A signature match is neither necessary nor sufficient.
That's false. There are situations where it is both. And, yes, there are
situations where it is neither; but they don't prove this negative.
On the contrary, my statement is completely true, and you have said nothing to
demonstrate otherwise.
Signatures are not necessary, in that there are circumstances where if such
operations are allowed it a signature should not be required to perform these
operations. Netnews is the obvious example and proves the point.
Signatures are not sufficient, in that there are circumstances where even if
something is signed twenty times by all the angels in heaven cancellation,
obsoleting, or whatever should not and cannot be allowed.
The fact that there are also circumstances where signatures are both necessary
and sufficient is irrelevant -- all I had to do is provide one counterexample
to prove the negative, which is exactly what I've done (twice now).
Cancellation of messages in transit is absolutely unacceptable
False again. It's sometimes acceptable and sometimes not.
Well, in my opinion it is unacceptable, period.
You are happy to point out situations and systems where cancellation would
be bad or difficult; great--important thing to do. Just don't deny that
there are other situations and systems where it would be good and easy.
I have yet to be convinced that there is a single, solitary case where
cancellation in transit would prove useful. And it is easy to come up with
cases where its use isn't acceptable.
Given the lack of an upside and over-abundant downside, I stand by what I've
already said.
MIME defined an 8bit CTE, even though there were TONS of systems where it
can't be used. Same thing here.
What on earth does this have to do with anything? 8bit MIME has a considerable
upside, and, properly negotiated, no downside. I cannot imagine anything more
dissimilar.
It may be that, after some discussion, we will agree that there are so many
inapplicable systems and situations that the whole idea is just not worth
the trouble; but one or two examples don't make me believe that.
Bzzt! You are now distorting my conclusions in order to justify your position.
There are of course cases where obsoleting messages is quite useful and
beneficial. I never said there weren't. However, this doesn't mean that these
operations need to be part of the transport infrastructure. This is what I was
objecting to.
It isn't a question of one or two negative examples proving anything. It is the
complete absence of positive examples showing why transport support for this
would be at all useful, especially given that support in the recipient's
mailbox is a requirement.
Ned