Proposal for an International Grid Testbed for the BaBar experiment.
====================================================================

Milestones and targets
======================
1.  Fix list of participating sites
-----------------------------------
     SLAC, RAL, IN2P3, Manchester, Dresden, Bristol 

2.  Arrange system for authentcation 
------------------------------------
    All participating sites agree to recognise certificates issued by:
           CNRS, UKHEP
    This list will be extended by mutual agreement. 
    Until the US gets a recognised CA, US BaBar users can get CNRS certificates.

3.  Arrange system for Authorisation
------------------------------------
    A Virtual Organisation ldap server  will be set up
             (Tim Adye and Andrew McNab)

    All participating sites agree to use this with the dynamic account system
           or to issue individual accounts to certified BaBar grid users 
           freely and with minimal user action.
    
4.  Arrange specific hardware in participating sites as required.
-----------------------------------------------------------------

5.  Draw up detailed definition of applications with use cases
--------------------------------------------------------------
     5.1 Running a kanga analysis job remotely
     5.2 Status monitoring
     5.3 Data Import
     5.4 Repeat Analysis
     5.5 Site problems
See 
 http://www.slac.stanford.edu/BFROOT/www/Computing/Offline/BaBarGrid/usecases.ps

     5.6 Data replication
 See http://www.slac.stanford.edu/~hasan/ppdg/SRBPrototype.html 
        (though this needs to be extended to replicate outside SLAC)

     5.7 MC production
     5.8 Analysis using Objectivity
     5.9 Analysis of large ntuples

Not yet written - volunteers sought

Volunteers also sought to add other use cases.

6.  Demonstrator tests showing it works in principle
----------------------------------------------------
    6.1 Globus-job-submit between specific sites
    6.2 Solve the input sandbox (afs) problem
    6.3 Set up ldap server with resource details (Marc)
    6.4 Set up resource broker
    6.5 Job submission as in EU datagrid

7.  Develop into useful and robust tool
---------------------------------------
    7.1 Output merging and retrieval (as desired)
    7.2 Common error handling protocols
    7.3 Documentation
    7.4 Provision of user-friendly interface. 

8.  Performance tests identifying bottlenecks
---------------------------------------------