> Even for software where ‘zero bugs’ was important they’d just delete a whole bunch of bugs and see if any bounce (come back). Eventually people get sick of refilling so they get to zero but bounce by exhausting the very people eying time help them.
The open-source equivalent of this behaviour is "stale bots" that close or lock issues with no activity. Or "moving discussion to a separate tracker", with all bugs getting closed and a polite request to re-open them. Or a "locking bugs older than X months, please open a new one if still applicable".
Sure, opening it again isn't a big deal. Re-opening all bugs ever opened by all humans is just pointless work for no obvious benefit.
Having zero open issues should never be a goal, any mature project has open issues. Trying to reach zero is chasing a number that won't make a product better; it's just a number.
Those bots drive me insane; it punishes the most helpful the most. I would much rather they leave it open. I often find myself digging around closed issues to find people with similar problems.
Until some maintainer writes "Please stop replying with +1 etc, we know the issue. Use reaction emojis instead.". Bonus points if the issue also gets limited to contributors only.
Agree, but forever open issues doesn’t give any value too. Because you don’t know if this issue still applicable for latest version of app/library.
I think stalled issue should be closed, especially minor one.
It’s the opposite. The older the bug, the less likely it will be fixed. Those old bugs are one of the most reliable sources of information how the software behaves.
Especially when it’s the developer that is inactive. Like, we’ve ignored you for 7 days and now we have committed to ignoring you forever. If I find this I won’t even bother filing a report. I get that we’re not owed attention, but they’re not owed bug reports. As a dev I have won’t fixed issues due to lack of dev time. I think the problem is when it’s paid software, or an open source project connected to paid software. There seems to be a push of testing onto the users which is disastrous when combined with a policy of ignoring them. Microsoft went through a period of aiming for two testers for each dev. Post my tenure it seems that devs are supposed to the test themselves.
The open-source equivalent of this behaviour is "stale bots" that close or lock issues with no activity. Or "moving discussion to a separate tracker", with all bugs getting closed and a polite request to re-open them. Or a "locking bugs older than X months, please open a new one if still applicable".
Sure, opening it again isn't a big deal. Re-opening all bugs ever opened by all humans is just pointless work for no obvious benefit.
Having zero open issues should never be a goal, any mature project has open issues. Trying to reach zero is chasing a number that won't make a product better; it's just a number.