Sam Hartman <hartmans-ietf@xxxxxxx> writes: >>>>>> "Eric" == Eric Rescorla <ekr@xxxxxxxx> writes: > > Eric> bad decisions we have a mechanism for unseating them. > > Eric> 3. Decisions of the IAOC should be appealable (following the > Eric> usual 2026 appeal chain) on the sole grounds that the IASA's > Eric> processes were not followed. Those decisions should NOT be > Eric> appealable on the grounds that the decisions were simply bad > Eric> ones. > > Eric> I recognize that point (3) above is somewhat controversial, > Eric> so I'll attempt to justify it a bit. The simple fact is that > Eric> the IASA (both IAD and IAOC) will be constantly making > Eric> business decisions and if we are to keep the job manageable > Eric> (and in the case of IAD keep the costs down) they have to > Eric> know that they will not be constantly second-guessed by the > Eric> community on what kind of cookies they purchased. > > I agree so far. > > However people seem to be making a jump I don't agree with from they > should not be constantly second-guessed to they should never be > second-guessed. The problem with constant second-guessing is that it > takes up too much time and resources, not that it doesn't sometimes > improve quality of decisions. Well, sometimes it improves the quality of decisions and sometimes it makes them worse. One common pathological case is what's known in the economics literature as rent seeking. > So, my goal is to find the minimal > restriction that prevents a resource consumption attack against the > IASA in practice. I'd prefer to err on the side of allowing a > resource consumption attack and fixing it later than on the side of > being too restrictive in what review we allow. I'm not sure why you think this is the best approach. It's precisely in the current moment where things are in flux and, frankly, not functioning that smoothly, that it's critical to be able to move effectively. In my view, it makes sense at this point to give the IASA a fair amount of latitude to form its own processes (under the supervision of the IAB, IESG, and ISOC board, of course.) So, why not err on the side of effectiveness now and then fix it if we find good decisions aren't being made? > Eric> The Trustees' decision upon completion of their review > Eric> shall be final with respect to all aspects of the dispute. > > Eric> It seems to me that this is a good model to follow for IASA. > > OK. I don't like this model because I don't believe it is minimal > restrictive. I'd like to cite another parallel from RFC 2026. Under > RFC 2026 section 6.5, the IAB does not have the power to assert a > specific decision, only to void an existing decision. I'd rather see > limits on what the results of a review can be than on the subject > matter of reviews. I think that is a more minimal restriction and > believe it will be sufficient to avoid DOS concerns. I'm not convinced that this will do the job. The goal of this kind of influence behavior is often not to get any particular decision reversed but merely to make the cost of making the decision one doesn't like excessive. That said, it seems to me that having a decision one doesn't like voided, even if not reversed, is a pretty dramatic outcome when we're talking about cancelling a proposed contract. -Ekr _______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf