[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: New Phonetic Design

At 12:47 PM 9/22/2004, Howard Chu wrote:
>>I don't think it appropriate for a matching rule's behavior
>>to depend on the attribute description associated with the
>>value (assuming there is an attribute description associated
>>with the value), especially in situation where one description
>>could be a subtype of another description.  I think that would
>>be pushing the X.500 directory information model beyond its
>>established bounds.
>But approximate matching is implementation-defined anyway, so I don't see that X.500 rules this out. 

The comparison involve an assertion value and the attribute value
in an implementation-defined manner.  However, regardless
of what that implementation-defined manner might be, in
should not involve other information such as the attribute
description(s) which may be associated with the values.

Though this may not be explicitly stated in X.500 specifications,
I believe it to be intrinsic in the X.500 model.