W3C home > Mailing lists > Public > w3c-dist-auth@w3.org > January to March 1999

RE: Delete, Move, and Copy for References (Yaron's Issue #9)

From: <ccjason@us.ibm.com>
Date: Fri, 26 Feb 1999 19:51:57 -0500
To: w3c-dist-auth@w3.org
Message-ID: <85256725.000483B0.00@D51MTA07.pok.ibm.com>

Yaron writes...
> What really scares me is a scenario where I have a directory filled with
> references but I'm using an RFC 2518 client...

Good observation...


> I would also invoke precedent here. In every system I have ever heard of
> that supports references (read: links) a COPY always copies the link not
the
> destination. I would be very hesitant to go against three decades of
> accumulated experience without a good reason.

I philosophically agree and that this is this is the way that I'd want it
to work when I use it...


But the other folks seem to have a good point also.  Let me toss a
hand-grenade that way though.  Not expecting a modification of one resource
to affect another is a universal problem for down-level clients.
Regardless of COPY.  If someone builds up a site with a newer tool and
creates (direct) references... and if someone with an older tool starts
working on the site... the person with the older client is going to
experience aliasing... which he and his tool don't expect.  And if we
design the spec to avoid the aliasing for downlevel tools... the user will
unknowingly do (less dangerous, but still unsatisfactory) things like
turning references back into resources.     A person with this older tool
and older expectactions is going to make a mess of the site one way or
another... regardless of COPY.   And a person with an old client, but new
expectations, will quickly upgrade tools...

So I'm sort of supporting Yaron here...  but...

And I'm also suggesting that perhaps (I don't know) that some of our
"downlevel-client" arguments might be red-herrings.   It's something to
recheck.   ...and if not a red herrings... perhaps (shudder) we should take
the philosophy of limiting the operations that a downlevel client can
perform in the presence of references... rather than trying to guess at the
least dangerous or least inconsistant behavior.   As a site administrator,
that's REALLY what I'd want the server to do.

Jason.
Received on Friday, 26 February 1999 19:49:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:43:49 GMT