Ideally I’d want the bounty to be proportional to the value produced. Some problem with this are:
any such estimate is going to be noisy, so this only works if people recognize that the estimate of value is going to be uncalibrated some of the time—e.g., “have $100 bucks for writing something which took you 50 hours and you view as both traumatic and deeply important” isn’t a great look
it’s unclear how you would even go about calculating that estimate, maybe value of community team * percentage improvement. The first item is unclear, but it could be bounded by the money that’s spent on the team.
So, e.g., $40k to $90k * 1 to 5 FTEs * 0.1% to 3% improvement * 1 to 3 years = 170 to 9.4k (mean 2.7k).
Not expressing an opinion on whether a bounty is a good idea, but it seems you would have more of a basis to set the amount if OP published their complaints anonymously first rather than basing it on the broad class of possible complaints.
Ideally I’d want the bounty to be proportional to the value produced. Some problem with this are:
any such estimate is going to be noisy, so this only works if people recognize that the estimate of value is going to be uncalibrated some of the time—e.g., “have $100 bucks for writing something which took you 50 hours and you view as both traumatic and deeply important” isn’t a great look
it’s unclear how you would even go about calculating that estimate, maybe value of community team * percentage improvement. The first item is unclear, but it could be bounded by the money that’s spent on the team.
So, e.g., $40k to $90k * 1 to 5 FTEs * 0.1% to 3% improvement * 1 to 3 years = 170 to 9.4k (mean 2.7k).
Ehh, this seems decently valuable.
Not expressing an opinion on whether a bounty is a good idea, but it seems you would have more of a basis to set the amount if OP published their complaints anonymously first rather than basing it on the broad class of possible complaints.
Makes sense