From: Stephen D. Williams (sdw@insta.com)
Date: Sat Mar 18 2000 - 12:03:16 PST
Jeff Bone wrote:
> ...
> The point is, despite any religion and / or ulterior motives driving particular
> IETF or other efforts, we've *got* the technology necessary to build open,
> interoperable applications of arbitrary kinds. The standardization of things at
> the *application* level should not --- and does not, except by the poor choices
> of application developers --- necessitate having new protocols for everything.
>
> The punchline to all of this is, I actually think DAV is rather an abortion.
> Ick. The more crap like that we cram into our star protocol, the less tenable
> it becomes.
>
> This kind of stuff should live above the wire, indeed, above HTTP. If all
> distributed systems can be implemented in terms of RPC, and HTTP provides a
RPC, in the sense the term is normally used, is NOT sufficient to implement all
distributed systems. Certainly not efficient and competitively scalable
distributed systems.
> *standard* and rather friendly RPC just with GET, and XML provides the
> marshalling format, then why re-invent the wheel? It's purely an ego and
> competitive game. The only good argument --- performance, etc. --- isn't too
> convincing to me these days. Simplicity and speed of implementation trumps.
For some segments of the market this is true, for others it is not.
While I believe that a good generic protocol is possible, it is going to need a lot
more than what you have suggested. In particular, it must be streaming message
based with windowing, flow control, reconnection/restart/fragmentation/partials,
will full support for pipelining and, when it makes sense, out of order processing.
Demands for near optimum data representation are also more involved. Logically,
XML, but optimized.
sdw
> jb
>
> Jim Whitehead wrote:
>
> > Hi Ernie,
> >
> > > He mentioned how Microsoft was using WebDAV as an alternative to POP/IMAP.
> >
> > Yep, just sniff the packets between Outlook Express 5 and Hotmail.
> >
> > We've finally come full circle. In 1968, Doug Engelbart demoed the NLS
> > system at the Fall Joint Computer Conference in San Francisco. NLS provided
> > email, outline processing, document editing, and the ability to hyperlink
> > between them. Email and documents were treated in almost identical ways.
> > In NLS it was not, go to the email program for email, go to the documents
> > program for documents. It was a fully integrated environment. Now,
> > finally, by using the same protocol to access email and documents, we're
> > starting to move back towards a more holistic view of information
> > processing. Theoretically it didn't require a new protocol, but since POP
> > and IMAP have led to programs specifically for email, it was helpful to
> > introduce a new protocol that really blurs the distinction between email and
> > all other documents.
> >
> > The same is true for Calendar data. I can see very little benefit to
> > defining a specific protocol just for use with calendar data, such as the
> > current Calendar Access Protocol (CAP) effort. Every new MIME type does not
> > need a new protocol. It makes sense to have hypertext links between
> > calendars, and email, and Web pages, and spreadsheets, etc., and using a
> > single protocol to access and author all of these makes this more clear.
> > You could have a separate protocol for each MIME type, but what's the point?
> > Unless there is a compelling advantage, it is just a huge duplication of
> > effort.
> >
> > > We thought it
> > > would be really cool to document some standard way of structuring mail
> > > messages as WebDAV properties, so everyone could do this. And we could
> > > maybe even get Microsoft involved, too.
> >
> > I would be very interested in seeing a standard developed here.
> >
> > >
> > > Do you know of anyone who'd be interested in coordinating something like
> > > this?
> >
> > Not off the top of my head. Undoubtedly Microsoft would want to get
> > involved to ensure the final design was close to what they have shipped.
> > Probably the best thing to do would be to form a mailing list, hold one or
> > two face to face meetings, and then organize a BOF session at an upcoming
> > IETF meeting.
> >
> > Such a BOF meeting would unleash a firestorm. There is very strong
> > sentiment in the IETF for having specific protocols for accessing mail
> > stores. POP and IMAP have strong followers, and there are many in the IETF
> > who feel that each domain should have a separate protocol, that one size
> > does not fit all. Still, I think it is a path worth taking.
> >
> > - Jim
-- Insta.com - Revolutionary E-Business Communication sdw@insta.com Stephen D. Williams Senior Consultant/Architect http://sdw.st 43392 Wayside Cir,Ashburn,VA 20147-4622 703-724-0118W 703-995-0407Fax Jan2000
This archive was generated by hypermail 2b29 : Sat Mar 18 2000 - 12:08:32 PST