RE: JW9: search arbiter locations

I think it would be a mistake to force DASL to deal with discoverability.
This is a general problem that is being dealt with in a number of groups at
the IETF. There is no reason to drag DASL into this mess.

I could imagine, for example, discovery information being made available
through LDAP, through SLP or perhaps even through my newly submitted
proposal SSDP
(http://search.ietf.org/internet-drafts/draft-cai-ssdp-v1-02.txt).

		Yaron

P.S. I think that was a plug, but I'm not sure. ;)

> -----Original Message-----
> From: Jim Whitehead [mailto:ejw@ics.uci.edu]
> Sent: Thu, June 24, 1999 5:02 PM
> To: 'DASL'
> Subject: RE: JW9: search arbiter locations
> 
> 
> 
> > What is the mechanism used to discover collections?
> 
> Typically, you'll receive this information from the 
> administrator of the DAV
> server.
> 
> > Why wouldn't that mechanism work for discovering search arbiters
> > other than "/"?
> 
> It could work, but in my view, it's really undesirable.  
> Having to receive
> the name of the top-level DAV collection from a server 
> administrator is
> unavoidable, or, perhaps more precisely, it would have been 
> expensive to
> develop a discovery protocol for this information.  However, 
> it would be
> very inexpensive to allow the search arbiter to be 
> discovered, either by
> querying a property, or by receiving a 302 redirect when 
> trying to perform
> SEARCH on a non-search arbiter resource.  This, in my view, 
> is much less
> expensive than having the search arbiter be a manually 
> entered configuration
> item.  Creating another manually entered configuration item 
> would seriously
> hamper the usability of the DASL protocol.
> 
> - Jim
>

Received on Friday, 25 June 1999 20:45:27 UTC