On Tue, Apr 13, 2010 at 11:46 AM, Ryan Delaney <ryan.delaney(a)gmail.com> wrote: [snip] But the opposite approach is as bad or worse: If every issue must be argued anew when someone brings it up then the ultimate outcome is that by sheer pigheaded persistence you will eventually get your way once everyone saner has tired or repeating the same argument, — or even ignoring a single persistence force — that we'll always eventually conduct things according to the initial impressions of a typical uninformed person (because, again, the informed people will drop out). I don't think the regular references to old discussions are a prohibition against a new discussion, but you certainly shouldn't spin up a new one without a reasonable understanding of what has been argued before. Going back to the software and pulling in Martijn's comments— I actually see something of an opposite effect here. It looks to me like this system would leave people in a position of having to perpetually continue and maintain an argument for fear that a slow drip of based counter-opinion won't erode the standing of the position. For example, as a community we might think really hard about the use of, say, ellipses in article text and after careful analysis and vigorous debate come to a widely held conclusion that they shouldn't be used. As far as I can tell every time some newbie bumps into this rule— before they've had a chance to calmly consider their position— they'll head over to the page and vote down the decision. Eventually the page reflects a conflicted (or opposed) state, even though that may not accurately reflect the project consensus at all, it may not even reflect the current views of the people who opposed it now that they have some distance from the incident that first brought them to the rule and additional experience with the project. I think it's important to for discussions to have closure, so that we can count on the most sane and considered people being willing to invest significant time making excellent arguments, but the closure ought not be absolute because situations and opinions change. I can think of no way that software could systematize a balanced system like that.