Re: Whether 8-bit SMTP? And how?

[kankkune@xxxxxxxxxxxxxx (Risto Kankkunen)]

> But I think we should not restrict the RFCs as this may break
> some perfectly valid implementations out there.

I agree.

I understand the problem with control characters is that they might
cause havok with some hardware. Because the set of control characters
varies in each hardware, I would think the best place to get rid of them
is in UAs, or MTAs when it delivers a message to the spool. That is,
make the filtering completely local operation, not needed to mention in
the RFCs.

If the character set must be restricted in some RFC, I'd rather see it
in the RFC822++. It doesn't make MTAs any simpler or more robust (they
don't display anything: no device to break), so I wouldn't like this
kind of restrictions in the SMTP spec. Possibly there could be a note in
RFC822 telling that sending one of these characters may cause trouble
and isn't polite.


-- 
 Risto Kankkunen                   kankkune@cs.Helsinki.FI (Internet)
 Department of Computer Science    kankkunen@finuh          (Bitnet)
 University of Helsinki, Finland   ..!mcsun!uhecs!kankkune   (UUCP)