At 11:39 AM 10/26/2012, John C Klensin wrote: >In principle, I have no problem with setting up a list of >repeated/ long-term non-feasance, non-appearance, or >non-responsiveness conditions that are treated as equivalent to >a more formal resignation unless the body of which that person >is a member votes to waive the rule (i.e., not accept the >resignation). I actually think that would be completely >reasonable and that we probably should have established such >principles years ago even though I'd expect a lot of quibbling >about setting the thresholds for "repeated/ long-term". This is exactly what I'd like to avoid. No matter how we set the conditions, someone will argue - successfully - that the conditions are not met. (For example, I attend one out of every 3 meetings for exactly 5 minutes - I obviously haven't abandoned the position, but I really have abrogated my responsibilities). Etc etc etc etc etc... Or something will come up that the conditions just never anticipated (the current example springs to mind). Let the bodies decide if the member is doing their job. If 2/3 of them think no and they can convince 2/3 of the confirming body, then its time for him to go. This handles pretty much any "emergency" condition I can think of and does it in only a little more than a month. And note, that unlike a recall, there is no bar to being re-appointed by the Nomcom in the following term. The recall process remains for the body of the IETF to be able to affect a removal with some amount of formalism. I'm pretty much going to object to any condition based model that anyone proposes, because we're really bad at a) figuring out the complete list of all possible conditions that could ever happen, b) writing conditions that can be objectively evaluated, and c) figuring out how to decide when specific conditions are met (because of the lack of objective criteria). In addition, people have been carping on the mailing list about how we need to be flexible - and condition lists by their very nature are not flexible.