ZDNet Compares Linux Desktops: Concludes KDE Best, But Not Good Enough

ZDNet has published a review by Jason Brooks of eWeek Labs comparing the Linux desktops. He writes: "eWeek Labs found that KDE (K Desktop Environment) comes much closer to
delivering the sort of smooth interface that users have come to expect from the
Macintosh and Windows operating systems than does GNOME (GNU Network
Object Model Environment). In tests, KDE delivered snappier and more polished
performance than did GNOME on the same hardware."
However, he continues, "neither desktop interface has yet reached parity with the established
players-pervasive support for features such as cut and paste across the
interface can still be unpredictable."
Strange to have picked out cut-and-paste, as that should be much improved in Qt-3.0.

Comments

> *Hmm, KDE crashes. OSS Rule #1534: If it seems like everything is crashing, then something is misconfigured.

Oh yeah? I had KDE crashing all the time last year. My file system was getting corrupted and KDE would not even compile without taking down my system. Know what the problem was? Bad RAM! That's right! I tested 36 bytes on one of my 128 MB DIMMs choking badly.

Frankly I'm extremely dissapointed that KDE, the kernel and other OSS software could not operate correctly with defective hardware! Of course if I had been running windoze I would not have noticed anything unusual since I expect it to crash.

Since I have not gotten a response on my bad RAM bug report I'm beginning to suspect that, as I don't run winblows, OSS software must have broke it! Clearly this proves that KDE is not only buggy but dangerous.

;-)
To the rational empirical evaluations yield results. To the rest the revelations of cause and effect are just another conspiracy theory and illusion is more substantial than proof.

>To the rational empirical evaluations yield results. To the rest the revelations of cause and effect are just another conspiracy theory and illusion is more substantial than proof.

Well said. Is that a quote from somewhere?

Hi!

I know i should do some more detailed failure descriptions, but maybe you are only playing with your linux box. I try to get some work done, and if i would report every bug i discover this would be a fulltime job *gg*

And the main thing I wanted to say is that the linux-community should realy stop to release every buggy extension to the kernel as so called release. And also kde 2 should IMHO be released as 2.0 in maybe a half year....
And what should be relly done is some serious testing. Testing is one of the most importand parts of software engeniering.

You can go on an play with every little kernel quirk and recompile all your apps. But I've got some work to be done.

cu Robert

You completely missed my point.

Here, I'll make it easy : what you are doing is trolling. Trolling is a waste of bandwidth, completely unhelpful, and very annoying. You are trolling because you are doing this : complaining endlessly in a completely unhelpful way. You seem to have nothing to say but "Haha, I'm better then you, cause my software is stabler then yours, so there".

Moreover, I do not recompile all my apps regularly, and I do not play with every little kernel 'quirk'. Neither do the people who run Linux boxen and serven on a production level basis.

If you have no time to waste, and want to get to work, then please, by all means, get to it! There is no need to waste more time on dot.kde.org displaying your extreme superiority in using a "certified" OS.

indeed, this is totally useless. you keep saying the same things without even trying to make some arguments.
what 'buggy kernel extentions" that are called 'release' are you talking about ? what bugs are you talking about ?
Also your point of view the entire oss community is a bunch of idiots playing around while being useless is at least a little bit unrealistic, not to say completely insane. think about it (if you dare).
Maybe you should try to get some work done instead of trying to work out your frustrations of your missed attempt to install linux (caus thats what you wanted to do didnot you) ... by the way we also get our work done , fast and efficient even. its not because you can't use linux/x/kde/... noone can.

Hello;
I simply agree with every single word you 've just said.

Well... I wouldn't say it's all true.
First of all, kernel 2.4 is *VERY* stable (for me). Its up-time is still longer than any other windows system (I'm not even referring Win9x, since its no competition for Linux or WinNT).

Another fact: as the Linux kernel goes up in
its version, it becomes faster on the same machine. In windows, a new kernel would probably
eat up all you system.

Besides, I know KDE has its crashes from time
to time, but 2.2.1 is pretty nice (and stable,
comparing with earlier versions), and we shouldn't forgert its a freeware (and most of
the people behind KDE contribute, rather than
get any money for their work). I think that if
the resources ever put in Windows/Office would
have put to KDE/Linux, things would probably looked different.

I work both with both Win2000 and KDE on my Debian box. On a PII-300, KDE runs fast and
flawlessly.

Not good at all ? That's not fair to say such thing....

All the best,
Elad.

JEP , i killed *all* windows stuff and installed
ext2 on the whole computer , with kde , and its has so many GRAVE bugs , its nearly not usable.
it does not even startup. ihave to try 2 times beforce it goes up, and so on .
kwin reaaly suxx..

The problems you are describing are *not* bugs in KDE. Problems like that have to do with problems in KDE's installation or the installation of some other program that KDE depends on (X windows, etc). If everyone was experiencing problems like that, it would be a KDE bug. But since almost everyone here runs KDE fine, it is obviously a problem with your installation.

That doesn't mean it's your fault, though! It probably has to do with bad packages from your distribution. What distribution are you using?

by ac (not verified)

Since you bring in the next QT version as a remedy for the current problems:
Is there any release date for the first beta ( stable API, not to many crahes ) on the horizon?

by nap (not verified)

According to this comment posted on 10th April
http://linuxtoday.com/news_story.php3?ltsn=2001-04-09-011-21-PS-0065
it should be coming coming pretty soon

I have a rather simple solution to cut and paste. I'm not sure if it's the same as the Qt-3.0 solution, as that wasn't described in much detail.

Have 2 clipboards. One for when you highlight something, and this would be pasted with the middle click. That would keep people like me who are used to that style and find it more efficient happy. The other clipboard would function more like the traditional Windows clipboard. You highlight, but it isn't added to that clipboard until the user chooses to copy it (using control keys or a button). If a user chooses to copy highlighted text, that text should then be removed from the middle-click clipboard and the old contents of that clipboard be restored. Does that make sense? That would satisfy both worlds, and, in many cases, come in handy for anybody. I just hate when i accidently click and move the mouse, and highlight something else right before i'm about to paste a big document, and have to re-highlight it.

Yes, it seems to be just about the same as the current agreement and Qt3.0 solution.

I have found a bug in KDE. When will it be fixed?

Then fix it dear Henry. fix it.

> > There's a bug in the KDE, dear Liza, dear
> > Liza
>
> Then fix it dear Henry. fix it.

With what shall I fix it, dear Liza, dear Liza?

This is getting silly, now to something completely different.

> > > There's a bug in the KDE, dear Liza, dear
> > > Liza

> > Then fix it dear Henry. fix it.

> With what shall I fix it, dear Liza, dear Liza?

With your brain, dear Liza, dear Liza, dear Liza
- with your brain:

The KDE 2 Development book
http://kde20development.andamooka.org

Have fun! :-)

> > > > There's a bug in the KDE, dear Liza, dear
> > > > Liza

> > > Then fix it dear Henry. fix it.

> > With what shall I fix it, dear Liza, dear Liza?

> With your brain, dear Liza, dear Liza, dear Liza
> - with your brain:

> Have fun! :-)
> Charly

But there's a hole in my mind, dear Charly, dear Charly, there's a hole in my mind.

What shall I do then?

go play with gnome :o)

yOu aRE a fAGGot

right after you report it.

...and sometimes sooner !

it's already fixed.

> it's already fixed.

Ouch!

My cat was "fixed" once.

He didn't like it at all!

What's that bug ?

Ban his IP! Otherwise he will troll at both KDE and Gnome!

by TrollKiller (not verified)

You don't have to be Kreskin to predict trolls' future.
The hand writing is on the wall: trolls faces a bleak future.
In fact there won't be any future at all for trolls because trolls are dying.
Things are looking very bad for trolls.
As many of us are already aware, trolls continues to reduce in numbers.
Red ink flows like a river of blood.

Let's keep to the facts and look at the numbers.

Troll leader ac states that there are 2000 trolls. How many usenet trolls are there?
Let's see. The number of Slashdot trolls versus usenet trolls posts is roughly in ratio of 5 to 1.
Therefore there are about 5000/5 = 1000 usenet trolls post.
Usenet trolls posts are about half of the volume of Slashdot trolls posts.
Therefore there are about 500 usenet trolls.
So there are only 1500 Slashdot trolls, compared to the 3 bilion Internet users.

All major surveys show that trolls have steadily decreased in numbers.
Trolls are very sick and its long term survival prospects are very dim.
If trolls want survive at all it will be among trolling hobbyists, Microsoft zealots,
and people with sick minds.
Trolls continues to decay. Nothing short of a miracle could save it at this point in time.
For all practical purposes, trolls are dead.

As for fixing KDE Usability and design issues, I founded a project with Lee Jordan to identify usability issues and problems in KDE.

We are just in the process of building the KDE Usability Study up, but you can keep tracks of it at http://kdeusability.sourceforge.net/ and on kde-usability at http://lists.kde.org/.

We welcome any help and volunteers, so just get in touch if you are interested.

by Asif Ali Rizwaan (not verified)

I have been yelling at bugs.kde.org for good system and network configuration modules for KControl, by watching and Using sources of DrakX, Yast, LinuxConf and other such configuration utilities. But No body cares! I Think that only these reviews will force KDE team to think about it!

This is what distribuors should care of. KDE is not desktop for Linux it is for Unix. Remember this.

I still agree very much with those who ask for this. Whatever practical, principled or technical obstacles there is to a unification like this, it *must* be done one way or another. The usability benefits are just too great. I'm sure there are ways to get around these problems. For example the shared platform nature of KDE could be "solved" by with a Linux specific extension, or in extreme, simply give up the goal of several platforms if there is no other way about it.

They do care, but they have too many other things to tend to. The KDE team is working very hard, and they will get to everything eventually.

It's not like they're doing something unimportant instead. KOffice, KDevelop, Konqueror, and KDE2.2's new printing framework are all very necessary.

Also, like _many_ have said, right now it is the job of the distributions to make tools like these. Maybe KDE can try to do this later with separate platform specific packages.

So for now, let them hack. And a word of advice: "Yelling" at bugs.kde.org (or anyone anywhere for that matter) gets you nowhere. This is mostly volunteer work. Please, send a single bugwish and then get on with your life. Or go read a HOWTO if you're having trouble configuring your system.

-Justin

I sadly can't agree with you. True, KDE team works hard and stuff that you mention is important, but it's not like everyone works on that.

New printing framework is more or less developed by one guy. I don't know about KDevelop or Konqueror, but I can remember that few months back, before 2.0 release, there was a lot of discussion about KOffice.

The general agreement was that it was important to release a stable KDE2 first, after which a lot of effort would be moved to completing KOffice. That didn't happen. KOffice development community stayed more or less the same and those who did join were newcomers and not old KDE developers.

I do think that good theme support is less important than good configuration utilities, but that's just my opinion and IMHO this is the catch. Everyone works on what is important to him. And that's fine.

But saying that it's distributions job to make these tools (and good packages) probably isn't. You see, there's strong competition between makers of distributions and configuration tools are one of the rare things they can use as a competitive advantage (this is probably the reason, why YAST doesn't have a free license). Therefore they don't have a real reason to work on them.

Plus, they don't see the world through KDE's eyes. Caldera also wants remote configuration and aims to get it with Webmin. RedHat puts its money on GNOME. Basically you only have Mandrake and Suse and even Mandrake is very agnostic these days and the last time I looked at it, its utilities were GTK based.

It doesn't make any sense to leave a known problem open for years, just because you think it should be responsibility of someone else to fix it. Especially when these problems are a big part of user experience.

by Jelmer Feenstra (not verified)

I'm just wondering, why couldn't a completely new distribution based only on KDE and Qt be succesful ? I imagine this distribution having all the tools you are asking for and also having a great system of updating KDE. This distribution would be for the beginner user that wants to get some work done and not spend a lot of time installing / setting up things. What are the cons of such an idea ? Anyone ?

Just a wild thought.

Jelmer

The solution for me is Mandrake minus ANY link to GTK. That means that their own maintenance utilities are dropped. Also, you lose The Gimp. Any other GTK based applications are irrelevant.

Long live KDE.

There's tons of distro like that, mandrake, caldera, SuSE, corel, and more, they use Gtk because RedHat uses Gtk and most of the distros are RedHat except they use KDE instead of GNOME. Qt is great but distro's arent willing to spend the money writing the stuff themselves when they can just use RedHat's stuff and have it be almost as good.

There's tons of distro like that, mandrake, caldera, SuSE, corel, and more, they use Gtk because RedHat uses Gtk and most of the distros are RedHat except they use KDE instead of GNOME. Qt is great but distro's arent willing to spend the money writing the stuff themselves when they can just use RedHat's stuff and have it be almost as good.

by not me (not verified)
by Jelmer Feenstra (not verified)

hehe, I can see the sarcasm here (I think :)

But seriously, I'm just talking about the people generally want to use an easy system, more advanced users would probably be using dists like debian or slackware ofcourse. Why do you think windows became so popular ? KDE itself is rather userfriendly, but I just can't imagine installing this on my dad's system as I know he won't be able to install new programs, set up his dial-up connection, add user profiles for me, my sister and my mother etc etc. Is it really such a bad idea ?

Jelmer

by AC (not verified)

Just becuase people are advanced users dosen't mean that they want a distro that is difficult to use(ie slackware and debian, both suck).

by Jelmer Feenstra (not verified)

Well, I think configurability inevitably comes along with difficulty. I just think the kind of tight integration of the interface and system that windows has clearly helped it becoming so rapidly popular. In order to take over the world (ahem) we need to make the other part of linux user friendly as well, and that is the distribution, configuration of the _system_.

Jelmer

by not me (not verified)

No, I'm really serious! Really!

Go look at www.redmondlinux.org! They are an acutal Linux distribution, and they are actually trying to do what you say!

Really!

by Greg Brubaker (not verified)

I've been using Linux for over a year, and never once have I actualy loaded up Gnome... I just never needed more than KDE. For this last year, I've also had Windows as a safety, and I would say that I only load up Windows 0-1 time a month... (unless it crashes really fast :)

KDE is great, and who cares what some editors had to say about their first use of linux. You wouldn't give a Mac a review on your first use, would you? If you did, you'd be too stuck on "How do I get my floppy back?", and "Where did my other mouse buttons go?"

by Beth McConn Norris (not verified)

Greg,

I apologize if I have the wrong Greg. I am trying to find my brother who was adopted by Brubakers in Ohio over 30 years ago. I met him once while in Middle School and would like to reconnect with him again now that I am a grown woman with my own family. Our Father died in January of '73 when I was 9 months old. I have been in touch with our sister Kelly, but would very much like to find my brother Greg again and talk. Hope this is you, but if not I will keep searching

Love Ya,
Beth

by TrollKiller (not verified)

You don't have to be Kreskin to predict trolls' future.
The hand writing is on the wall: trolls faces a bleak future.
In fact there won't be any future at all for trolls because trolls are dying.
Things are looking very bad for trolls.
As many of us are already aware, trolls continues to reduce in numbers.
Red ink flows like a river of blood.

Let's keep to the facts and look at the numbers.

Troll leader ac states that there are 2000 trolls. How many usenet trolls are there?
Let's see. The number of Slashdot trolls versus usenet trolls posts is roughly in ratio of 5 to 1.
Therefore there are about 5000/5 = 1000 usenet trolls post.
Usenet trolls posts are about half of the volume of Slashdot trolls posts.
Therefore there are about 500 usenet trolls.
So there are only 1500 Slashdot trolls, compared to the 3 bilion Internet users.

All major surveys show that trolls have steadily decreased in numbers.
Trolls are very sick and its long term survival prospects are very dim.
If trolls want survive at all it will be among trolling hobbyists, Microsoft zealots,
and people with sick minds.
Trolls continues to decay. Nothing short of a miracle could save it at this point in time.
For all practical purposes, trolls are dead.