>> >I think it is more fruitful to think of the system between us >> >maintainers being anarchy, not democracy and therefore we should be >> >aware of and work with consensus. >> > >> >For a nice description and one implementation of consensus, see: >> > >> >https://www.noisebridge.net/wiki/Consensus >> >> It does look that currently it is unattainable as such on some >> questions. > >On which questions do you think we currently do not have or cannot reach >consensus?
For overrides, some people explicitly consider that it is better not to have too much possibilities for overriding package settings, even if it would come for free. It rules out true consensus - there could be some tradeoff declared consensus, but it is more honest to say that we look for a tradeoff between two groups of developers with directly opposite interests. I cannot be sure about anything in the case of the parallel build fiasco. I cannot understand the values of the people who committed what we have now, so I cannot say whether a true consensus would be possible there. _______________________________________________ nix-dev mailing list nix-dev@lists.science.uu.nl http://lists.science.uu.nl/mailman/listinfo/nix-dev