Re: Tw bobs worth on TOP, LDAP standards process and subschemasub entry attribute usage in rootDSE

[Robert Allen <robert.allen@eng.Sun.COM>]

	FWIW, I've hesitated to enter into this discussion for fear
of adding more heat than light, since LDAP and schema in particular
seem prone to such in discussion, but I believe Alan and others have
brought up a point worth discussing, or more accurately, coming to
closure on:

Mark Wahl wrote (in part):

>>The IETF standards process does not enforce conformance.  The IETF does not 
>>operate a branding program, have an enforcement arm, trademark or license the 
>>elements of protocol.  The IETF does not tend to sue implementors when their
>>products are nonconformant.  Furthermore, an application protocol such as LDAP
>>would have either Recommended or Elective status: even FTP or SMTP do not have 
>>the Required status. 

	My comment here is: is it not the case that Naming and Directory
	Service protocols are a bit different than a File Transfer Protocol?
	I can see that you wouldn't need or want FTP if (say) FTAM works
	for you, but it's much less clear to me that a Directory which is
	deployed in order to facilitate storage and retrieval of human-named
	objects shouldn't be subjected to a significant amount of rigor
	in specifying what those names are, at least for common objects.
	
	Perhaps this would be less the case for a simple access protocol,
	but LDAP is being created as a complete Directory, and such naming
	issues would appear to be part of the Directory standard.



Andrew Probert wrote (in part):

>>To date, it appears that it is only Microsoft that is non-conformant.  A
>>clear message needs to be sent about this.

	And since MS is a very large player in the software business,
	the fact that it is `only one' of a number of LDAP implementors
	doesn't seem to be a very compelling argument towards avoiding
	resolution of this "top" issue, and future related schema issues.
>>
>>I think *some* people are emotively defending the right for IETF to diverge
>>from ISO/ITU, rather than objectively looking at the facts, the problem, and
>>the ramifications.
>>
>>I request that it is put to a *VOTE*, that the TOP used in LDAP conforms to
>>the ISO/ITU definition of TOP.  

	Precisely.  I believe a vote on this would be the appropriate thing
	at this point.  The only alternative would seem to be that everything
	in the world no longer inherit from a polluted "top", which would
	be very painful to do.
	
	In a larger sense I think that the level of protocol which people
	are now starting to address, that of the human-computer object
	naming interface, must follow at least some basic rules, one of which
	would be that the same name refers to the same object, in compliant
	implementations.  "top" is the proper noun name of a particular type
	of object.  If it doesn't always mean that, two disjoint groups of
	"people", in this case applications, can't have a useful conversation
	about that object, thus it becomes useless.  Naturally, replication
	of data objects is also very hard if two servers don't agree what the
	contents of a named object are.
	
	
Best regards,

Robert Allen
rja@Eng.Sun.COM