In particular, the actions
associated with the conversion of RFC822 messages to Grey Book format
very much depend on comprehensive knowledge of headers and are
therefore the actions of a gateway, not a simple MTA.
Agreed. (I think I have been using the term "gateway" too.)
There are various ways this could be done:
(1) The formats are close enough that MIME could probably be adopted with
very few changes.
(2) Some type of encapsulation could be defined for various multimedia
formats.
(3) Something completely new and different could be invented.
The Grey Book user that I talked to (Mr. X) seemed to think that
simply extracting MIME headers from the body and putting them in the
real headers would be sufficient. Disregarding for the moment the
reversed addresses (e.g. uk.ac.oxford and uk.sun.com :-) and such,
MIME and Grey-Book-MIME would be "identical" except that the top-level
MIME headers would be placed in the body.
Mr. X seemed not to know of any "multimedia" format for the Grey Book
world. That does not mean that there *are* no formats for doing
things like multipart, content typing, etc. (He may not be aware.)
So, it seems to me that it would be in everyone's interests to start a
dialogue between the Internet and Grey Book network communities.
But that's not my main point. My point is that we should not only
start a dialogue with the Grey Book people, but also with any other
"enclaves" that might have problems with putting MIME headers in the
top-level headers. Why? Because we want to make sure that *if* the
internal formats of the various enclaves ever spill out of those
enclaves (quite likely, since there are often multiple Internet
gateways on the borders, and you cannot expect flag-day switch-over of
those gateways), then at least we on the Internet side don't have to
deal with *multiple* not-quite-MIME formats.
This assumes that there actually *are* multiple enclaves like the Grey
Book world, and that 2 or more of them actually intend to "do" MIME by
simply placing the top-level headers in the body itself, being
otherwise identical. Of course, I do not *know* of any such enclaves
(otherwise I would have mentioned them!), so if you are *certain* that
there are no other enclaves, then the problem is simplified, though
not solved.
I don't have a problem with coming up with solutions to problems. I do
have a problem with the imposition of these solutions on other
communities.
I agree. I'm sorry that my message wasn't clear.
In the case of Grey Book, I'm not sure we have enough
knowledge to even make a reasonable recommendation; there are simply
too many choices.
Are you saying that they already have a way of doing e.g. multipart?
Just to make sure that my intentions are clear, basically what I'm
saying is that when you design something it is always a good idea to
try to anticipate problems. We have already seen multiple proposals
and/or working solutions to the problem (the Andrew way, the stuff
Harald mentioned, my proposal and maybe others). Shouldn't we try to
harmonize these?
Actually, the solution doesn't need to be limited to MIME headers.
The solution could easily cover whatever headers people want to, but
cannot, include. In that sense, it may be better not to put this
proposal in MIME, but in a separate RFC. (Perhaps such an RFC already
exists? :-)
Erik