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

RE: drop QSD?

From: Babich, Alan <ABabich@filenet.com>
Date: Mon, 17 Apr 2000 14:23:33 -0700
Message-ID: <C3AF5E329E21D2119C4C00805F6FF58F0398E8F5@hq-expo2.filenet.com>
To: "'Jim Davis'" <jrd3@alum.mit.edu>, www-webdav-dasl@w3.org
I disagree that QSD is complex. Of course, complexity is relative.
For example, if you read the DMA spec. and then you read the QSD 
section in the DASL draft, you will remark on the extreme simplicity
of DASL QSD relative to what you have to do in DMA. You can make
up your own mind on how complex it seems to you. However, the
correct question to ask is, "Is QSD more complex than it has to
be, given what it tries to accomplish?" I don't think so.

I strongly disagree that no one will ever implement QSD. Anyone
who implements a quality general purpose query function (as opposed 
to a hardwired query function) needs QSD. How else can the
UI present pulldown lists of legal properties to query without
retrieving all the properties of every object? Note that the
exhaustive retrieval approach doesn't even work if it happens
to be the case that no object in the database currently has
a value for one or more supported properties. 

I am predicting that many implementations of general purpose query 
facilitities will eventually occur. Therefore, QSD is essential. 

There is great doubt in my mind that DASL will ever be approved 
without QSD, because there would be such an obvious hole in the 

Alan Babich

-----Original Message-----
From: Jim Davis [mailto:jrd3@alum.mit.edu]
Sent: Sunday, April 16, 2000 1:56 PM
To: www-webdav-dasl@w3.org
Subject: drop QSD?

I would like to propose that we drop Query Schema Discovery.  I do this
with great reluctance, in part because I put so much effort into the
design, but to be honest I do not believe anyone will ever implement it,
and it is complex.  

Does anyone object to this?  If you do so object, please state whether you
either HAVE implemented it, or have a firm commitment to do so, and whether
in a server or client.

I suppose I would be willing to be told that this feature should remain in
the spec, even though no one implemented it, but I would then also like to
have someone explain how we will get DASL to ever be accepted by the IETF
if there is no implementation of this portion of the protocol.
Received on Monday, 17 April 2000 17:25:57 UTC

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