Re: [Editorial Errata Reported] RFC5322 (3400)

> Ned Freed <[email protected]> wrote:
> >
> > The text in RFC 3030 doesn't explcicitly say "all lines MUST be
> > terminated". Rather, it's saying that when terminators (it actually uses
> > delimiter in another place) are used they have to be CRLFs.

> That isn't how I read it:

>   "In particular, it is essential that text be canonically encoded with
>    each line properly terminated with <CR><LF>."

The word "each" gives some weight to your argument. But not a lot - it would
have to be "every" to actually impose the restriction.

>   "In particular, text/* MUST be sent with <CR><LF> terminated lines."

This says nothing about unterminated lines.

> I agree with your interpretation of the intent of RFC 3030 but the way it
> expresses that intent is not quite correct.

Well, the bigger problem is that the entire style of the RFC 3030 prose is
obsolete. We no longer mess around with discussions of local storage formats
and how to convert to and from them. We now focus on what's on the wire and
leave the implementation details to implementors.

MIME has some of this too, but not as bad.

In hindsight we should not have included this sort of language, but it seemed
like a good idea at the time.

				Ned
_______________________________________________
ietf-822 mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/ietf-822