I have one thing to say for all your cries.
TRY KDE 4.2 >:)
My experience of KDE 4 on even old hardware did not suggest something klunky and sluggish, actually the opposite.
The KDE4/QT4 combo seems to reduce memory requirements and give a more modern look.
How many of those complaining about KDE4 speed, are using Nvidia binary drivers, which are known to peform poorly with features KDE4 is using?
Have you got links that back this statement up?
Likewise with Dolphin. I’m hoping they fix this in later releases, but initially, trying to revert to Konqueror in KDE 4 caused faults.
Filed a bug to 4.0, and the answer was that Konqi didn’t lose file manager functionality, and that same module was useable by both, but it was crashing in that version due to issue with Konqi integration.
I blew off Gnome years ago because the Gnome developers caught this same, “make it stupidly-simple” attitude and refused to even listen to complaints
They have to develop to a HIG, and these get designed by corparate usebility experts. It would appear a large number of ppl, cannot focus in on relevant items, and get disturbed by choices. You probably like me, focus and don’t even see irrelevant items, which can lead to embaressment on occassion, where I miss an ‘obvious’ option, that I’ve never needed.
If I have to, unless KDE 4 comes back with ALL of the functionality and reconfigurability that I need, I’ll be one of those who pitches in to keep KDE 3.5 alive.
Scrambled spagetti code doesn’t scare me. I assure you, I’ve worked with worse.
Supporting KDE 3.5 in 3 years time, is likely the reason to not officially include KDE 3, rather difficulties anticipated next autumn and year.
But if KDE 4.3 doesn’t satisfy meet my needs, then I’ll choose an alternative environment.
As a developer, think how you realistically implement change.
For maintainability reasons, you decide to redevelop, when the library you depend on, goes through a major change. Then, you look at what is possible with the newer library, otherwise you could just use the tried and trusted version in maintenance mode.
So you do new things with canvasses, support compositing, and you don’t test these by eye candy demos, which show off new features?
What happens, when you reimplement and clone the old desktop, without new features? How would that be received?
I’ve been using KDE since version 2.0 because I like it better than Gnome.
KDE 1 myself, it was rock solid, most reliable desktop I’d ever used. Konqueror’s handy rendering of web pages, and the ioslaves was a killer feature.
GNOME 1 used to crash a fair bit, even in RedHat 6.2 (their stablest release). GNOME 2 was the corporate take, after Sun came on board, and the useability experts got the HIG in, which we find so restrictive and annoyingly “dumbed down”.
But if KDE 4.3 doesn’t satisfy meet my needs, then I’ll choose an alternative environment.
That’s what I’m thinking. I’ll certainly give it a try. But I do wonder what alternative I would choose?
As a developer, think how you realistically implement change …
Actually, I do. In fact, I’ll bare my soul for a moment, and if you can reassure me about this, then I will happily listen. I apologize in advance for the length of this.
I love Open Source. I choose to use it for any number of reasons, but one important one is open standards. The data on my computer is MINE. I should know how it’s stored, and if I want to write a program to access it, I should be able to. That’s just one of many reasons.
But to date, the penetration of the F/OSS desktop into businesses, much less private homes, is dismal. GNU/Linux and the BSDs are doing much better in servers, but Microsoft is slowly gaining on them there.
The typical F/OSS developer will object, “I’m not competing with Microsoft.” Ummm … yes you are. Whether you realize it or not. Why? BECAUSE THEY CERTAINLY SEE YOU AS A THREAT AND ARE DETERMINED TO DESTROY YOU. When the day comes that Linux is only on 1-2% of desktops again, hardware drivers (already a problem) will disappear, Linux will become a marginalized, minor “fanboy” hobby and that’ll be that.
Simply put, I think F/OSS is in danger. Further, I think there are reasons for this. If you ask why more vendors haven’t ported their applications to Linux, one thing you’ll hear over and over is that “it’s a moving target.” They’ll build up a package, release it … and then watch as it shatters as soon as the kernel is updated or the KDE people decide to move to a completely new library.
That costs them money for support. People will call up, angry. “I upgraded from Opensuse 10.3 to 11.1, and your crappy software doesn’t work anymore!!!” Easier just to say either that (a), we don’t do a Linux version, or (b), we do, but it’s ONLY supported on a very limited number of distros and versions.
Now: in the following, I’m speaking IN GENERAL. Of course there are exceptions, but IN GENERAL:
You speak of developers and how they target the “new library.” Sigh. My friend, my development background is in Windows. There, we wouldn’t DREAM of jumping onto the latest, greatest library just because it had new and cool features. In fact, the standard there was to deliberately develop to the previous release (or even TWO releases back) to ensure that our stuff would work on as many platforms as possible. Ex.: Even after Windows XP came out, it was common for commercial houses to target Win98, so as to work on as many desktops as possible.
That type of thinking is utterly foreign to most F/OSS developers. They want to be bleeding edge! (This same desire, where their coding fingers outrun their doc’ing fingers, explains why so many F/OSS packages are released with dismal or non-existent documentation.)
A new library comes out and the developers go, “wow, cool, let’s USE IT!” If the library has new eye candy and “gee whiz” features, of COURSE they target those immediately. After all, that’s more fun, and they’re not being paid for it anyway, right? 
So … in the case of KDE 4, we get Windows that can be rotated 360 degrees. Cool! But what good is it for MOST USERS? The time spent to develop that feature might have much more wisely been spent on other things.
The fact that this leaves carnage in their wake, as dozens of things get busted, dozens of new bugs appear and users everywhere are screaming “wuh happ’n??” is ignored. PROGRESS! We’re the BLEEDING EDGE!
This is a real problem. And yeah, this is why I reacted so strongly to the claim, “you’re just opposed to progress!” Not so. But I’ll finish with a true story (and on final apology for the length of this).
The guy who turned me on to Linux back around 1999 was a true-blue gunslingin’ hacker who started out compiling Slack from source in the old days. He went from Red Hat to Mandrake (now Mandriva), which is what he recommended to me. He could certainly have continued hacking and building from source, but (like me), he ENJOYED being able to just install, click and run.
In recent years, from his point of view, Linux has really become infected with this, “let’s release the latest and greatest thing, fix bugs on the fly, beg others to do the documentation (because WE don’t have TIME) and be BLEEDING EDGE!”
My friend moved back to Windows XP. He still uses Linux for servers, but says that he’s totally uninterested now in Linux as a desktop, for the reasons enumerated (badly and disjointedly) above.
Yeah, that’s just one person. But if you knew this guy, you’d know why I was DEEPLY concerned. If we lose someone like Ed, who’s a hacker’s hacker and a geek’s propeller head, then F/OSS is in trouble.
Further, he might just be an indication of why the penetration percentages for desktop Linux continue to stagnate. For every new user who jumps in and tries it, there’s an Ed who got tired of it (and got tired of trying to explain his concerns to people who would blow him off, or call him – HIM! – anti-progress) … … and quietly said, “screw it,” and switched back to Windows.
Edited to add one other point: look at Apache. They still release security patches for the old, ancient 1.3 release. They still maintain the 2.0 release, even though 2.2 is the current version. They have the right attitude. Progress is cool, but it should be done in steps, incrementally, with maturity and eye toward EXISTING USERS. I think what concerns me about the KDE developers, more than anything else, is their decision to just stop maintaining the KDE 3 branch. I think they’re going to regret that. Wait and see.
What I found appealing about Suse 11.0 and KDE4 was aesthetics. A modern and attractive desktop is important to me as I use my computer quite a bit. 3.5 is okay, but it is dated, clunky and to me it is slower than KDE4.
I had KDE4.1 running great for a few months. The updates all worked and I usually would run an update once or twice a month because I found they would post part of an update and I would get dependency errors.
Recently, KDE4.2 RC1 came up on an update but gave me a long list of dependencies. I skipped it. A week later I tried again and the list of errors were shorter, but still there. I skipped again. Last night, the list of errors was even shorter and I had the option of just not installing the particular module that had dependency problems.
Despite my past experience with this I installed anyway and now I have a borked system. I am in KDE 4.2 RC1 now, but I have no plasma desktop. I installed Cairo-Dock and it comes up so I can open and run programs.
I’ve tried zypper and Yast and I still get a Plasma Segfault when I restart. I’ve now downloaded 11.1 and I am going to go a fresh install and see what happens. My //home is on a separate partition.
Lets see if 11.1 is an improvement. I’ll install KDE3.5 as a backup but I much prefer KDE4 or Gnome.
Its a common problem.
Just go to /home/*username/.kde4/share/config/ and rename plasmarc and plasma-appletsrc to plasmarc.old and plasma-appletsrc.old respectively, logout, login back.
Or if you can get a terminal run these:
mv ~/kde4/share/config/plasmarc ~/kde4/share/config/plasmarc.old
mv ~/kde4/share/config/plasma-appletsrc ~/kde4/share/config/plasma-appletsrc.old
KDE 4 was been good for one thing, at least for me. It made me use Gnome for the first time and I discovered I really like it. I doubt if I would go back to KDE now. So thank you KDE developers. 
So … you’re admitting that the KDE project essentially released alpha or beta-quality code and called it 4.0 and 4.1?
That they’re finally getting it right on the third “point” release?
The question that pops into my mind is, why should I believe them? Or you, for that matter?
I’m not being deliberately contentious, and I apologize if I sound that way, but in a way, you’ve just helped make my point. 
I, actually, happened to like KDE 4.2 (RC1)
The more I use it, the more I love it.
Yea, it was a bit odd in the beginning, but I held on 
Everyone knows and admits 4.0 was a pre-alpha quality release. But the KDE devs got back the trust with 4.1. Massive improvements over 4.0. And now 4.2 is the best so far.
Dont believe anyone 
Try it yourself and see if it meets your requirements. Just try.
You’re not alone. Just sit back and read the comments on dot.kde.org (especially on this news). So much hatred towards KDE 4 and KDE devs.
Not without a reason though.
If Microsoft or Apple released software even remotely as unready as a .0 release as the KDE guys did, you all would be ripping them apart and mocking how unfinished buggy poop they put out, how they’re ripping your freedoms and how Linux is the saviour of the universe.
But when KDE does the same, they’re merely trying to get more people to test the software and how the future versions will land from the heavens on a silvery chariot and take you to the promised land.
Aaron Sieger was very keen to have, wider testing and get a release out of the door. They needed applications to port to KDE 4 library, which was now well enough developed.
The problem was that Plasma, was still fairly new, with few months of development, and just wasn’t going to be ready in time, to offer a respectable Desktop. To some extent there is a chicken & egg. You need the window manager and library, before you can create the look & feel aspects, and things like the panels. Also to be useable there’s a need for applications. They only really got it “off by 1”, 4.0 should have been 3.9.666 (Developer), 4.1 ought to have been 4.0 for early adopters.
What happened was, very many ppl were encouraged to download and try 4.0 beta. Found it obviously broken, and riddled with bugs that any developer eating his own dog food would run into. That leads to rancour, and a huge disappointment for many developers who have good standards.
But issues with drivers like Nvidia’s are getting addressed, and that likely would not have happened without a significant user base and drive towards the new desktop. So actually you don’t need to believe KDE Developers or anyone else, just judge the 4.3 release in OS 11.2, when it comes out. That’s the time to see if they delivered or not.
That’s very important. For me the sheer quality and useful features incorporate into FOSS server applications was also important.
Think Apache, improved shells, languages like perl & python.
But to date, the penetration of the F/OSS desktop into businesses, much less private homes, is dismal. GNU/Linux and the BSDs are doing much better in servers, but Microsoft is slowly gaining on them there.
This is absolutely not true!! Linux has a huge penetration in server space, there’s figures around, that tend to show Linux competing against UNIX, with Windows competing against “other”. Latest figures for example show M$ losing market share again now below 33%, with yet another expensive Worm outbreak to bring some more holdouts to their senses.
When the day comes that Linux is only on 1-2% of desktops again, hardware drivers (already a problem) will disappear, Linux will become a marginalized, minor “fanboy” hobby and that’ll be that.
Linux is likely only about 1% desktop according to figures I see. Furthermore it’s a hard market to sell binary proprietary software into.
The Netbook sector is first GP consumer PC, widely sold with Linux pre-installed, and Asus have made a big success of it.
Linux runs on a whole host of embedded platforms, and Point of Sale, where you don’t even see that Windows CE has been remorselessly crushed into the 1% lunatic fringe of the market.
Simply put, I think F/OSS is in danger. Further, I think there are reasons for this. If you ask why more vendors haven’t ported their applications to Linux, one thing you’ll hear over and over is that “it’s a moving target.” They’ll build up a package, release it … and then watch as it shatters as soon as the kernel is updated or the KDE people decide to move to a completely new library.
Lack of sales is the main reason. The kernel system call interface actually doesn’t change break compat, unless they abuse GPL with legally questionable attempts to create proprietary modules loaded by the kernel.
That costs them money for support. People will call up, angry. “I upgraded from Opensuse 10.3 to 11.1, and your crappy software doesn’t work anymore!!!” Easier just to say either that (a), we don’t do a Linux version, or (b), we do, but it’s ONLY supported on a very limited number of distros and versions.
That’s what QA ppl have always done. Specify supported configuration hardware and software, and then do certification.
You speak of developers and how they target the “new library.” Sigh. My friend, my development background is in Windows. There, we wouldn’t DREAM of jumping onto the latest, greatest library just because it had new and cool features.
KDE applications are a grey area. KDE 3 & 4 applications should run, under GNOME, or other KDE version, but it is better in a DE to run native.
Qt4 was I think an upheaval, KDE 4 didn’t immediately follow, because there were significant changes and challenges to solve.
In fact, the standard there was to deliberately develop to the previous release (or even TWO releases back) to ensure that our stuff would work on as many platforms as possible. Ex.: Even after Windows XP came out, it was common for commercial houses to target Win98, so as to work on as many desktops as possible.
Mostly if you link against the oldest library version, then binaries will run.
Where I have built rpm’s myself, I could generally install them on newer OS releases without problems, except where due to standardisation efforts the names for directories like /etc/init.d were changed (1 off for FHS and LSB compliance).
That type of thinking is utterly foreign to most F/OSS developers. They want to be bleeding edge! (This same desire, where their coding fingers outrun their doc’ing fingers, explains why so many F/OSS packages are released with dismal or non-existent documentation.)
FOSS thinking is like this. You develop the source on a platform, and list the base requirements. Your deliverable is a compilable source package, which is installable via configure, make, make install. Then you let someone else package it, in another format, and that puts config, and any support scripts in distro specific locations. Often it is convenient to host that contribution on your site, but it’s secondary to the source tarball.
A new library comes out and the developers go, “wow, cool, let’s USE IT!” If the library has new eye candy and “gee whiz” features, of COURSE they target those immediately. After all, that’s more fun, and they’re not being paid for it anyway, right?
Not my experience. Generally support for newly developed libraries is a config option, and not a requirement.
So … in the case of KDE 4, we get Windows that can be rotated 360 degrees. Cool! But what good is it for MOST USERS? The time spent to develop that feature might have much more wisely been spent on other things.
KDE 4, mixes infrastructure enabling features, with a finished environment with policy defined. Rotatable windows may be used by UI designers in future, in ways that you and I cannot anticipate.
Furthermore, KDE did not specify much of Qt feature set. Troll Tech (now Nokia) obviously care about KDE, but it’s rare (Phonon is an exception) that KDE specs, define Qt features. So now… You have a new library Qt4, and need to use it, in future for support reasons, and technical advantages that may advance state of the art.
When re-developing, do you conduct experiments, and trials of “risky” new features early on, or do you just “stick” to low risk things that are known to work?
The fact that this leaves carnage in their wake, as dozens of things get busted, dozens of new bugs appear and users everywhere are screaming “wuh happ’n??” is ignored. PROGRESS! We’re the BLEEDING EDGE!
Frequently it’s distro’s like Fedora and openSUSE who are pushing forward with bleeding edge, to keep the environment improving as fast as possible. Installing Debian “Lenny”, and note it is not using libata/pata_* drivers, it offers you KDE 3.5 not 5. There’s by default a 2.6.26 kernel, not 2.6.28 despite it being pre-release.
In recent years, from his point of view, Linux has really become infected with this, “let’s release the latest and greatest thing, fix bugs on the fly, beg others to do the documentation (because WE don’t have TIME) and be BLEEDING EDGE!”
What happens when a distro like openSUSE puts out “stale” package versions? How many ppl rushed to install 11.1, despite having 11.0 on their system? I’ve called it “Version-itis” for a long time, it’s always been there, and there’s less and less reason for it now, the core is much more mature.
Frankly I’m disappointed in many Documentation efforts. Web HOWTO’s that are distro specific, not integrated with upstream release, and lack of effort on docs.#
But Commercial software is worse! Manuals when provided are useless, they just explain the obvious, for example listing menu options, without detailed explanations. Furthermore they are generally not read, and the trend is to abandon them, less money is spent on them, so they are even less worth reading, making a vicious circle.
For library code, and things like ALSA, the project should not be releasable, without proper documentation, and it should be developed as they go, something like perl handles it.
Shame, I find the FOSS desktop works very well for my needs, but I do run some consumer stuff that is just better supported under Windows, with Mac/Linux development an after thought.
If we lose someone like Ed, who’s a hacker’s hacker and a geek’s propeller head, then F/OSS is in trouble.
Windows is a major PITA in my book, XP is like a toy IMO, and is very hard to keep secure, and there’s poor information on driver updates. The UI is poor, KDE 1 had superior functionality. Obviously MS have done a great job with Vista.
Further, he might just be an indication of why the penetration percentages for desktop Linux continue to stagnate.
Stagnation? The figures I saw was showing growth, but not as fast as Mac OS.
Edited to add one other point: look at Apache. They still release security patches for the old, ancient 1.3 release. They still maintain the 2.0 release, even though 2.2 is the current version. They have the right attitude. Progress is cool, but it should be done in steps, incrementally, with maturity and eye toward EXISTING USERS. I think what concerns me about the KDE developers, more than anything else, is their decision to just stop maintaining the KDE 3 branch. I think they’re going to regret that. Wait and see.
They did stop maintaining it? Since KDE 4, I’ve seen 3.5.9 and 3.5.10 released to fix bugs. 3.5 is better now than any past release.
Now explain to me why, with the source code, and given an annuity to spend on FOSS, that you could not arrange for KDE 3.5 to have security fixes? If KDE developers are so awful, why is it that noone has forked the project off and made a KDE3.org? Given that, if it’s important enough, it will be done.
Apache is in server space, and may have been used in “devices” like routers. It is simpler, to support, than a desktop used by the masses.
Furthermore this “fail” project, despite all the controversy won this award for example : KDE Voted Free Software Project of the Year
Linux Format magazine has unveiled its annual Reader Awards (PDF) for 2008 and KDE won a ‘landslide’ victory in the category of Free Software Project of the year in recognition of the ‘incredible’ work done with KDE 4. Amarok, Qt, Konqueror and the KDE-based Asus Eee PC were also recognised in the awards.
Please, try to have a sense of proportion about this. KDE 4 is not the only UI in the distro, we can work in degraded mode with GNOME or an alternative.
… and you have no problem with this? KDE lost a lot of respect and trust when they released 4.0. That software should have clearly been called a “beta.” More to the point, they should have recommended that production systems continue to use KDE 3 until 4 was truly ready.
PC vendors who might consider pre-installing Linux at the factory get badly burned by this, too (because they get a flood of support calls). And the F/OSS community wonders why preloaded Linux hasn’t become more common? This type of thing is one big reason why it hasn’t.
They did say it was a “Developer” version, in the hoop-la over 4.0 that message was predictably lost.
There was internal controversy, but that’s in the past, all those numbers are just labels.
What matters is that :
4.1.x - currently is usable as a desktop for many
4.2 - will satisfy a higher proportion of KDE userbase
4.3 - offer yet more improvements.
Did you know that Qt3 is not supported any more? Far from rushing to Qt4, KDE are being pushed.
Why did the distributions kill off KDE 3 so early?
There’s also a wise comment, talking about “loss of our control” Why did the distributions kill off KDE 3 so early? being the reason why we feel so threatened by disappointing Desktop releases.
“For me, this whole process has been a lot more emotionally harrowing than it really ought to be. And for others too, clearly, this just hits us where we live, and for me, it really brings out the crazy. We’re Linux users because we want to have control over our system, and when we feel like we’re losing control, it’s a big deal, even if what we’re really faced with is nothing more intrusive than the possiblility of a couple of extra clicks from time to time at worst… and at best, the best destop GUI ever! To anyone who has been reading my posts, there’s no point in attempting to conceal the fact that this whole process has occasionally turned me into a whiny six year old boy… or maybe a resentful, annoying old lady. Sorry about that.”
When I first tried the Beta of KDE 4.0, I felt threatened and concerned, having fears about DE direction. Answers in bug reports, were reassuring. Looking back, I know that what I could perceive through the outer face of the software under test, was not really enough information to make informed opinions.
IMO the very last thing FOSS world needs, is yet another DE or WM project. It seems to be more attractive than doing applications.
Personally I’d like to see decoupled application engines interfacing with the actual UI, so the core runs unchanged. Each native UI driven via shared library code for each major desktop. Having “best of breed” applications matters much more than, whether we have plugin features to krunner, or the exact menu style, and what Theme of the Day is fashionable.
Well, this super awesome KDE 4.2 made me completely forget my frustrations with 4.0
Totally agreed there.
Let us all stop bashing KDE4 and help it gain the lost trust. Let us all stand together for the future of one of the greatest open source projects ever, KDE.
Well said, Hear, hear!!
The only reason I can see for all this mess is the missing “alpha” or “beta” in front of “KDE 4.0”.