CURRENT MEETING REPORT
Reported by Peter Resnick, Qualcomm, Inc.
Minutes of the MIME Content for SGML Documents
Working Group (MIMESGML)
The original agenda was suspended. It had
been to discuss, in turn, what it took to make each of the two
proposals, encap and exch, workable. As a proponent of only one
of the proposals was present, it did not appear to be a useful
exercise. Instead, the working group reviewed the current status
and activity.
Terry Allen summarized the recent mailing
activity and noted the participation of James Clark. Mr. Clark's
participation served to recenter the discussion. Some progress
is being made now. Mr. Allen noted that we sometimes find the
discussion in the collision zone between IETF and SGML.
Ed Levinson reported on some off-list work.
He requested that James Clark and Don Stinchfield cooperate to
resolve the issues that Mr. Clark identified within the exch proposal.
These relate to features in encap but not to any of those in exch.
Mr. Levinson has gotten a response from Mr. Stinchfield and is
waiting to hear from Mr. Clark. Charles Goldfarb has offerred
to participate in that effort. John Klensin requested Ed Levinson
contact Paul Grosso, Technical Chair of SGML Open, to have SGML
Open take on developing an agreement for the content of the extend
catalog contents being proposed in the exch proposal. Mr. Levinson
agreed to do so.
The possibility was raised to rewrite exch
to be independent of SGML Open Catalogs. Reservations were expressed,
particularly in relation to not having the IETF act as a party
in actually defining SGML. Instead, we should work toward getting
the SGML Open folks to move on this.
The problem was raised of dealing with references
to [the sender's] local files that are not included in the message.
Two proposals were suggested: changing those local references
to global (public identifiers) or indicating via a parameter that
the document is "clean," i.e., has no local references.
The use of Multipart/Mixed in the current
exch draft was discussed. John Klensin said he had pushed for
that initially, but that it might not turn out to be a good thing.
Current mailers, he said, "flatten" Multipart/Mixed.
Multipart/Related will go to standards track when an application
that uses it goes standards track; the same is true for the content-id
URL draft.
Clark Anderson discussed the application of
distributed catalogs to whois++. Maybe transport can be done with
MIME or an index server could solve the problem. Mr. Anderson
asked if this was applicable to the problem. Mr. Klensin observed
that this may be an external catalog.
The work of the HTML in email group was explored. That work was considered to be a special case and not generalizable.