Re: WG Charter Proposal v2

Chris Apple wrote:
> Mark,
> 
> Sorry for being obtuse. You used the word extension in
> relation to the LDAP models.
> 
> "...an argument could be made that any form of 
> multimaster replication requires an extension of
> the LDAP models..."
> 
> I was looking for some clarification about what you
> could have meant by extension in this context.
> 
> The suggestion below seems quite reasonable for editors to
> consider during their final revisions of LDUP documents.

I meant "extension" in the most general sense of the word. I think 
Kurt's point is that if multimaster replication (MMR) is deployed, there 
are certain things that LDAP clients can no longer assume to be true. 
For example, a client can no longer assume that only one writeable copy 
of an entry exists. Because that knowledge may be really important to a 
client, LDUP should perhaps provide a way for clients to discover that 
MMR is in use. But I don't think most proprietary MMR implementations 
both providing such a discovery mechanism so one could argue this is not 
a big problem in practice.

Certainly it seems to me that experimental RFCs are just that: 
experimental. If problems are discovered by the brave implementers and 
deployers who adopt such as RFC, then the specifications can be revised 
or we can all conclude that the experiment was a failure. No one HAS to 
use any of this stuff.

-- 
Mark Smith
Netscape Directory Product Development
My words are my own, not my employer's.