After further thought and several comments (some off-list) from
people who don't feel that it is a showstopper one way or the
other but would prefer to see it go, I propose the following.
This will make no sense unless you have read notes from the
earlier thread today.
The current sentence reads
"Such responses could, in principle, be covered by an SMTP
extension in the future."
Since there is a normative DNS prohibition on CNAME targets (in
RFC2181), that sentence is, strictly speaking, untrue. It could
be corrected by saying:
"Such responses could, in principle, be covered by DNS
changes or extensions and an SMTP extension in the future."
Or we could drop it. My instinct at the moment (i.e., I
reserve the right to change my mind) is that dropping it would
be better. Given the language that already occurs elsewhere in
2821/ 2821bis about the things that extensions can change, it
doesn't really add any information. And, unlike the original
form, this is started to get convoluted. 2821bis already has
enough of a surplus of convoluted text and another instance,
while not a horrible precedent, is not needed.
I'd like to hear from John Leslie, who originally suggested
putting something like that in. If he is now persuaded that it
should be dropped, I'd find that convincing and perhaps others
would as well. On the other hand, if he has a strong argument
for retaining it (presumably in the revised form above), we
should all hear and consider it.