not so. 2822 can be advanced in grade without revising it. I think
it would also be reasonable to publish a separate list of errata and
declare 2822 + that list to be a draft standard, though I don't know
whether IESG would go along with that.
I think that would be needlessly annoying to implementors. Reading
errata lists is frustrating and time-consuming.
and trying to read through a long document looking for minute changes
isn't frustrating and time consuming?
(not to mention the tremendous amount of work required to revise that
document)
obviously this depends on how many things need to be changed. but I
recommend to start working on a list of errata. if that list gets to be
long enough, it's not much more work to fold those changes into the
document. just try to avoid the mindset that the entire document is
being revised.