| Summary: | Broken pipe in KIO after network connection lost | ||
|---|---|---|---|
| Product: | [Unmaintained] kmail | Reporter: | Juha Tuomala <tuju> |
| Component: | general | Assignee: | kdepim bugs <kdepim-bugs> |
| Status: | RESOLVED WORKSFORME | ||
| Severity: | crash | CC: | des, faure, kollix |
| Priority: | NOR | Keywords: | triaged |
| Version: | 1.9.9 | ||
| Target Milestone: | --- | ||
| Platform: | Fedora RPMs | ||
| OS: | Linux | ||
| Latest Commit: | Version Fixed In: | ||
| Sentry Crash Report: | |||
|
Description
Juha Tuomala
2008-07-10 00:20:16 UTC
it actually appears that i've lost net connection as my irc nicks have changed. Then let's edit the summary to say so. ;-) *** Bug 166331 has been marked as a duplicate of this bug. *** David, any insight into this and #165540? It seems that the SimpleJob * stored in the coSlave (these are other slaves of the same type, right?) is invalid. This happens using disconnected imap when the network connection breaks. keyword bug #165540 might refer it as link... > This happens using disconnected imap when the network connection breaks.
for the record, mine imap is on-line one.
Could not reproduce on SVN trunk r949397. (KDE 4) Tried to reproduce this by killing my courier-imapd while KMail was doing I/O to the IMAP server. Tried several times, and each time KMail recovered flawlessly. It transparently reestablished the connection without any indication that anything had gone wrong. (And yes, I confirmed that the imapd was, in fact, killed.) Well, then let's close it until someone can reproduce it. Sorry, i have been nonresponsive. Actually i stopped reporting bugs to kde's bugzilla this summer since last release of kdeaddressbook closed tons of bugs including wishes (not reported by me) for no reason. So until there is a clear policy for bug management here, agreed by everyone, i'm not going to participate anymore. It's unrelated to this report, bug: the kaddressbook author closed all kaddressbook bugs because he rewrote completely the GUI while also porting kaddressbook to a new technology, akonadi. There is therefore -very- little left of the old kaddressbook (just a few dialogs). If you find that an old bug still applies in 4.4 once it's released, feel free to reopen it. I was a bit surprised by this move too, but it kind of makes sense: rather than the maintainer having to go through 500 old bugs to keep only the still-valid ones, why not share that work with others who can do it: those who reported them. (In reply to comment #10) > It's unrelated to this report, bug: the kaddressbook author closed all > kaddressbook bugs True. > because he rewrote completely the GUI while also porting > kaddressbook to a new technology, akonadi. There is therefore -very- little > left of the old kaddressbook (just a few dialogs). True. > If you find that an old bug still applies in 4.4 once it's > released, feel free to reopen it. Cannot since I'm not the owner. > I was a bit surprised by this move too, but it kind of makes sense: > rather than the maintainer having to go through 500 old bugs to keep > only the still-valid ones, why not share that work with others > who can do it: those who reported them. I agree with you fully. But you missed that he flushed completely valid feature *wishes* which in most cases don't have anything to do with rewrite which doesn't change features that much. And that was the exact case, *those* he closed *were* valid, had good discussion history which at least I consider somewhat valuable community participation. Another point is to modify vanilla bugzilla so much, that community cannot take part to clean up those 500 bugs. That's what people exactly would like to do, but are not even trusted for that - apparently. Lot of those wishes were also so old, that original report might have disappeared (at least I've not seen re-openings). I've noticed that some developers ignore the reports completely and let the reports pile up. Makes then wonder, why to have such database at all. Showing such ignorance and arrogance towards the community drives people away by just plain disgust. In my opinion at some parts KDE has grown too big and popular and healthy community doesn't exist anymore. Sometimes it reminds me from GNOME-users-are-stupid-episode. I used to report most of the crashes and thought that it would make difference, but whatta hell, now I've more time to something else. in last i mean s#original report#original reporter# > I've noticed that some developers ignore the reports completely and > let the reports pile up. Makes then wonder, why to have such database > at all. > Showing such ignorance and arrogance towards the community drives > people away by just plain disgust. In my opinion at some parts KDE > has grown too big and popular and healthy community doesn't exist > anymore. Sometimes it reminds me from GNOME-users-are-stupid-episode. It is certainly true that there are problems with the bug handling. We developers do KDE development in our free time, and there are so many reports that it is impossible to deal with all of them. For example, I do not ignore KMail reports, I read all incoming reports and comments, but I don't really have the time to actually deal with them. The Bugsquad helps us there a lot, especially in detecting incoming duplicates. Still, the database is useful, I personally find it especially useful to detect regressions in new releases or really bad bugs (which I admittedly didn't find the time to fix, for example the famous Slave::deref() crash). With the limited resources we have, it is natural that the reports 'pile up', I wish it would be different. That doesn't mean however that there is no use in bugzilla. In the KAddressbook case, there is simply not a single developer working on the old KAddressbook, so having those reports open would not help. It was a mistake to close the wish reports though, we should have thought about that a bit more. Please do not attribute our bug handling to ignorance and arrogance, that doesn't reflect the truth. We care about our software in our users, it is just that we don't have the resources to express that by dealing with bug reports in a better way. I wrote somewhat lengthy response and konqueror crashed. I can feel the irony here. :-/ Perhaps i try later again. > Cannot since I'm not the owner. If you're interested in helping with bug triaging and/or reopening the kaddressbook wishes that still apply, I can you give the bugzilla permissions for doing so. The deal is: no more ranting, though :) > I used to report most of the crashes and thought that it would make > difference, but whatta hell, now I've more time to something else. We always (try to) fix crashes, I don't see what the closing of old kaddressbok bugs has to do with that. The old bugs don't apply anymore, and the old wishes were closed by mistake, I agree. Why not open the rights for everyone? I never saw any problems in Red Hat's bugzilla - for example that someone would have been setting wrong versions or components and vandalizing it. Current strict policy is just damaging the community itself more than non-proven vandalizing would do. Giving me the rights is not going to solve the lack of manpower in needed scale and thus doesn't really solve anything. Because we have seen people being unreasonable, like reopening bugs over and over again, for instance in the few "controversial" bugs (e.g. those related to politics, like which countries should appear in the country list, and a few other similar issues). We open the rights to everyone who asks and who appears to be reasonable. But we cannot just open it for everyone passing by. This solution does scale - we have a growing group of people who work on triaging bugs, whereas we had none just a few years ago. David. Whatever, I'm done with this bugzilla unless I see a written policy that I can agree with. |