Version: unspecified (using KDE 4.7.4) OS: Linux I archive my mail on an IMAP server and access it with Kmail from my workstation and my notebook, both of which run Kubuntu 11.10 / KDE 4.7.4 from the Kubuntu PPA. Today I was working on my notebook, moving folders inside the IMAP archive. Everything seemed to go well: a new folder created, moved a couple of emails and folders into it. But when I accessed the IMAP server from the desktop, nothing was there. I logged on with SSH to the server running IMAP and indeed the maildir folder structure is still the old one and there is no trace of the couple of emails that I moved. It turns out that the IMAP server was not running. I start the IMAP server and the desktop can access the old folder structure and the old emails. Then I turn on the notebook and start Kmail. A large number of messages flash so quick that I can not see them. I find a few of them in the notification tray. Many of them read "Failed to delete the folder, restoring folder list." One read: "Append failed, server replied A000014 NO Must create mailbox before append [ TRYCREATE ]" And then a last ominous one: "Item query returned empty result set" Kmail on the notebook restored the old folder structure, and the folders are *empty*! On the workstation they were gone as well, and when I logged into the server with SSH, they were gone too. Did Kmail destroy one year worth of email archives? If not, where can I hope to find / restore them? Reproducible: Didn't try Steps to Reproduce: * Stop the IMAP server with whom Kmail2/Akonadi are communicating. * Move around emails and folders in Kmail2 * Shut down the client * Restart the IMAP server * Restart Kmail2 * Watch Actual Results: emails went missing in action Expected Results: either warn that you can't do the operations; or try to replay them; but keep a backup of the concerned messages until there is confirmation that the operation succeeded; and make such backup easily accessible.
This is easily reproduced, e.g. on a laptop connected to the internet via wireless. disconnect the access point from the network, move some messages around. no apparent feedback that something went wrong, so assume everything went well. go to a different location / access point, connect, and bye bye the messages were lost in translation. even when messages do move from one IMAP server to another, they are tainted so that signatures no longer verify properly. absolutely unacceptable. moved to Thunderbird. Whoever decided to introduce Kmail 2 made a big mistake. Kmail 1 was working just fine and could have been left alone while Kmail 2 evolved. Instead, Kmail 2 has been released too quick and now, as far as I am concerned, it is Krap. I moved to Thunderbird and will not come back. The few months of Kmail 2 usage left a scar in my emails archive that will remind me not to trust anything *K*rap in the future.
Looks like a duplicate of 290363. I also lost some thousands of mails, but they were not important, so I do not care that much. But it made me stop using KMail, I am a Claws Mail user now. Too bad, I like KMail, but such a loss of data is totally unacceptable. And it does not seem that anyone cares about these reports. I do not understand this.
I am not familiar with the code, so mine are all speculations. The behaviour described in bug 290363 looks similar. I am not sure it is a duplicate: the author of 290363 write that he can't reproduce consistently. I can reproduce consistently by turning temporarily off the IMAP service. I also do not understand the lack of reply to the reports. To me it seems that Kmail/2 is abandoned/unmaintained. While I understand that open source software development has its pace and that we users must adapt to it, I do not understand how a "mature" community such as KDE makes such misguided decisions like replacing well working and tested software (Kmail/1) with unfinished prototypes (Kmail/2). That replacement has come too early and has done no favour to anybody. In my view, KDE's reputation has taken a serious hit and I am now looking to switch to another desktop environment all together.
*** This bug has been marked as a duplicate of bug 290363 ***