[Prev][Next][Index][Thread]

Draft Minutes from IETF WEBDAV BOF



I finally have enough time to report on the activities of the WEBDAV group
over the month of December.  Here is my first pass at minutes from the IETF
meeting.  Please send comments/corrections on these minutes to Jim
Whitehead <ejw@ics.uci.edu>.

- Jim

----------------

WWW Distributed Authoring and Versioning (WEBDAV) BOF Meeting Minutes
San Jose IETF Meeting
December 11, 1996, Reported by Jim Whitehead (U.C. Irvine)

Jim Whitehead, U.C. Irvine chaired the meeting.

The BOF on WWW Distributed Authoring and Versioning (WEBDAV) was held on
Wednesday, December 11, 1996, from 09:00-11:30, with 79 people in
attendance over the duration of the meeting.  Jim Whitehead presented an
overview and history of WWW Distributed Authoring and Versioning, including
information on how to become involved in the work of the DAV group, which
was followed by a detailed presentation on the current Distributed
Authoring and Versioning (DAV) functional requirements.

During the presentation of the DAV functional requirements there was lively
discussion on many points:

LOCKS.

Meeting participants expressed the opinion that remote lock administration
functionality is needed.  At present, there seems to be an assumption that
the webmaster will perform this lock administration role, which could
become onerous if there are many active locks.  There was some discussion
on whether lock timeouts would help address the lock administration
problem.  Several comments addressed the need for some authentication
technology to be used to identify and validate the user or agent operating
on a lock.  Existing HTTP authentication schemes (basic, MD5) were
discussed in this context, and there was a suggestion that public key
technology should be investigated for this use.  There were some questions
concerning what error codes and conditions might result from locking
activities.  Several participants stated that it would be helpful to have a
model of access control which matches the locking model.  It was observed
that the proposed "read lock" was really a convenience access control
function to modify the read permissions on a resource, and that a read lock
might not really be a lock.

MOVE.

There was a thread of discussion concerning how intelligent an operation
"move" should be.  Some participants felt that HTTP should recognize the
special nature of HTML media types, and automatically perform some link
maintenance on links which are broken by a move.  This intelligent move
operation was referred to as a "magic move."  Three link cases were
identified: 1) links to a resource, 2) links from a resource, and 3) links
within a resource.  While it is impossible to correct all links pointing to
a resource (since in general you do not know or have control over all of
the links pointing to you), it is conceivable that links from a resource,
and links within a resource might be corrected.  Also, it might be possible
to correct all links in documents that you do have control over (such as
the documents on a single server).  However, some participants expressed
serious concerns over allowing a move operation to have side-effects which
could modify the resource being moved, or even modify resources which were
not being moved.  These participants stated that what was most important
was being able to know exactly what the consequences of an operation would
be before executing the operation. As a result, each DAV operation must
have precisely defined semantics. Finally, it was noted that this
discussion had some similarity to the symbol table and link table issues
which occur in compiler/linker development.

MISCELLANEOUS.

There was a recommendation that the DAV protocol be supportive of links to
a discussion forum, so that remotely authored documents could be embedded
in an issue discussion space, or a rationale space.   One participant
recommended that even if the DAV group is not planning on developing
DAV-specific authentication or security technologies, that we should still
develop requirements by which authentication or security technologies can
be assessed for their usefulness to DAV activity.  It would also be helpful
to develop the hooks necessary to use some existing authentication and
security technologies.  There were recommendations of
related protocols and systems to investigate: Andrew File System (AFS),
Internet Message Access Protocol (IMAP), and Application Configuration
Access Protocol (ACAP) were recommended to be examined for their models of
access control.

FORM WORKING GROUP.

At the end of the session, the attendees were polled on whether they
thought it would be worthwhile for the IETF to work on WWW Distributed
Authoring and Versioning. Overwhelmingly, the attendees thought the IETF
should pursue work in this area (the vast majority were in favor, with 2-3
opposed, and a handful of abstentions).

*** Meeting Adjourned ***



Follow-Ups: