Chuck Zmudzinski <[email protected]> writes: > The theory behind my proposal is that there is a diversity of viewpoints > that stems from a diversity of roles within the project. If the voting > members are restricted only to the contributors who volunteer for the > role of developer, then the full diversity of the Debian community is > not reflected in a vote that only comes from the pool of formal > developers.
I think it depends a lot what the viewpoints are being diverse *over*. If you've formed a group of people to go pick up litter, it's not helpful to have a diversity of opinions about whether or not litter should be picked up. That is just silly. And, somewhat more on point, it's not helpful to have a bunch of members who have no intention of going out on the street and picking up litter, but who have strong opinions about how other people should do so. This is not useful diversity. There are numerous Linux distributions with different viewpoints about how to make a Linux distribution. That's diversity, real diversity, in terms of distribution goals and structure. It's the kind of diversity that involves completely independent organizations making completely independent decisions with different financial structures, different founding principles, and so forth. As a Linux user, you can explore that diversity by picking and choosing which distributions you install and use. Within any given project, including Debian, we're all trying to do something together, which inherently requires some amount of agreement and consensus. At some level, this is the opposite of a diversity of opinion. We have to roughly agree on a course of action, at least at some level and with some rough, in order to make forward progress. One of the things that's helpful to avoid endless decision paralysis, or to avoid a bunch of people who are not in a position to do the work voting for things the people who are doing the work don't want to do, is that we ask everyone who is voting on major decisions to have significant skin in the game. The implicit assumption behind GR voting is that to some extent you're then going to go help implement what we all voted for, because we're the only people who can. The voters are the same group as the implementors. We're using voting to help guide consensus among the same group that has to do the work. If no one does the work, the GR is useless and pointless and there's no reason to have held it. That tends to significantly inform our voting. It's something that I think about with every GR. To make that concrete, you can hold as many GRs as you would like saying that every bug in Debian with a tested patch should be applied and uploaded, and it will accomplish precisely nothing. Changing who votes on such GRs is useless; the whole theoretical model behind such a vote is wrong. Work happens in Debian because someone does it. Voting doesn't change that; what changes that is more people doing work. If you think a package is being neglected by its maintainer, we have processes for that, but they mostlly require that you, the person who wants to make a change, step up and be the maintainer. If no one is willing to step up and be the maintainer, the package will continue to be neglected. Voting is not a magic spell. > Michael Stone followed me to this list and condemned for me asking > questions here on this list. There is no way *he* considers me a member > of the Debian community who has a formal voice as a Debian user. This is Michael thinking you're being annoying and wasting people's time and telling you so, while warning other people who don't follow debian-user that arguing with you may be futile. It has nothing to do with whether or not he considers you a member of a community. I can assure you that he'd say the same thing to a DPL if he thought they were being annoying and futile to argue with. -- Russ Allbery ([email protected]) <https://www.eyrie.org/~eagle/>
