I have a copy of Microsoft's RTF specification in front of me.
I'd be interested to hear other people's opinions, but mine is
that it is only marginally adequate as a personal computer
document interchange format. I would be loathe to see it
introduced as any kind of Internet standard.
I agree wholeheartedly. Face it, Microsoft RTF was designed to
allow some portability of documents between systems which couldn't
handle the binary format. It doesn't have the flexibility of an
SGML or an MIF, nor the utility of a TeX or Scribe. There is nothing
(in the spec) to keep someone from including RTF as an application
subtype but I can't imagine it as an Internet standard.
On the other hand, there really exists no canonical document form which
enjoys widespread support between the academic and business communities
and Ed's point is well taken in that it doesn't make sense that writing
a specification for such should fall within the scope of the IETF. Last
summer I attended the AIM (Image Management, not AI in Medicine) conference
and was astounded when one of the leaders of the Image Management community
stated that the *bitmap image* of a printed page was the universal
electronic document storage format because "other technologies may come
and go but we'll
always be able to print bitmaps").
From a technical point of view his argument is garbage but from a practical
point it makes some sense. The Internet community has gotten a little more
sophisticated in that we take one step back and use Postscript although if
you have too little memory and the document contains too many bitmap fonts,
or if it was prepared with psdit and you don't have the prologue or... Whatever
the case, Postscript has become our sort of de facto standard for
formatted document interchange (some people still use TeX-dvi) and the
printer people are starting to support the Adobe specifications so maybe
that isn't so bad.
So what is the point? Well, Ed sort of hit on it. We need something! And
we aren't the best people to be developing it because practically everyone
is supporting some type of word processing application and the business/real
world is going to be some split of Word and Word Perfect and all the other
guys and as much as SGML might make sense to a lot of people, who actually
has an SGML compliant authoring and rendering tool that is affordable and
widely available to Internet users?
I mean we *are* the Internet. Shouldn't we develop, test, deploy, and use
something, first, before we decide to make it a standard. And if we can't
do that ('cause face it, who is gonna pay to have SGML authoring tools
developed for widespread Internet distribution), maybe we should pick
something which is widely supported by industry in products which are
competitively priced so that most of us can afford to use them. And the
problem is, there ain't nothing like that right now. And maybe there won't
be for awhile.
None of this is relevent to the MIME specification but it is important
to the issue of how we support application interoperability with the majority
of the desktop community. I don't have an answer but I can tell you that
after one year at the helm of the HIS department of 500-bed hospital we won
the technical battle for Internet standards in our network even as we were
losing the desktop war because none of what we could show on our workstations
had anything to do with what the desktop users were doing. We could send
multipart medical record information around the world if we wanted to (and
did), but I couldn't get it from my office to that of the desktop user
running Word Perfect office, cc:mail, or CE QuickMail without losing a whole
lot of functionality. We even have a DTD for this document but until WP comes
out with their SGML compliant editor I dunno how end users will edit or
author the damned thing.
We gotta work with (or through) the vendor community on this one. Maybe not on
the transport, but certainly on the content.
Sean McLinden