--On Monday, October 30, 2017 19:26 -0400 Alia Atlas <akatlas@xxxxxxxxx> wrote: >> I think its very mature in theory, but every time you dare to >> interoperate different versions, not even to say different >> implementations, i was often in for a fail. Not to mention >> ASCI-8859-UDF issues and the like. >> > > Then that would be feedback from the WG not to build on top of > it - or to require a better specification. > This is a judgement call from the WG, not an anything goes. That would be a reasonable answer if our WGs were always balanced among perspectives and interests and/or the IESG really did a careful audit each time a document was sent to it as to whether the WG has actually considered all of the issues related to a particular spec _and_ no one every argued strongly that a WG had approved of a piece of work and therefore the IESG and the community had to standardize it unless harm could be conclusively proved by those who objected. Unfortunately, while we often get those things right, we often don't and I've seen very little evidence that the IESG is really willing to take on that particular reviewer/ certifier/ gatekeeper job to the point of rejecting -- not just asking for modifications to-- a WG product when all of the ducks are not lined up. Again, I see the normative reference rules as protecting the IETF and the general perception of the quality and stability of our standards, not just a question of the approval (or not) of particular documents. best, john