RE: draft-reyes-policy-core-ext-schema-01.txt
"Pana, Mircea" <[email protected]> Fri, 7 Mar 2003 17:05:37 -0600
| Newsgroups | gmane.ietf.policy |
|---|---|
| Message-ID | <A33EE5A81E634B488B099FD31F65196153CBD7@SRVOTEMAIL> |
> I question whether it is absolutely necessary to mark the several > components from PCLS as "OBSOLETE". I realize that PCIMe describes > these as "deprecated". But the "OBSOLETE" designation is harsh. It > implies that a Directory implementation may not simultaneously > support both PCLS and PCELS. I don't see any naming clashes between > PCLS and PCELS, so why mark the PCLS components as OBSOLETE? You are right, it is not *absolutely* necessary to mark the deprecated schema items as "OBSOLETE". The "OBSOLETE" mechanism has been chosen as a way to formalize the restriction for PCELS information trees not to (SHOULD NOT) contain any of the deprecated PCLS items. In other words, PCELS and PCLS information trees SHOULD be disjunctive. This restriction originates from the desire to avoid backward compatibility issues for PCELS-only compliant applications ;-) PCLS schema items marked as "OBSOLETE" in subschema subentries governing specific PCELS administrative areas indicate that instances of such schema items are to be ignored. Implementations that do not support subschema subentries and specific administrative areas may ignore the "OBSOLETE" statements in the schema definitions but should have these restrictions enforced by other means. Regards, Mircea.
(unnamed)
(application/ms-tnef, 2.6 KB) - not displayed