My understanding is that Brandon's system previously required no
out-of-band label at all. Whenever a message was processed in their
infrastructure, they would use a single algorithm, and that algorithm
contained some heuristic for dealing with messages which were using 8bit
data in headers. Everything worked pretty well for them.
Now, the EAI introduced another syntax, and that syntax breaks their
heuristics. Ned's suggestion is to make use of the out-of-band labeling and
extend the heuristic to only apply the old crutches when the label says
that this isn't an EAI message, while Brandon's complaint is that they use
no such labeling at this time because it isn't required at all. Brandon
also suggests that there are many other places which process these messages
without the external metadata.
By "out-of-band" or "external", I mean what gets set in the ESMTP envelope.
Hope this helps, and please correct me if my understanding of your
positions is not correct.
From my point of view, it would be most interesting to see what happens
when I deploy e.g. a Postfix version which understands this new RFC along
with an older Dovecot which doesn't do it. Suddenly, an IMAP client is in
the same position as Brandon describes because there's no way of accessing
the ESMTP envelope through IMAP.
Cheers,
Jan
--
Trojitá, a fast Qt IMAP e-mail client -- http://trojita.flaska.net/
_______________________________________________
ietf-822 mailing list
ietf-822(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf-822