FYI, several existing proprietary mail systems use private labelling
methodologies in RFC822 messages rather than MIME. The nature of these systems
is such that they obviously don't always work, but they do work much of the
time. (GTW, even when they don't work, the result is fairly harmless. The same
thing cannot be said of auto-detect failures on printers...)
And guess what? Some of these were part of the original MIME specification.
Heck, at least one of them was presented as an alterative of MIME.
But there isn't a trace of any of them in any standards-track IETF document
right now. There are some informational and experimental RFCs but that is all.
Why is this the case? Because had they been included the specification that
included them would have never made it to draft standard.
> A scenario where this is very helpful, is if the user gets a file
> of some unknown file type e.g. as attachment in a message or found on a web
> page and wants to attempt printing it. In another scenario, a user might
> know that a file is Postscript, but not whether it is Postscript 1, 2 or 3.
> The auto-sensing can often help sort that out once the file arrives at the
> printer.
I'm well aware that auto-detect mechanisms are widely used today -- in fact
I've developed print spooling software myself that does this. And I'm well
aware that that this can work -- most of the time, that is. But unless I'm
mistakem, this WG is trying to develop a standard. And when you build a
standard, most of the time just isn't good enough.
I will also reiterate that I see no problem with having an informational
supplement that describes this approach.
> We obviously do not want to make IPP less powerful than existing products
> in this respect. It seems to me though, that we should define our own MIME
> type for a document file that we want to have auto-sensed.
On the contrary, I think you absolutely do want to, at least insofar as
standards track documents are concerned. If you want to have an informational
document describing this, that's fine. But
In any case, I've said my piece. If this WG wants to include this sort of
thing in their documents, that's fine. But don't be surprised when objections
are raised that cause the documents to get tossed back at you for revision
later on.
Ned