Re: Update of the MIME-MHS Specs

[Ned Freed <[e-mail]>]

> Actually, most of them are not, and for good reason: they need to be defined
> by the people who own them, e.g., Microsoft, Word Perfect, et al.  But that's
> going to be slow in coming -- see below.

It would be nice if the authorities responsible for the types would define 
them. However, if they don't they'll get defined one way or another. Even if
the IETF refuses to allow official registration people will define them. Count
on it.

> > If it is already happening then why does it need to become more complicated
> > by dragging in another organization into the picture?

> Because it's a big win for the Internet and MIME.

Agreed.

> It clearly better serves E-Mail end users if X.400 BP15 and MIME recognize a
> similar set of application types.  I'd hope there is no disagreement here.
> Filters, user interfaces, and gateways all become simpler to implement and
> manage.

Agreed.

> This will happen faster and better with EMA and IETF working together.  No, I
> am not kidding.  Like the Internet, the EMA is very end-user oriented, and has
> a good reputation of getting things done.

I agree that it would be nice for the EMA and IETF to work on this together.

> But the end-user base served by
> EMA is completely different from that of IETF:  business, software houses, and
> large corporations.

Actually, this is exactly the same end-user base that's served by the IETF.

> Most of these concerns are both excited and frightened by
> the Internet.  They want to get involved, are willing to spend research money,
> are willing to make technology available, but they don't know how or where.
> EMA and IETF working together would greatly boost the credibility of IETF with
> the business community, making them far more willing to get behind MIME.

I don't think this is valid. For starters, the business community is involved
with the IETF in a big way already. The most recent example of this was the EDI
over MIME working group meeting in Seattle. The room was packed to standing
room only with representatives from major corporate EDI users, who seemed
collectively willing if not anxious to buy into MIME as a form of EDI
transport, if only the IETF would get the standards for doing it in place.

Second, I don't see the technical side of EMA that's doing the FTBP work as a
useful means of connecting with business users. It is, however, a useful means
of connecting with service providers and developers in other parts of the
electronic messaging community. This is something that the IETF has not been
particularly successful in doing.

> Try listing all the companies with more than 200 employees that have made a
> committment to MIME.

Am I infer this to mean that you think this will be a short list? If so, you're
dead wrong. The degree of comittment to MIME at many large corporations is
already substantial and increases daily. This is precisely our customer base
which is constantly growing, so I know what I'm talking about here.

In fact, you have it backwards. The problem is the companies with less than 200
employees. This accounts for a majority of businesses and revenue, and I quite
frankly don't believe any of us (IETF, EMA, Internet, etc.) are doing a very
good job getting to these people. (No, I don't have any bright ideas about how
to do it better either.)

> *All* the companies, not just software developers.  You
> can put IBM right at top, of course, but where is everyone else?

Buying into MIME in a major way, actually.

> They are
> waiting for something to happen, that's where.  EMA end IETF working together
> could be that something.

Again, I may disagree with your reasoning but I still think this is an
excellent idea. It certainly cannot hurt.

> Need specifics?  At least one Very Big Software Company has been emphatic that
> they will register no types for anything -- nor allow anyone else to register
> types for them -- until they can do it once for both X.400 and MIME.  If they
> can do it only for one, it will be X.400, but that is considered a last
> resort.

Well, now you've switched from talking about companies in general who use email
to companies who write email software. There's a very big difference here!

> There are forces in that company that could switch the focus from
> X.400 to MIME, but since X.400 was there first, the MIME camp has to
> demonstrate that they can do the job better.  The credibility gained by the
> IETF and EMA joining forces would be a very strong push for them.

I certainly agree that the EMA and IETF working together would be a plus.

> Does Word Perfect Corporation know that someone is defining MIME content types
> for their data formats?  Word Perfect's prorprietary E-mail system preserves
> external data references and OLEs, but the current MIME proposal does not.  So
> what is Word Perfect going to do when they implement their own MIME gateway?
> Why, something proprietary and incompatible, of course.

> We've *got* to get the different camps working together, or nothing is going
> to interoperate.

Actually, this brings up an important tangential point. One of the things MIME
has done is markedly lessen the control that either the IETF, EMA, or any other
organization has over the whole organic process that's now taking place.

MIME software is by nature table-driven. Nowhere does the MIME specification
say it has to be this way, but starting with Metamail this is just how things
have been done. In contrast, previous solutions tended to be hard-coded to
handle some specific set of situations in a specific way.

This has profound consequences. Let's say the random-giant-software-company
refuses to register MIME types for their very popular document formats because,
say, they don't like silent people in whiteface (its a company policy or
something).

Fine. So what? In the final analysis, end-users just don't care that much. They
add an entry to their tables, and presto, some nonstandard MIME content type
pops out. A message reaches some other site and isn't handled properly. No
problem -- the administrators there just add the same type to their tables.

Conflicts may arise (although they are far less likely than you might think).
Fine. Businesses understand conflicts and deal with them every day. They'll
work it out somehow.

The net result is that organizations like the IETF and EMA, to say nothing of
the random-giant-software-company, are smoking something if they think they can
say yay or nay to this process. Maybe they had this level of control with X.400
(much less than they thought, if my experience with nonstandard X.400isms in
any indication), but MIME blows almost all of it away.

Mind you, I'm not saying that this outcome is a good one. It isn't. There will
be problems if this approach continues. I try to stick to the party line, be it
EMA, IETF, OIW, or whatever. But the customer is always right, and besides, I
can't control what they put in my tables anyway!

This is why I think this work is of paramount importance. We have a small
window during which we can exercise some control over this ongoing process.
Miss it and the process is going to control us.

				Ned