--On Monday, March 25, 2019 13:13 -0700 S Moonesamy <sm+ietf@xxxxxxxxxxxx> wrote: >... >> I'm not certain I understand the distinction you are making. >> As Appendix A points out, RFC 3777 raised the number of >> signatures from one to 20 because of concerns about what I >> summarized as frivolous recall attempts and intentional >> attacks on the system by people whom others might >> (retrospectively at least) classify as trolls or worse. I >> like "frivolous" better than "capricious" to summarize that >> concern, but either is consistent with the discussions about >> making recalls sufficiently more difficult to deter or >> prevent bad behavior that brought about the more restrictive >> issues of 3777. > > There is the following sentence in Section 7.1 of RFC 7437: > "The Internet Society President must not evaluate the recall > request". Who is going to evaluate whether the persons who > signed the petition made a frivolous request? In my opinion, > it is better to leave it to the Recall Committee look into the > request. Ah. Misunderstanding, not disagreement. The intent of the changes in RFC 3777 was to make it significantly harder for someone to start the rather heavy-duty and costly (in community time if nothing else) recall process if they did not have a serious complaint. That was specifically intended to discourage efforts that I would describe as "frivolous", Adrian might describe as "capricious", and others might described in sentences involving terms like "troll". In no case was there any intent to have anyone other than the recall committee evaluate the content of the appeal in terms of any of those categories; the intent was simply that, by requiring multiple members of the community -- members who met some standard for participation and absence of being an obvious organized cabal -- to endorse (by signing) the request, things would get a lot harder for one person whose concern was not shared more generally to initiate a recall. I think we all understood that the new (as of 3777) model would prevent all such problem recall efforts, but the motivation for raising the threshold was to at least require more effort. If that isn't clear enough in the document, I hope you/we would welcome proposed text. The current draft doesn't change any of that, it just recalibrates the threshold and requirements for generating/ endorsing a recall effort. > There is a threshold for signatories. The minimum number of > signatories ensures that the justification is supported by > more than one person and the primary affiliation restriction > ensures that the persons are not from the same > company/organization. Right. > As a comment about "frivolous", it is currently possible for > anyone to appeal a decision. Does the IETF want to make > appeals difficult given that it is possible to file an appeal > which might be viewed as frivolous? I hope not. And, fwiw, two key differences between appeals and recalls is that the former don't involve the burdens on the community that setting up a recall committee, etc., do _and_ the appeal process is intended to accommodate "hey, have you considered the following issue in making your decision and, especially if not, please reconsider the decision" situations rather than any claim of misbehavior. Very different and I don't think having much to do with each other. > The question which could be asked by a non-participant is > whether the first step of the current procedures is there to > protect IAB or IESG management members from being recalled. > Has there been any frivolous recall attempts between RFC 2727 > and RFC 3777? About twice as many as obviously frivolous recall attempts since 3777, i.e., approximately twice zero (although there were a few threats). One could argue that even the change from "anyone can initiate" was not justified by history and experience, but it appeared to many people that the "anyone" provision was just a problem waiting to happen. >... best, john
