Everyone who uses software knows that bugs exist: bugs in hardware drivers, in operating systems, in normal applications, hell, even in bad designed web pages.
Users of Open Source software are quite used to be able to report bugs in appropriate forms – the best known is probably bugzilla. Since I use quite a lot of Open Source software I try to submit bug reports of all kinds to all kinds of projects: KDE, GNOME, Fedora Core, Tango Project, freedesktop.org, to name the biggest. Overall I submitted umpteen bugs, I think I’m quite close to 100 bugs which I submitted or made significant contributions to the report.
Between all these bugs, there is one bug number one for me. This bug does not prevent me from doing something – like the PAP feature wish-bug for NetworkManager – and it is no cosmetic bug like the firefox finder bar for konqueror wish. It is not even a “doesn’t work” bug like my WLan problem.
It is much meaner: it only slows you down. And that’s the reason why I could go nuts from time to time, and throw my computer out of the window: Just from time to time, not predictable but often enough, input fields like blog edit flieds and wiki edit fields in konqueror, the chat field in kopete, the mail window in kmail or once even the field in kate, slow down in the display of characters. I continue writing, but suddenly the characters are not displayed as I type, but with a delay. And since I continue typing with normal speed, this adds up more and more. After some words typed with normal speed while the slowdown is happening, I have to wait for 2 or 3 seconds until all characters are displayed and I can check if I spelled something wrong, etc.
Perhaps you cannot imagine how frustrating *that* is. But it drives me nuts, I would even pay money to get rid of this bug.
But one of the main problems is that it seems to happen very seldom, and so far only on a few set of machines. Additionally, there had been other bugs which caused similar effects, so bug reports in this area mix different problems because they have the same result. And yes, I tried to sort out some of these mixtures, but failed horribly.
However, I tried to fill a new bug report, and it was marked as duplicate, so I entered that bug and tried to post as much information as possible. Until today these were some descriptions and 3 debugger outputs. Today I read something about xtrace (which should not be confused with xtrace from the Fedora Core glibc-utils package), which records all signals an application sends to X – and I produced an output. Approximately 11 MB. Not bad, and maybe not that amount of data a developer can go through :(
But, to use one of my #1 quotes from Lord of the Rings: there is still hope! I mean, that’s the only thing left if there is nothing else ;)
Oh well, I try to be confident, perhaps one day there will be a developer who ever wanted to get used to xtrace outputs, and analyses this output. And with some luck the needed information are right there, and there will be a bugfix upstream, and my world will be perfect again…
Maybe I should offer a significant amount of money. Hope moves thoughts, money the other things.