Ian Clarke writes: > One example that sticks out is this justification for saying a website > redesign > would have absolutely zero value: "We have just redesigned it. Throwing that > away too soon would disappoint the volunteer authors". I find this rationale > infuriating.
You ask for the value people assess for features. I think you have to live with honest answers, though you don’t need to agree with them. There is no way to use this poll as a basis for any "rational" decision, but that was clear before we started. At best it provides a picture of the range of opinions in the project. The only way I see to get a semi-rational from this would be some kind of Condorcet voting using the Schulze method to resolve ties¹: 1. Use the average cost estimate for each item from the next step to weight the items in each voting separately (i.e. xor value item 1 * average cost estimate, …). 2. Then use those values to calculate an order for the items in each vote by ROI: highest return of investment first (i.e. xor ordering: 5, 3, 22, ...) 3. And finally do condorcet voting¹ to get an estimate which item should be the first for which we should check whether we find someone who would take part of the funding to complete it. Doing it separately for users and developers would at least provide an interesting metric of the differences between both groups. Add some normalization and this is the best option I see. It has tons of flaws, however (I won’t start to describe them, they should be pretty obvious) and is not suitable to really take decisions based on it. In short: It is a horrible system, but all others (I can think of at the moment) are much worse. Except for just using this as a non-binding picture of the range of opinions and then actually taking decisions. Best wishes, Arne ¹: https://en.wikipedia.org/wiki/Condorcet_method — more exactly: the Schulze method: https://en.wikipedia.org/wiki/Condorcet_method#Schulze_method -- Unpolitisch sein heißt politisch sein ohne es zu merken
signature.asc
Description: PGP signature
_______________________________________________ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
