With KDE SC 4.7.3 recently introducing the latest of a continuing pattern of debacled KDEPIM/Kmail releases I think it's crucial for the KDE development community and release authorities to completely, honestly and thoroughly reassess the release model.

Of course, with such a provocative statement, an inevitable flood of supporters will immediately jump the KDEPIM defense, touting all of its vision and glory. I won't pretend to try to understand why, especially in the face of the current broken release situation, unless simply blinded by their allegiance. While I too have held a strong allegiance to KDE for more than a decade (and remain so), riding out the storms and always looking to the grander vision; at least in the case of KDEPIM, one has to call it for what it is. Broken! We can argue over the minutia of what's broken (Kmail, Akanodi, etc.) or what's so amazing about it, but the bottom line is that the 4.7.3 release irreversibly broke my primary means of communication. Period! Again! And... I never saw it coming. After faithfully taking the KDE SC 4.7.3 plunge, once again like many others likely did, I had to spend frustrating hours reverting back to the previous 4.6.5 release and then restoring my email from backup (luckily I do have a fair backup system), although I did loose a few new emails in the process. Others may have simply changed to a different tool as I eventually did to ride out this storm. The 4.4 release was an equally problematic struggle for me and for many and is well documented.

This release methodology is arguably only one of any number of reasons why KDE is, sadly, not taking a decisive lead among Linux desktops and at least partially why nearly all major Linux distributions lead with GNOME and not with KDE. And, as much as instant messaging, web based email and social networking sites (Facebook, Twitter, et al) may appear to be the trendy new replacement communication medium, they are an incredibly long way from completely replacing the good old-fashioned desktop email client as an essential communication tool among the computing masses. Arguing otherwise is wishful thinking at best. I'm as plugged in to the new social communication medium as most people are (well almost), but still rely very heavily on my local email and have no plans for changing it any time in the near future, if only for privacy reasons. On the job, email is easily 70-80% of my critical communication, for a host of additional reasons. Suffice it to say that the desktop email client/PIM is here to stay for a good long time.

That's what makes this latest (and repeated) KDEPIM release failure so disappointing and unacceptable. So much so that, like many others, I've had no choice but to switch to another email client, Thunderbird in my case. Hopefully only until KDEPIM matures sufficiently. Else I was prevented from migrating to KDE SC 4.7.3 at all (KDE SC 4.7.3 + KDEPIM 4.6.5 is not even a viable option). What's even more disappointing though is how much better Thunderbird seems than Kmail (at least in performance and setup). Although it is lacking some nice features from Kmail that I'd grown quite fond of. Even still, it's like having a veil lifted from my eyes, having been so committed to Kmail for so very long. As a long time, loyal KDE user, this is heartbreaking - Time after time, the same overly aggressive release mistakes. Not at all very comforting.

For those interested, here are the main upgrade problems I experienced:

- The migration tool was completely fault intolerant, requiring Google and command-line intervention and still did not successfully migrate my email account logins to kwalletmanager, forcing me to manually re-enter them unnecessarily.
- The migration tool migrated all of my local mail folders to Akonadi (or somewhere), effectively making them irrecoverable after my decision to back out the upgrade, necessitating restoration from backup. Suggest a reliable un-migration/recovery tool.
- Kmail decided to no longer honor my Local Folder sort order and wouldn't allow manual sorting. I have only a moderate number of local folders, but found this highly inconvenient.
- The Kmail/Kontact and Akonodi/Nepomuk relationship are overly confusing, apparently still immature, too loosely integrated and wholly misunderstood by the masses, including myself.
- Akonadi/Nepomuk performance is abysmal, not only impacting Kmail/Kontact performance, sucking CPU, but running constantly and re-indexing on every login. This was the nail in the coffin.

Either KDE is a quality product or it is not. Quality products rarely make these types of release mistakes, at least not repeatedly (as with KDE 4.4), especially in a point release and offering no easy means of recovery or alternative means of upgrade survival or even an upgrade opt-out.

[End of rant]

I sadly dumped KDE PIM after much heart searching round about 4.2 or some such and haven't gone back to it (now on 4.7.4) - the mere fact that Akregator is not able to handle tags is a big a no no for.

> Either KDE is a quality product or it is not.
It is definitely not.

Kded dies in 4.7.3 randomly (Fedora 16). On my notebook, thanks God, it dies after "boot" of KDE is complete, but on desktop it dies when splash screen is active and KDE does not start at all. On notebook KDE is hardly usable and on desktop I switched to XFCE on obvious reason.

It is maintenance release, not beta or RC.

Kmail 2 is just piece of ..... After upgrade I moved to Thunderbird because KMail2 drives me crazy in an hour. Keyboard layout indicator crashes randomly... and so on and so on and so on.

I think that QUALITY is not a priority in KDE development at all. The priority is "bells and whistles" and false pride of authors. If things will move the way they do KDE will become toy for it's authors only.

Hm, how comes that Feodora boasts two unsatisfied customers? Please help me out as I had only one run-in with an rpm based distro - do they patch KDE at all?

It's one thing to complain, but another to help. When you run into a bug, are you filing the bug, and filling it with helpful information? Developers often have no time to read the forum, but they will work with you to get bugs fixed, if the information they need is there.

For more about how to create a useful bug report, see: http://blogs.fsfe.org/myriam/2011/10/wh ... rt-useful/

moved to KDE Café

I'm not alone.
http://www.itwire.com/opinion-and-analy ... se-of-lxde

I was using KDE and the desktop failed to load after an upgrade. At the time it happened, my workload was such that I had no time to fiddle and find out what was wrong. I looked for the easy way out, a DE that would allow me to keep working.

valoriez wrote:It's one thing to complain, but another to help. When you run into a bug, are you filing the bug, and filling it with helpful information? Developers often have no time to read the forum, but they will work with you to get bugs fixed, if the information they need is there.

I don't think anybody here disagrees on the importance of filing bug reports. That's not really the discussion here. Doubtless there are already a number of bug reports filed on KDEPIM 4.7.3 issues. Rather, my rant is really about the repeated releases of alpha quality software to the unsuspecting user community. And... just reading these forums only confirms this as there is an over-abundance of kmail2 frustration posts, just as there were with the equally frustrating release of KDEPIM 4.4.

Trying to draw some focus to the bigger problem.

Exactly. I don't complain on BUGS. The problem is overall QUALITY and total absence of QA work.

With regard to the KDE Daemon (kded4) problem - this application mostly loads others code, without a backtrace it is impossible to track down the issue. It could be caused a bad configuration file, inconsistent installation or be a distribution specific breakage.

I haven't seen it crash on me in a long time (except once when it was trying to load a plugin which it was not compatible with, so that was expected).