Thursday, September 10, 2009

KDE4 - Hype & Reality

KDE3 vs. KDE4

It feels good to see KDE 4.3 mature. Again it feels good to see some of the major but conservative Linux distributions turning up to KDE 4. PCLinuxOS has set a complete KDE 4.3 repository ready for those interested in this new and much-hyped desktop environment. Slackware 13 is loaded with KDE 4.2. However, it feels very bad that KDE4 doesn't deliver the stuff hyped much earlier. Read on to know what I exactly mean by "hyped stuff" but don't start kde4-fanboyism.

The worst hype about KDE4 is that it's more responsive than KDE3. It's a complete false statement. I have tried OpenSuse, PCLinuxOS, Mepis, and sadly, Fedora, on top of KDE4 and 3. Sorry friends, I could not see any major difference in performance. On the contrary I have seen lack of performance while using KDE4. "KDE4 uses around 40% less resource than KDE3" is another biggest hype that moved across various OSS forums. It's again a false statement. True, KDE4 manages memory more intelligently, but not 40% less.

I am sure, KDE 4.3.1 is way more polished and advanced than KDE 3.5.10. It looks stunning. It's enviable to all - KDE 3.5.10, Windows 7 and Mac OS X Snow Leopard. But it is not as tightly integrated.

Good news is that Linux Kernel 2.6.31 has much possibility in terms of desktop performance improvement. And it seems, KDE 4 will stood up to its hype when it couples with this nice kernel. Till then I am sticking to KDE 3.5.10 on PCLinuxOS MiniMe.


Anonymous said...

kde4 runs slower on your hardware(hardware you failed to mention by the way) and "It's a complete false statement [ that kde4 runs faster than kde3 ] "

the "i dont see it on my hardware hence it doest happen on everybody else hardware and its a hype" comment doesnt add up

Anonymous said...

I find KDE4 to be more responsive, but then again, mine is only really a base install with a lot of the optional stuff disabled when compiling from source.

manmath sahu said...

I don't say that KDE4 is always slower than KDE3, but it simply doesn't stand up to its hype that it's faster.

jajaja said...

You said you "sadly" tried Fedora - is it a particularly bad implementation of KDE4?

Anonymous said...

Last night i installed debian lenny (net install) and then installed kde 4.3.1 (from the ustable repository). On my old machine, a pentium IV, 320 Mb RAM and a Geforce2 MX 200, kde4 was a little more responsive (less HD work) than KDE3 on the same machine. After i booted KDE4, it taked only 150Mb of RAM. I can say that KDE4 with compositing disabled is faster that kde3 on my machine. Unfortunately it doesn't look good with the compositing disabled :)

Anonymous said...

Wow another blog generating page hits by bashing KDE4.

I dumped XP for Gnome on Ubuntu, Also I got a Mac at work. I dumped both when KDE4 came out. I could never appreciate the powerful but chaotic and old-looking KDE3.

IMHO KDE4 has the big-$-designer look and feel that previously only the mac had - minus the corporate BS like company logos everywhere or pop-up advertising for new or less crippled versions. Also no registration or activation.

The amount of high quality apps available for KDE4 is amazing and growing very fast indeed. The standard is high and because of the high level of integration the user interface is well structured and seems logical or natural to me.

Instead of looking for magical 40% performance boosts, try doing some actual work on this desktop. It's fun.

Anonymous said...

It seems you have no idea of what you're talking about.

manmath sahu said...

Thanks you dropped by! And Mr. Anonymous I have much ideas of what I talk about. I know KDE4 is better than KDE3 in many aspects. But all I wanted to convey through this post was that - KDE4 is glorified more than what it deserves. The false hype of performance and resource consumption was just two points in this regard.

blackbelt_jones said...

Okay, I have a Dell Optiplex Pentium 2.4 GHZ with a gigabyte of RAM. I installed Kubuntu Jaunty with KDE4.3. I reied to watch a movie on HULU the reception was really jerky, so I booted a SLAX 6.1.2 live CD, with KDE3 and got much improved flashplayer reception.

the "i dont see it on my hardware hence it doesn't happen on everybody else's hardware and its a hype" comment doesnt add up

Actually, it does. I don't remember hearing about how KDE4 ran faster on some hardware. All I kept hearing about was how it ran faster. Well, apparently, it also runs dramatically slower on some hardware.

We were told that it was faster, and some of the time, apparently it is. Not for me, or for manmath, but apparently for anonymous, anonymous, anonymous and anonymous.

We were told that KDE3 is obsolete because of QT3, although software continues to be released with qt3, including opera 10.

We were told that KDE3 would continue to be supported as long as it had users. That's apparently not true either.

Now we're being told if anyone complains, they're working for Microsoft or trying to "generate blog hits." No one could possibly have a problem with having this monster shoved down their throat.

I've been frustrated and angry and anious for the better part of two years, but only now do I begin to suspect that we've have been had.

I tried Kubuntu, Opensuse, Lenny, Sidux. Just because it runs badly on my computer doesn't mean that it runs badly on anyone else's computer. Fortunately, I've got my slax cd archived, along with a bunch of modules, and all the packages from slackware 12.2

It's the crazy stampede to implement that occasionally pisses me off, especially the headlong rush to toss a prexisting classic down the crapper, but I still support KDE4 as the proper direction for development. I'm not saying that it's a piece of shit. It's emphatically NOT a piece of shit... but I have come to believe that if it was a piece of shit, they would still be telling us how great it is.

manmath sahu said...

Hi Jajaja,

Yes, I found KDE4 is very badly implemented on Fedora in comparison to OpenSuse and Mandriva.

Hi BlackBelt,

It's good if it runs well on others' systems. BTW, this post is not another flamebait. It's just a personal opinion that KDE4 has been hyped beyond what it really is. You may like it or not. So, take it lightly.

blackbelt_jones said...

Okay, I tried a google search for "KDE4 uses less resources than KDE3". Here's what appears to have happened. In late 2007, someone did a survey that concluded that KDE 4.0 used 40 per cent less resources than KDE3. But that turned out to be incorrect. Someone checked the math, and it used 20 per cent more resources than KDE3, but the original uncorrected story got was more play on the web, and gets repeated to this day.

To their credit, KDE developers were instrumental in the correction, but Searching through four or five pages of results for "KDE4 uses less resources than KDE3" produced no references to any other survey or study.

These two links are telling. Note the big KDE4 uses less resources headline at the top, and the correction that come much further.

So what we see here most definitely looks like to me, although it's important to note that the KDE organization is not directly culpable in spreading misinformation, but it became part of the fanboy dogma, so that you were attacked, accused of disingenuous motives (generating hits) for questioning what appears to have no basis in fact. I invite someone else to try this google search, and see if they come to a different conclusion.

I think that sooner or later, we're going to have to face up to the fact that this is some kind of debacle, and yes, the comparisons to Vista are beginning to look valid. I believe that, to some extent, hype and misinformation have entered into the free software method, along with a cultlike dogmatism that corresponds to the corporate groupthink that probably drove Vista. It's not right for urban legends to be guiding software implementation. It's not right for everybody to be saying it's faster when it isn't. We've got to be more cooly critical in the future.

But less brutal! The haters are as much to blame as the fanboys. The ferocity of the initial attacks could not have produced any other result than a further digging in. Has it EVER produced any other result?

I'm not going to be using KDE4 any time soon, but I've seen it run on more high end computers and it can pretty dazzling. So let's just say there's room for disagreement and leave it at that, but if software is to be developed efficiently on the internet with the participation of a community, we're going to have to guard against those perennial internet problems, misinformation and net rage.

manmath sahu said...

thanks blackbelt for such well thought reply. hope the fanboyism will diminish as the OSS matures.

blackbelt_jones said...

I used to be one of the KDE4 haters, so I have my own share of responsibility for the debacle. Now I'm a KDE4 critic, of course, if you're a KDE4 fanboy, you can't tell the difference.

Anonymous said...

KDE4 is faster on my hardware, a LOT faster.

I use the open source radeon driver for my ATI graphics card. According to benchmarks, the open source driver is much faster in 2D graphics operations than the closed source driver:

The nVida closed source binary driver is also abysmal for 2D desktop-type graphics operations.

But whatever the reason, KDE4 is much faster than either KDE3 or GNOME on my system.

Even if my system was cripled by an abysmal 2D graphics card and driver, then KDE4 would be no slower than KDE3.

Anonymous said...

Genial post and this mail helped me alot in my college assignement. Thanks you seeking your information.

How about this