W3C home > Mailing lists > Public > www-webdav-dasl@w3.org > April to June 2000

RE: JW2a, JW2b: Search Arbiter resource

From: Kevin Wiggen <wiggs@xythos.com>
Date: Wed, 19 Apr 2000 19:20:34 -0700
To: Jim Whitehead <ejw@ics.uci.edu>, "'DASL'" <www-webdav-dasl@w3.org>
Message-id: <ONEOJMKKAIDAGPLOPJEDMEBACDAA.wiggs@xythos.com>

I definitely think that a new resource type should not be created.  On
Xythos, the ENTIRE Webdav server responds correctly to SEARCH methods.  I
don't want to have to respond in every request that the resource is a search
arbiter also (overkill and very wordy).

Thus I agree with Jim.

By the way, where is the issues list?  I will go through it also.


-----Original Message-----
From: www-webdav-dasl-request@w3.org
[mailto:www-webdav-dasl-request@w3.org]On Behalf Of Jim Whitehead
Sent: Wednesday, April 19, 2000 1:53 PM
To: 'DASL'
Subject: RE: JW2a, JW2b: Search Arbiter resource

I'm returning to a post I made on August 16, 1999
to see if we can start closing down some of the open issues in DASL.  The
crux of this issue is whether the DASL specification is defining a new type
of resource, the search arbiter.  After some discussion, Yaron and I agreed
that a search arbiter does not necessarily need to be DAV compliant, it just
needs to support the SEARCH method.

I recommended:

> A search arbiter is any resource that supports SEARCH.  A
> search arbiter is not a new resource type.
> In general, if a resource is a search arbiter, no conclusions
> can be made concerning what other methods it supports.

Yaron suggested, in his response, that a search arbiter could be a new
resource type, and that DAV resources could have multiple inheritance.  That
is, the DAV:resourcetype property could responde with DAV:searcharbiter, and
potentially some other value as well.  I tend to think that adding in
multiple inheritance is overkill for this situation, especially since there
is no strong need to report DAV:searcharbiter inside DAV:resourcetype (a
client can discover that the resource is a search arbiter by performing an
OPTIONS request).

So my recommendation still stands: a search arbiter should just be any
resource that supports SEARCH, and it is not a new resource type.

- Jim
Received on Wednesday, 19 April 2000 22:26:28 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:22:41 UTC