits" and just hanging on some background service dependency. XFce4 quite nicely fills in the secondary spot - it's always fast, always happy with GTK or QT stuff, doesn't hang, highly customisable, usable, and not primitive in graphics or function (sorry IceWM and LXDE)... If you really get down to brass tacks, there's always WindowMaker - most stable of 'em all.
YankDownUnder Veni, Vidi, volo in domum redire
Go back two or three years, and I think you had the same situation when KDE 4 came out. Everyone and their dogs hated it and 'switched' to Gnome. Sounds like it's reverse that's happening here. For what it's worth, I appreciate it when developers exit the status-quo and create something new. Remember when Firefox first came out? Then Chrome? Diversity is a good thing in my opinion.
Goodbye Slashdot. You've changed.
GNOME 2 or KDE 3.x with compiz were very close to the ideal desktop. If GNOME would have made 2.x's yelp help browser startup faster, it would have nearly been a perfect "minimal" desktop. KDE 3.x was a little further away but was still close to a perfect "power user" desktop. Now we are stuck with two less than optimal desktops that, despite the goal of being easier to use, seem more confusing for beginners. Devs MUST learn that past some point of complexity, evolutionary change is the only way to go.
KDE 3 was basically finished. It got as good as it could possibly get. At that point it had a bunch of enthusiastic developers who wanted to code but didn't have anything worth doing. They kept coding anyway and replaced sensible stuff with less sensible stuff, they kept braking things and turned a good product into an ugly mess.
This happens to software sometimes, nobody tells the developers that they have finished and it's time to stop.
I won't belabor the point and of course, you use what you want. It just seems that by sticking with the upstream projects, not only will you get more timely updates but, the upstream stuff seems to be more feature complete anyway. The only downside I see is lxde has some custom configuration guis but it's really not that big of a deal (to me).
The soylentnews experiment has been a dismal failure.
In particular see Aaron Seigo's rant on how GNOME ignored "status notifiers", a cross desktop specification submitted to Freedesktop.org and with an existing implementation by Canonical.
And don't forget informative overview of this drama by Jeff: http://bethesignal.org/blog/2011/03/12/relationship-between-canonical-gnome/
Gnome was *more* customizable before Ubuntu came. Now I'm not going to blame it on Canonical since the urge to reduce configuration stems from Gnome itself.
My favorite example is the size of the buttons in the Window List applet, the Gnome 2 equivalent to the taskbar.
When I started using Ubuntu back in Warthy, in other words from the very beginning, I configured the Window List to display the widest possible buttons, meaning that no matter how many windows I had open the taskbar never had empty space.
Eventually they removed that option from the preferences dialog. I realized it was still there in gconf since it inherited my preferences after an incremental upgrade.
Later I made a clean upgrade and simply changed the gconf keys I wanted.
Then it stopped respecting that setting. It's setting there its just deprecated (since Gnome 2.20) Why? It was deeply hidden in gconf so they can't argue it was crowding the preferences dialog.
It was removed just because. It was an option they wanted me not to have.
KDE on the other hand infuriates me by insisting on hogging the corner hot spots for its own use, in other words, KDE has become as more customizable than Gnome to me.
It shocks me that I can't configure what icons or actions I want for the corners, but I can freely rotate my rss reader!!
WTF? WhyTF would I want to rotate a rss reader? And why would I want to use a semitransparent feed reader that is only displayed as a desktop widget?
It doesn't have the immediacy of a panel applet, nor the capabilities of a full reader like akregator.
Lately its getting harder and harder to set my linux desktop "my way" /rant
But... the future refused to change.
Use Dolphin or KWord: massive toolbars and small content area.
Good news:
Dolphin in the upcoming version will change that: http://3.bp.blogspot.com/-9ujy04d0LMc/TY-FyUfXOuI/AAAAAAAAAeA/e6QxAfTjTXM/s1600/dolphin-default-4-6.png
KWord is dead, btw. Its maintainer supposedly was a dickhead so the KOffice crew left him altogether and created Calligra Suite with a new word processor forked from KWord. It'll take a while for the first Calligra release but some GUI aspects may change especially considering that the Calligra crew is also targeting mobile devices with small screens (something the old KWord maintainer fiercely fought against because he wanted to "concentrate on desktops with big screens").
Deep in my heart I'm a WindowMaker/GNUstep guy. Unfortunately, that environment is particularly 'all or nothing' and without a reasonably browser, office suite, image editor, I have to use non-GNUstep apps and the experience breaks down quickly. I also really really don't want to go without 'scale windows with window title filter'/'present windows' now that I have it.
XFCE/LXDE are nice enough, but lacking certain features I want that come with a larger user base.
Gnome has been quite sufficient and gvfs with fuse does a *lot* for having arbitrary applications enabled for non-admin access to network resources. The problem has been they have been fighting a war against configurability. It's bad enough they don't want to present a UI, but they don't even want to add 'hidden' gconf options even when given patches. Gnome 3 has been the last straw for me, going too far in forcing the specific vision of the developers.
Unity offers an alternative, but suffers the same fate of their way or no way (not even able to move their 'dock'.
Currently I'm in the KDE4 camp. A lot of the defaults were not what I wanted, but I was able to configure it easily enough to fit my preferences. One issue I do have is they are on their high horse on KIO, and have outright refused to embrace some fuse based bridge to ease life on people forced to use applications that aren't KDE. This is even worse because out of the box most distros select the Xine phonon backend instead of gstreamer, meaning KDE's own media players cannot even use KIO. Embracing fuse out-of-the-box to provide a POSIX entrypoint into KIO would fully get me inte KDE.
XML is like violence. If it doesn't solve the problem, use more.