Hi Dave, On 12/11/2015 03:45, Dave Crocker wrote: > Brian, > > Were this a reading comprehension test, you'd get a failing grade. > You've misinterpreted or invented, rather than dealing with the plain > text as I wrote it. It says what I meant. We disagree profoundly about what your words mean. I don't think it would be productive to continue mutual textual analysis. I do want to say this: We have given the ADs power of decision over what gets published. They take this power very seriously; that's intrinsic in the way they are selected and appointed. It's the first thing you learn as a new AD: the buck stops here. If we want to stop the ADs spending large amounts of time on document quality, we have to take away their power of decision over what gets published. I don't advocate this, because I don't know how else to get rid of the serious errors that survive WGLC. Brian > > > On 11/10/2015 11:14 AM, Brian E Carpenter wrote: >>> That an AD occasionally catches some important error in a >>> draft is a distraction, not a justification. >> >> It is rather beside the point whether an AD notices an error personally, > > Given the amount of time that ADs spend on reviewing specs themselves -- > they are supposed to read them all -- no it's not. > > >> (b) it is the AD who has the power to ensure it gets fixed. > > I didn't comment on IESG or AD authority. If you want to introduce a > new issue, please decouple it from my note. > > >> Also, I think that "occasionally" is plain wrong. A lot of of quite serious >> errors survive WG Last Call and are detected during the part of the process >> that the AD supervises. I am not saying that all documents have such errors; >> I estimate that ~20% have significant issues when they leave the WG. > > A lot of serious problems survive AD review. That was my point: We > publish flawed documents. Lots of them. With lots of flaws. > > I didn't say that ADs don't catch any flaws; I said they don't catch > that many and they don't come close to catching them all. > > Rather than treat the occasional, significant problem that is caught by > an AD as justification for all the effort ADs put into doing their > review, treat all the time spend /not/ catching errors as a strategic > mis-expenditure of a very scarce resource. > > Cost/benefit: consider carefully. > > >>> Nor does it succeed because of an AD's "leadership". >> >> The scare quotes are tendentious. > > Use of the word tendentious was redundant (and inappropriate). When > someone would publish a note offering criticisms and advocating for > specific actions the essence of the exercise is "bias". Advocacy does > not reflect neutrality. > > >> The point isn't that the AD is in a >> leadership position. It's that the AD has the power hold up the document >> until it's fixed. > > Calling it leadership does not make it leadership. Calling it > leadership inflates the actual nature of the work and imparts a role > onto ADs that they don't have. Again: the initiative and the work come > from the community, not ADs. > > >>> The reality is that IETF work does not succeed or fail because of the >>> technical input from an AD. >> >> Correct, but completely beside the point. > > Here's the nub of your comprehension failure: It is /exactly/ the point. > > ADs wind up trying to do lots of technical work as ADs, when that's not > the f'ing job. > > >>> Defining the AD job to be fundamentally more modest and fundamentally >>> one of facilitation will simultaneously make the job accessible to a >>> much wider range of candidates and make the job more useful to the >>> community. >> >> If you advocating removing the power of decision from the IESG, > > Again: Please don't introduce issues that I neither raised nor intended. > > Homework assignment: Try re-reading my note, but this time, try to take > the words as they are written. > > d/ >