There are a variety of remote remote services available on the 
Internet.  These include file servers, telephony gateways, white pages, 
and many others.  They are currently very difficult to discover.  Often 
the name or network address of the service must be explicitely 
configured with a client system.  
 
There are a variety of protocols available which could ease the 
configuration of wide area services.  These would primarily function 
through the use of storing a pointer to the service indirectly 
and providing a resolution service (URN, DNS, LDAP, etc.)  None of them 
are currently being used to solve the problem of wide area service 
location, though each is used to solve specific service look-up 
problems.
 
WASRV will specify the required architectures and protocols to make it 
much easier to discover remote services.  In particular, we will seek to 
define a mechanism which allows automatic configuration of clients which 
are only aware of the attributes of a server they need in advance, not 
its location.  The more dynamic and decentralized this service is, the 
better.

Participants in the WASRV BOF are encouraged to read an internet draft 
discussing the problem and suggesting directions for a solution to it:
   
   Rosenberg, Guttman, Moats, Schulzrinne, "WASRV Architectural
   Principles", draft-rosenberg-wasrv-arch-00.txt
 
Participants may also wish to read proposed solutions to the WASRV 
problem:
 
   Rosenberg, Suter, Schulzrinne, "Wide Area Service Location",
   draft-ietf-svrloc-wasrv-01.txt
 
   Moats, Hamilton, Leach, "Finding Stuff (How to discover services)",
   draft-ietf-svrloc-discovery-05.txt
 
   Moats, Hamilton, "Advertising Services (Providing information to
   support service discovery)", draft-ietf-svrloc-advertising-03.txt