15 ms·
I'm not a fan of the "stale bot" approach to bugs either, but I think what it's doing can at least somewhat be classified as part of the "facts" side of the div
by pm215 2y ago
I'm not a fan of the "stale bot" approach to bugs either, but I think what it's doing can at least somewhat be classified as part of the "facts" side of the division. The initial bug report conveys the fact "version 3.5 crashes when passed the --wombat option". A year later that fact is still true but is of very little interest to the dev team: what they want to know is whether "version 5.0 crashes when passed --wombat" is true or not. The stale-bot is pushing the obligation to test that new different fact onto the bug submitter. (I think it's better to leave that with the dev team; plus unless the devs plan to actually fix it then retesting on 5.0 is just gathering another fact that isn't going to matter...)
- hinkley 2y agoIf a problem hasn't been fixed in 1.5 versions then I'm likely to have moved on to a different tool, and I no longer have the context to reproduce it. If I'm the only one who felt that way maybe that's efficient. But that's not always the case and I've had bugs get closed when there were 4 other commenters with the same problem.