An Overview and Evaluation Re: Access Control

[Einar Stefferud <Stef@xxxxxxx>]

I too was at the ASID WG meeting in Montreal, and I will agree with
Mark that the discussions tended to be intense, but not out of
character with a recognized productive WG effort.  You should have
seen the MHTML WG.  Actually, I think that the ability of IETF WGs to
get down to it and come to consensus is very powerful and a great
asset.  Not unlike this discussion via EMail;-)...

My take is that everyone there (and here) see this as a major
breakthrough in several critical areas in need of progress, with a
great deal at stake in getting it right for the Internet.  

If this work does not yield regular and easy Application Information
Object Interchange protocols for all kinds of application objects, we
will have missed the boat in a big big way.  So, with so much at
stake, how could the discussions be anything but intense.

As I read the excerpt from the minutes, I see that the issues listed
are terribly critical for a whole variety of new developments,
including how we get EMail and WEB applications to smoothly and
regularly exchange information objects wihtout damage or ambiguity.

The handling of CR, LF, and CRLF are a major issue related to a
massive installed base of EMail systems which can only evolve slowly.
I am eagerly awaiting Dave Crockers two deliverables as listed, since
they promise to resolve the CRLF issues for lots of new protocols.

Another issue is the so called choice of MIME CHARSET "default".  My
take was that ASID decided, and HTTP decided to no longer use
"absence" to imply any default, as different transports have taken to
using different semantics for "absence" of CHARSET parameters.  MHTML
is now seriously considering that CHARSET MUST be present in all
Content-type: text/mhtml MIME parts.  

To fill all this out, I recall a decision somewhere, perhaps in DRUMS
to make CHARSET a MUST, and to define "UNKNOWN" as the value to use
when the content charset is in fact unknown.  With this done, we no
longer have ambiguity born of the attempt to use "absence" to mean
different things in different transport contexts.

That attempt was born of good will, to deal with the fact that lots of
mail was being generated without MIME-version: 1.0, or any other
MIMEisms, and all such mail is by (historical/hysterical) RFC822
definition, CHARSET=US-ASCII.  So much for history -- We can cinch
this down now by all agreeing that CHARSET no longer has ANY defaults
in new MIME types.

So, what I see out of Montreal was an amazing lot of consensus for a
lot of issues that have been plaguing some of us for a long long time.

That is my view from about 15,000 feet;-)...\Cheers...\Stef

>From your message 22 Aug 96 18:34:18 :
}
}Dr. Mark K. Joseph wrote:
}> 
}> I would also agree with this.  However, I remember the reception that
}> the ASID MIME-DIR draft received in Montreal.  It was unfavorable ?  I
}> was in the meeting that got pretty hot discussing how the basic vCard
}> format (similar to vCalendar format) that was being transported in
}> the ASID MIME-DIR structure was all wrong.  So if we proceed with ASID
}> MIME-DIR those concerns brought up in Montreal should be addressed up front.
}
}Mark: 
}
}Your comments are VERY misleading. I certainly would not have thought that that 
}is your intent here. I was also at the Montreal meeting and participated in the 
}discussions and action items out of the meeting. As the minutes will show, the 
}vCard profile work was addressed and progressed at the meeting. There were a 
}short list of comments that Patrik had on the Internet-Draft. They were 
}resolved to his satisfaction in the meeting with revisions to the 
}specification. 
}
}Not sure how this can be more clearly expressed.
}
}- - Frank Dawson