Zefram wrote: > Ed Murphy wrote: >> Proto-Proposal: Per-case panels > > Messy. It's a lot of extra complexity and room for bugs. It doesn't > interface properly with the rest of the judicial system, as evidenced by > your need to increase the power of R911 so that it can make exceptions > to the framework rules. > > The rationale of bending reality to match your outdated automation > is appalling. We have a cleanly-engineered judicial system, which > was brought in to do away with the former system's cruft, such as the > complex and sui generis appeal board assignments. You propose to undo > this progress, when you should instead be redeveloping your database to > reflect the new structure.
So long as appeals are handled by sets of three, they're going to be sui generis no matter what you call them. As for complexity, I've already pointed out how the old system (under which the individual significance of an individual action could be determined immediately) is simpler than the new (in which it cannot be determined until the panel acts); the rationale of /this/ added complexity is understandable (the desire for unanimous panel action), but debatable (appeals tend to drag out longer). I've also already pointed out how my proposed revision is driven not just by the state of the automation, but also by the concept of partial continuity (deliberately choosing a panel with some members in common with the previous one). Furthermore, the existing concept of panels existing independently of cases is confusing. Redeveloping the database is not in the cards. If the proposed rule revision is voted down, then I can simply enter recusal records for all three members individually, rather than just the ones who aren't members of the new panel. I may also switch to a policy of either supporting majority opinions pre-emptively, or assigning completely new panels when the members of the old panel fail to request such support.
