How hard is it to write a bug tracking system that doesn’t suck? Quite hard it seems …
I spent ten minutes trying to work out how to enter a bug in a Flyspray-based BTS. And I gave up. I gave up entering a bug! Isn’t that, like, the most important single task that BTSes should enable me to do?
Trac is a pile of crap too. The enter a bug button is hidden away somewhere (it’s called “New ticket”), and the main interface seems to be arranged around producing arbitrarily chosen “reports”, none of which I have ever found to be in the slightest bit useful.
Mantis … urrrgghhh. OCaml upstream use this to track bugs and it’s horrible.
And .. Bugzilla. Oh Bugzilla. When I get my own starship, I hope it’ll be simpler to use (and faster) than Bugzilla.
I would encourage everyone to try out (proprietary) FogBUGZ for an example of a usable BTS. It’s proprietary software, but you can learn from it. You can learn that (a) most of your users won’t be BTS experts, and (b) you should encourage as much feedback from all your users as possible. That means doing usability tests and concentrating mainly on making it as easy to use as possible. It’s the most important thing for BTS authors to do. Clue: if your users cannot work out how to enter a bug, then you should go back to the drawing board.
