Dave Cridland writes:
I'm mildly concerned you're infringing on (9) and (10), but if we're
going to dump any idea of addressing subaddresses further [if you can
follow that] then I'd like to canvas implementors on what they think
a subaddress is, whether it sometimes/usually fits this description,
and whether we might as well chop it further. (Say, by insisting on a
'+' postfix ).
FWIW, trying to bend in VERP and similar mechanisms strikes me as (5).
Hm.
That's not what I tried. I tried to define separator strongly enough
that it can be used, and weakly enough that no implementer thinks he
can process other people's subaddresses except as a 100% opaque
address.
Arnt
(who thinks 1925 should be on the standards track)
I think you could advance it straight to Draft status on the basis of
multiple independent interoperable implementations, at least.
Oh? The last adjective seems rather out of place ;)
Arnt