I have a maximum of 1 day of safety buffer set in the auto-rachet for a goal (number of pomodoros completed). I also have weekends-off. When I go crazy on Friday and blast past my goal, I would expect to be safe on Monday. But the weekend-off eats the 1 day of safety buffer and Monday is always an eep-day.
Is this the intended behavior? If so, I’d like to propose revising the design.
I see some earlier discussion at the following links but they’re pretty old:
Hi! Thanks for sharing – always useful to hear this kind of question.
Weekends off are no different from other breaks, which are no different any other buffer you gain except that a flat spot itself will not be ratcheted. There’s no way I can think of for distinguishing between “you have buffer because you had a period with a rate of 0/day” (other than the period with the flat spot itself) and “you have buffer because you’ve overachieved”.
So because any buffer you have before the weekend is not a flat spot, it can be ratcheted (and is).
So I wouldn’t say it’s exactly intended behaviour, but it’s also not surprising behaviour, and I’m actually not sure how we could easily change it – if anyone has any thoughts, it’d be good to hear.
I think the way proposed in the past has been true breaks, where the bright red line just stops existing for the weekend. Personally, I’m not sure how that would work in other ways (does the goal functionally not exist? can you enter data? should you be able to? is it quantified self if you can’t enter data? but if you can enter data, the graph needs to be plotting it somehow…) so I’m not a huge fan, butit’s something that’s been mooted as a good move for various reasons.
To be clear (just in case it trips anyone else up) autoratchet is aware of breaks. It should never ratchet away an actual break. What you’re asking for is that it take breaks into account and ignore them when calculating how many days of buffer you have and whether it should run, if I understand correctly – is that a good summing up?