@maaretp I agree. It really depends on tools in use when it comes to analysis time. Some are 90%+ accurate, so analysis is a waste of time. Some are ~20%, so jumping to fixing might be a waste.
@maaretp The tricky thing with fixing is that it can be “whack a mole.” If you have 100,000 or more, which 500 do you start with? The ones actually surfaced have greatest benefit. The “low hanging fruit” where you’re just bumping a dependency version are another great start.
@johncutlefish Maybe another category: Power / Status. Ex:
- Passive aggressive (feigned support)
- Rewarded sabotage
- Fools errand projects forced by leaders to undermine teams for political gain
- Self before business
@manicode Maybe pre-commit is fine for some stuff and pre-push is right for other stuff while other things are CI only. It’s a tradeoff between how big the workflow interruption is and the speed of feedback. The cumulative interruption may be worse than delayed feedback.
@noahsussman If mimicking people, the estimates should begin to drift toward the average of the session after being called for more than a few times in 15 minutes.