Re: Scope of DASL Charter

Saveen,

Yes DASL is mostly orthogonal to the WEBDAV protocol.  (Of course there are
also the connections of the origin of DASL) The document and property model
or at least parts of them are important to DASL.

There is also the issue of property discovery.  (Which fields are
appropriate to search on in this collection?)

DASL will provide a standardized syntax and semantics for searching
collections on the web.  The most obvious scenario of taking advantage of
this without needing the rest of WEBDAV is distributed search.

While distributed search applications exist, today the problems of
different searching capabilities, syntax, and ranking algorithms, makes
these applications either difficult to write or of dubious quality due to
things like incorrect merging of ranked results, different meanings
attached to the users original query, etc.  Standardized syntax and
semantics would solve the problem of translating the user's query into a
different form for each target, and provide for simpler processing of the
results.

--Rick

Saveen Reddy (Exchange) wrote:

> Rick,
>
>         Let me make sure I understand you correctly. Is your position
> that DASL is orthogonal to the WEBDAV protocol -- other than picking up
> the document and property model? If so, then I think I agree to some
> degree (maybe a large degree). However, I'm not clear on why this
> relationship unnecessarily narrows the scope. Could you elaborate on the
> scenarios that you think are interesting here but that the charter might
> consider out of scope? This would help clear it up for me.
>
> Thanks,
> -Saveen
>
> -----Original Message-----
> From: Rick Henderson [mailto:rickh@netscape.com]
> Sent: Monday, February 23, 1998 4:26 PM
> To: 'www-webdav-dasl@w3.org'
> Subject: Scope of DASL Charter
>
> It seems to me we are unnecessarily narrowing our scope.  Search is
> already a common thing while distributed authoring is a rarer more
> fancy thing.  It seems wrong to define a search protocol that only works
>
> on servers that have all that distributed authoring stuff.  DAV has lots
>
> of things in it that aren't at all necessary for searching. like
> PROPPATCH,
> MKCOL, ADDREF, DELREF, LOCKING.  There is no need
> to be able to do document management operations to do search.
>
> As part of the protocol spec we should probably define a subset of DAV
> that
> is necessary for DASL.  This might be only the underlying document
> model.
>
> --Rick
> *************************************************
> Rick Henderson            (Netscape)(650)937-3152
> rickh@netscape.com
> *************************************************



--
*************************************************
Rick Henderson            (Netscape)(650)937-3152
rickh@netscape.com
*************************************************

Received on Thursday, 26 February 1998 12:57:22 UTC