ietf-822
[Top] [All Lists]

Re: Internet Message Bodies DRAFT revisions

1991-06-29 08:37:21
I agree with Mark that we can do everything in 7-bit.
Do it by quoted-printable, base64 or quoted-readable.
I prefer quoted-readable as this notation is readable to the
end-user without any new software.

However we have made a new RFC using 8-bit mail.
If people are using this they should not need to know if the
receiver is able to accept this mail as 8-bit mail.
Maybe the UA can do it, but that requires a very intelligent UA
which can get in contact with all the receipients' UAs ad somehow
check it. I doubt that such an intelligent UA will exist for quite
some time, and also I find such a solution quite inefficient as the
UA shall do contacting of the other UAs and do intricate testing, which
is duplicated in part at the MTA level when the messages are delivered.

I have been one of the foremost people on this list to advocate
interoperablity of the new and old worlds, so I do not understand
your words about "incompatible manners". Could you clarify?

8-bit MTAs are fundamentally incompatible with 7-bit MTAs, and entirely too
much time has been spent in trying to cover the butts of lazy programmers who
unilaterally violated standards in kludging in 8-bit character sets.

Mark keeps talking about programmers, but now only the "lazy" ones.
Maybe he is hinting at me - I did some quite elaborate work on
8 bit and other character sets. I do not think I am entitled to the
term "lazy" - the code written for this in all is about 100 Kb of
tables and C code.

Mark, my name is *not* "Rob Ullmann" and I do *not* advocate just
sending 8-bit mail out there.

Yes, I think we have spent a lot of time to come to the conclusion
that we need to do conversion between 8 and 7 bit, to make the
whole of internet interoperate. I hope we do not need that more
discussion... Well Atlanta is just one moth from now, and I hope we can
reach a conclusion there.

The standards are on the side of 7-bit.  It was programmer laziness that
created 8-bit MTAs without going through the standards process.

I have created an MTA capable of doing 7/8 bit translations
without information loss, and I have been active on this list for
a long time - I have written a draft RFC for my specs.
My MTA is by default 7-bit. You are not fair in your arguments, Mark.

Either 2022 or Quoted-Printable.  It's only at the transport level.  WHY are
you so hung up on the esthetics of the transport level?

Well, there is a draft RFC out there which specificially allows
8-bit mail.  It is not aesthetics, but base interoperability
of the internet that is my concern.

I think 7/8-bit transformations without loss can be made

Yes, very easily, by keeping 8-bits at the UA level only and using 7-bit
transport throughout.

Well, actually I do not care if we run 7-bit only on the transport 
level, but then we could just discard the 8-bit mail RFC.

Keld

<Prev in Thread] Current Thread [Next in Thread>