Re: Backpointers

Jim Amsden wrote:

> Third, the ability to support backpointers is clearly server dependent. Some
> WebDAV servers will delegate to multiple backends, so the capability will also
> be resource dependent.

I agree that this is an important case; but, for most people, this does not
happen.  Today, and for the foreseeable future, most servers are not federations
of backends.  If I have a simple server, the fact that there exist other, more
complicated servers should not limit what I can do.  If I buy a server with a
monolithic DAV implementation, which chooses to implement backpointers, I would
prefer to find out that those backpointers are implemented in some standard way,
so that I can buy my client from anybody who implements the standard.  If I buy a
server with multiple backends, maybe the answer is that it just can't implement
backpointers.  (Even there, I expect it would be possible for the DAV frontend to
keep track of the MKREFs and DELETEs it forwards to the backends, and maintain the
backpointers separately.  Probably not the most efficient implementation, but it's
possible for those who want to spend the CPU.)

Mind you, as was discussed on the conference call this morning, there's nothing
saying that backpointers have to be part of the core spec; they could be split out
into yet another RFC.

--
/=====================================================================\
|John (Francis) Stracke    |My opinions are my own. |S/MIME supported |
|Software Retrophrenologist|==========================================|
|Netscape Comm. Corp.      |Sleep is for wimps--healthy, well-adjusted|
|francis@netscape.com      | wimps, but wimps nonetheless.            |
\=====================================================================/

Received on Tuesday, 6 October 1998 20:11:04 UTC