From: Russ Allbery (rra@stanford.edu)
Date: Fri Jun 01 2001 - 17:41:22 CDT
Bill Davidsen <davidsen@prodigy.com> writes:
> The whole issue of breaking existing archives is smoke and mirrors,
> there are *zero* archives implementing the Archive header, so we can't
> break them. And if they do implement it, we can specify that they SHOULD
> bloody well respect the author's wishes! We have no justification for
> anything stronger than SHOULD, but I don't see the argument that we need
> anything weaker, either.
The SHOULDs that I've seen proposed aren't "you SHOULD, if you implement
the Archive header, honor it" but instead simply "you SHOULD honor the
Archive header." The latter declares all existing archives not strictly
conforming until they add support for Archive.
Per most of the definitions that I've seen thrown around for an archive,
there are many existing archives that don't honor X-No-Archive.
-- Russ Allbery (rra@stanford.edu) <http://www.eyrie.org/~eagle/>