Both... > > If an AD cannot get cooperation from the WG and cannot enter a ballot position > that supports sending the document forward, then the AD should switch to > "abstain." > > > > That's completely inappropriate. A document reviewer should never be > expected to pretend like he doesn't have a problem with a document. To expect > an AD to change his vote to "abstain" is asking him to be dishonest and/or shirk > responsibility. > > iirc the IESG used to call this kind of abstention "holding > one's nose", but it can't be distinguished in the ballot from > an abstention "for cause" (conflict of interest). In theory, > you can find out which applies from the history in the tracker > or from the IESG minutes. You can tell the difference because one is a "recuse" and the other is an "abstain" AFAIK the process that Keith calls out has never needed to be executed. It exists to handle deadlock which might otherwise have to wait until the make-up to the IESG changed or until NomCom was woken from its peaceful slumber. ADs are not universal policemen and serve the will of the community. When an AD is clearly in the rough against the consensus of the IESG they should not have the power to individually block the progress of work for which there is clear IETF consensus. Note that when an AD does abstain he usually writes a strong comment in the datatracker explaining why he is abstaining. Adrian _______________________________________________ Ietf mailing list Ietf@xxxxxxxx https://www.ietf.org/mailman/listinfo/ietf