W3C home > Mailing lists > Public > www-webdav-dasl@w3.org > October to December 1999

Re: DASL and ordered collections

From: Jim Davis <jrd3@alum.mit.edu>
Date: Thu, 02 Dec 1999 13:26:19 +0100
Message-Id: <4.1.19991202131040.039cbdd0@pop.xs4all.nl>
To: www-webdav-dasl@w3.org
At 02:07 PM 12/1/99 -0800, Jim Whitehead wrote:
>During the advanced collections teleconference today, Kevin Wiggen raised
>the issue of what the results look like when a DASL query, without an
>orderby clause, is submitted to an ordered collection.  The consensus, at
>least among the people attending the teleconference, was that in this case,
>the DASL response should return the result set using the collection's
>ordering.

I do not think this makes sense.

1. if the client wanted ordering, they would have asked for it.  Unlike
ordinary WebDAV clients, a DASL client certainly knows about ordering.  it
may not know about AdvColl ordering, but it does know about ordering in the
abstract.  so absence of an orderby clause indicates that the client does
not want ordering.

2. I predict that the scope of a DASL search will often be an entire site
(or the root collection).  i dont think AdvColl style ordering makes sense
for the full recursive set of members of a collection, but only for the
level 1 members.

3. it may be expensive to compute.  a typical implementation of DASL might
use a RDBMS to store the properties of resources.  The list of rows that
match the query may be returned in a random order by the RDBMS.  the DASL
search arbiter would then have to sort.

4. I predict that Adv Coll orderings will mostly make sense in the context
of having the entire membership of a collection.  e.g. when you have a set
of pages you want to print in order.  I don't think they will always make
sense in the context of a random selection of a subset of the collection
according to some search criteria.  So I doubt the client would care to get
this ordering.
Received on Thursday, 2 December 1999 07:28:58 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 22 March 2009 03:38:05 GMT