Please do remember that individual people use their computers differently. I was not shooting the messenger, I was simply offering solutions and requesting information so that a solution can be found to the problems you listed.

1) Good to see this bug was actually fixed.
2) Has this bug been reported to KDE Bugzilla? If it has not been, then it is not likely a developer will be aware of it and fix it.
3) Did you install the patches through YaST? Non-system connections should not require root access.
4) Are you using KDE PIM applications? If so, then Akonadi will be started, and you will get those warnings
5) Did a fresh install include a completely fresh home directory? If not, then your old configurations could still cause Akonadi and Nepomuk issues.

I hope this comment can be a positive contribution. cmrntnnr and wotan have hit the nail on the head. I've used KDE for five years. I liked KDE 3. I accepted the initial release of KDE 4 as I was committed to supporting KDE. But KDE 4.6 is still too buggy. When I discovered that kmail was frequently not sending messages, that was the last straw. I'm typing this on a brand new Xubuntu install.

There may be some value to Akonadi, Neopomuk, etc, but from a user perspective,all they do is cause endless problems. The developers' responses to user complaints are condescending at best and frequently insulting. I'm not a noob, but my home computer is a tool. It must just work and Akonadi, Neopomuk etc don't. My day job is as a systems level programmer. I have zero desire to debug bad code at night when what I really want to do is check my email.

Akonadi, Neopomuk, etc are not the only problems I have with KDE 4, but they are the ones that made me switch xfce. cmrntnnr raises a number of very good points. The two which had the biggest impact on my decision to leave KDE are the developers' attitude and the poor quality of the product. Unless these both improve tremendously, KDE will never acquire a big marketshare and may loose much of the marketshare it has. I for one am gone and unlikely to return. Which makes me sad. Best wishes to those who remain.

Can you please provide links to examples of the developers "attitude"?
It would also be nice if you could explain the problems which you were having.

It could be quite possible that you are using KDE in ways which the developers did not think of - and thus did not test and ensure worked properly.

I did a bog standard install of Kubuntu and upgraded several times. I reinstalled several times after KDE 4 was released. I spent countless hours trying to get akonadi and nepomuk to work after each install or upgrade. The only thing even slightly nonstandard is that /boot, /, and /home are in separate partitions. With the upgrade to 11.10 nepomuk never once ran. In searching for relevant bugs I found this, just to pick one:

https://bugs.kde.org/show_bug.cgi?id=258171#c14

Not my bug but the attitude comes through. The core problem though is continuing to preach the virtues akonadi and nepomuk when they don't work for release after release after release. And then abusing users who are trying to find a way to continue using KDE in spite of the developers' efforts to drive them away.

Look for bugs related to akonadi and nepomuk. It doesn't matter how right the developers may be technically, the overwhelmingly negative user experience means that overall the developers are wrong and need to rethink what they are doing. For sure my upper management would see it that way.

Here's the problem that ended it for me.

viewtopic.php?f=20&t=97153

This was just the last straw among many. It's simply not acceptable for a released product to fail so utterly in so many ways in every release.

Well, in KDE's defense it is a free product. And IIRC it has no warranty at all (like most FOSS software). If it doesn't do what it claims... well, tough luck.

That said, KDE 4 has many, many regressions relative to KDE 3, and many of those have not been fixed (or are intentionally left unfixed, e.g. the all-important RTF export support in KOffice 2).

Honestly though this is nothing new. I see the same thing all over the GNU/Linux world. Gnome 3 is markedly inferior to Gnome 2 for most uses, E17 is a total mess compared to E16, and of course polkit/consolekit (and now systemd) has scads of bugs, misfeatures, and missing features compared to HAL. Everyone wants to reinvent the wheel for some obscure reason, and nobody seems to be able to gather enough naysayers together to make a difference.

I honestly don't expect the desktop Linux ecosystem (such as it is) to survive much longer. I'm kind of hoping Haiku matures fast enough take up the reigns as the major alternative desktop OS, because the past few years has seen Linux squandering most of its potential on eyecandy.

@dsurber: Please do understand that the developers do use the software they create - and annoyances such as the ones you describe do get fixed when the affect the developers.

From the bug I can see that the developer / bug triager in question has become extremely irritated with users not accepting the opinion that the "bug" is actually intended behaviour - because Nepomuk is a central dependency of Akonadi as it is used for search.

From what I can recall, fixes and automatic repair code have been included in Akonadi with KDE 4.8 to fix the mail dispatcher if it gets itself into trouble.

@Gullible Jones: Can you provide some examples of components of the main KDE distribution (workspace, edu, games, network, etc) where this is the case?

This thread has been underway now for nearly 3 months. It is worthwhile to return to the original question, "Why doesn't KDE acquire a big marketshare, ~15%". Bart, your goal was to get ideas for blog articles. The discussion has definitely accomplished that. Just to summarize what has been debated or proposed as factors limiting kde from wider acceptance, there are three sizable buckets:

* The number of options and features to control kde behavior is too large, confuses users, and slows the learning process
* Graphics performance of X-window is not sufficient. A side-effect is that kde is not a platform for gaming
* Insufficient quality for releases and lack of consistency between them gives creates a poor impression.

The greatest amount of energy seems to be around the last item.

Something additional that is worthy of discussion in your blog could be the goals of the kde developers and designers. None of the comments in this thread seem to be authored by a kde developer. If they had been, this topic would likely have been discussed. After all, kde is free software. Such an extensive ui project would not have been possible without the combined efforts and good will of a large team. Even with its quirks and bugs, it makes computing more fun. It is important to keep that in mind.

Given that this is free software, do the kde developers and designers care about the opinions of users? Said differently, do users have a voice in or any influence over the development of kde? Are there any standards for releases other than getting a new feature in and meeting a ship date. Would there still be a large team of contributing developers if there were standards? If the answers are no or not really, then we have the real answer to the question in the headline of this thread.

bcooksley wrote:@Gullible Jones: Can you provide some examples of components of the main KDE distribution (workspace, edu, games, network, etc) where this is the case?

Sure. Off the top of my head:

- KDE desktop is extremely slow to start, even on relatively fast computers. KDE 3 started faster on a 400 MHz Pentium II than KDE 4 does on a 1.6 GHz dual-core Intel Atom.
- Memory leaks in Nepomuk/Akonadi, which have been discussed elsewhere.
- Serious breakage in KMail, which has been discussed elsewhere.
- KOffice no longer supports RTF. I know this is intentional; I still consider it a regression, as RTF is very commonly used and KOffice 1 supported it.
- Phonon-xine and phonon-gstreamer seem to have bugs that sometimes make media players crash on exit.
- In general, KDE 4 performance is terrible on older machines, and KDE 4 does not offer a whole lot more functionality than KDE 3 (and offers less in the case of KOffice).

Phonon-Xine is no longer recommended as Xine itself is effectively unmaintained.
For Phonon-GStreamer, have you checked to see if these bugs have been reported on bugs.kde.org?

I myself use Phonon-GStreamer on my openSUSE 12.1 system and it operates fine. It could be a buggy decoder or similar component in GStreamer itself which is causing this crash (without seeing a full backtrace it is impossible to tell unfortunately)

Akonadi/Nepomuk have had further bug fixes in KDE 4.8 - which I believe continue to work towards fixing these issues. Nepomuk itself (excluding Strigi) seems to work fairly well on my system.

I find your observations about startup time interesting, as it usually takes a maximum of ~20 seconds to login on my system with cold caches.

Investigation into what is consuming disk or CPU time during the login process would be quite welcome.

Maybe this is part of the reason (from thread KDE doesn't remember window position.. • KDE Community Forums):

schnelle wrote:I asked Martin (main developer of kwin) on IRC about this feature and he responded:

"In general we ignore feature requests.
We have our own roadmap which does not align with user wishes.
If a user wishes for something like that he has to contribute.
Feel free to provide them this link http://blog.martin-graesslin.com/blog/bugfixing-service "

Is the KDE developer community too busy / too stretched with the visionary roadmap to attend to user wishes, like basic, natural feature requests and stability and performance across the wide user base?

I could understand that to some extent, since different users' desires could be conflicting. That said, it sounds rather out of touch!

Edit:OTOH this is a free and open source project, so the developers might as well be doing the whole thing for themselves, if they so wish.

As for me... I've switched over completely to LXDE, and I don't think I'll be looking back. 'Twas nice posting here!

Do we / does the world need a project on top of KDE that focuses on bug fixes and usability fixes and solutions for current real-world needs, while the actual KDE community focuses on its plasma workspaces vision? The way Mint done for Ubuntu and arguably Ubuntu originally set out to do for Debian?

Does Chakra have this aim? How good is Linux Mint's KDE work?

Good point. I mean, who wouldn't like a multi-million Dollar/Euro/Rand/Pound/Baht/Rupie/add_your_currency input? If we could only get the same attention as the English Premiership. I still wonder why Mark Shuttleworth opted for gnome but who is perfect