--On Friday, March 4, 2022 09:01 +1300 Brian E Carpenter
<brian.e.carpenter@xxxxxxxxx> wrote:
> David,
> On 04-Mar-22 01:08, David Noveck wrote:
> ...
>> The safe thing is to limit changes after consensus to those
>> that do not require such extensive review by the authors who
>> may, as someone pointed out, have lost the necessary context.
>
> This is a bit of a strawman, but it is entirely possible that
> an editor could resolve an ambiguous sentence in a way that
> effectively changes the meaning of the whole document; for
> example, changing a terminology definition that is used
> throughout the text.
Your supposed strawman is eventually what brought us to
pre-publication review of documents that, through a few stages
of review, brought us to AUTH48. And problems about authors
either tuning out or taking advantage of questionable changes to
substitute their preferences for approving-body consensus. In
other words "been there, tried that, and it did not work out
well".
Probably worth mentioning that those checks were first put in
place at a time when the editing team had considerably more
expertise and depth about the subject matter of RFCs than
today's RPC and any we are likely to see in the future. Even
they could got get decisions about what did or did not change
meaning right 100% of the time.
> We can't avoid this by issuing meaningless instructions to the
> editor like "don't change anything that matters". How we
> resolve it today is by the editor writing to the authors at
> AUTH48 time in terms like:
>...
If the editor even notices that a change might be substantive.
The current RPC team gets that right almost all of the time, but
final review by subject matter experts is still important (if
nothing else, to keep this a team effort and avoid pressure on
the RPC for absolute accuracy and universal knowledge.
john
