[Date Prev][Date Next]
[Chronological]
[Thread]
[Top]
Re: [ldapext] Fwd: I-D ACTION:draft-sermersheim-ldap-subordinate-scope-00.txt
Jim Sermersheim wrote:
It's a valid concern.
I assume different client SDKs will need to be updated to varying
degrees to support a new search scope. Some may just allow an int to
pass through, while others will require a new enumerated type (and thus
require that the SDK be revised). Something similar may have happened
(though likely to a much less degree) when the '+' all operational
attributes selection string was introduced.
I suspect many SDKs will need to be updated to support the new search
scope. For example, the Netscape/Mozilla SDKs validate the scope
parameter and therefore won't pass a new value through.
I personally feel that allowing there to be a ripple effect on client
SDKs is actually healthy. Client SDKs should be flexible enough to adapt
to future extensions regardless of the form of extension. Meaning, I
would prefer a world where the protocol may be extended without having
to worry about the impact to current client SDK implementations.
I agree, within reasonable limitations. A bigger question is: do we
need a new protocol feature to be available to people who are using
existing/older SDK implementations? To me, this is more of a "how quick
will people adopt the feature and be able to use it" concern than a "we
are going to break something" concern. But I may be overlooking something.
--
Mark Smith
LDAP Book Information: http://www.ldapbook.com/
What's Next: http://www.pearlcrescent.com/
_______________________________________________
Ldapext mailing list
Ldapext@ietf.org
https://www1.ietf.org/mailman/listinfo/ldapext