Re: RFC2253bis: "a published table"

["Kurt D. Zeilenga" <Kurt@OpenLDAP.org>]

I have one additional point I'd like to make.

The table can be viewed as strictly limiting what DNs an
implementation can produce.  It doesn't not limit what DNs
a implementation might liberally recognize.

Kurt

At 12:46 PM 11/1/00 -0800, Kurt D. Zeilenga wrote:
>At 11:35 AM 11/1/00 -0700, Jim Sermersheim wrote:
>>I like the removal of the "example" language, but this also stops any future attribute name from being used.
>>If, for example, I have an attribute called SSN and I want to use it as a naming attribute, there is no way for me to get it into the table so that people don't have to use its OID form in string representations of DNs.
>
>Given the problems associated with depending BER DN-> LDAP DN string
>generation upon the controlling schema or other DSA-specific (or shared)
>information, restricting the use of attributes names to those in "a
>published table" is sound and should not be changed.
>
>Given that this table likely needs to be embedded in implementations,
>a static table is a very good thing.  If a implementation where to
>use a different table than another implementation, I believe
>significant interoperability problems would arise.
>
>Kurt