> So when rough consensus is being assessed on a spec, the chair should > first ask who has read the spec and should then ask only those people > to do the humming. so if you want to exclude people outside the inner circle, all you have to do is make the spec so long and so opaque that nobody else can stand to read it.