Re: OSI-DS 40 [identifiers]

Mark Wahl <wahl@atlas.arc.nasa.gov> Fri, 19 March 1993 23:03 UTC

To: Tim Howes <tim@terminator.rs.itd.umich.edu>
cc: Colin Robbins <c.robbins@nexor.co.uk>
cc: Paul Barker <P.Barker@cs.ucl.ac.uk>
cc: osi-ds <osi-ds@cs.ucl.ac.uk>
cc: wahl@atlas.arc.nasa.gov
Subject: Re: OSI-DS 40 [identifiers]
In-reply-to: Your message of "Fri, 19 Mar 1993 09:32:23 EST." <9303191432.AA15695@terminator.rs.itd.umich.edu>
Date: Fri, 19 Mar 1993 14:14:46 -0800
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: Mark Wahl <wahl@atlas.arc.nasa.gov>
Message-ID: <9303191803.aa21878@CNRI.Reston.VA.US>

On the enumerated types Tim Howes says:
>new enumerated type attributes (for FileFormat and DocumentType).
>While it might not be as elegant in the X.500 sense, I think a
>more workable approach is to make these things CaseIgnoreStrings
>and include some text about what values are allowable.  This approach

I would suggest using an OID for these instead of enumerated or a
CaseIgnoreString. This is similar to the use of OID for encryption algorithms.
- more extensible than enumerated - anyone who owns a portion of the OID tree
  can define them
- allows 'bilateral agreements' for private types
- prevents collisions which would occur between CaseIgnoreStrings, and the
  related management problem
- 'well-known' OIDs (for text, PostScript, etc) would be in local OID table, 
  so that the DUA would interact with the user using simple names
- when a DUA encounters unknown OIDs, at least it is known who defined them,
  if not what they mean (with a CaseIgnoreString one doesn't know either)

		--------------------------------
	Mark Wahl,ARC,NASA,US; wahl@atlas.arc.nasa.gov; Sterling Software

OSI-DS 40 Colin Robbins
Re: OSI-DS 40 Tim Howes
Re: OSI-DS 40 [identifiers] Mark Wahl
Re: OSI-DS 40 Thomas Lenggenhager
OSI-DS 40 Paul Barker