E866 Analysis Upgrade Task List - MJL 6/10/93
   (first guess at manpower in days=d given in parenthesis)

A. Backend Analysis

   1. formulate plan and make decsion whether to proceed or not [>9d]
      a. make this list (3d)
      b. evaluate the time/manpower and expertise needed (3d)
      c. look for manpower and see if available manpower
         is adequate to do the job in time (>3d)

   2. obtain 150 MIPS of UNIX computer (from FNAL-CD) [29d]
      a. decide what is required, e.g. cpu power, disk space,
         network configuration, memory, licenses, root access,
         etc. (1d)
      b. negotiate with FNAL-CD (14d)
      c. install hardware (7d)
      d. set up system software (5d)
      e. setup network (2d)

   3. hoist data path [44d]
      a. hoist data from UNIX file to simple UNIX event pool (done)
      b. hoist data from UNIX event pool to simple dump program (done)
      c. same but to full analysis program (7d)
      d. establish UNIX event pool control and exception schemes
         so as to support multiple clients (14d)
      e. establish prescale scheme for hoist data path (2d)
      f. establish BOR/EOS hoist data path with it's signals
         including end-of-spill syncronization (7d)
      g. hoist from VME hoist server (14d)

   4. shared memory histograms [9d]
      a. use HBOOK histograms in shared memory (wait for CERN to
         fix existing routines) (1d)
      b. try PAW on these shared memory histograms (1d)
      c. try HISTOSCOPE on same (7d)

   5. reconfigure analysis for e866 [69d]
      a. remove silicon vertex code (unpacking, analysis,
         pictures, alignment, vertex code) (1d more)
      b. remove ETM driven spill and run analysis (2d)
      c. disable calorimeter analysis (1/2d)
      d. clean up swim code and put absorber back in (7d)
      e. replace old MULTI/ONLINE with new backend analysis interface (2d)
      f. test new program for variety of configurations (need monte carlo) (14d)
      g. update PIX (14d)
      h. check VAX compatability (7d)
      i. make new Patchy PAM and convert to CMZ (7d)
      j. test new program for variety of configurations (need monte carlo) (14d)

   6. reconfigure monte carlo for e866 [21d]
      a. remove silicon aperture, hits, and packing (2d)
      b. remove soft pions (2d)
      c. update and cleanup event generators (Drell-Yan, J/Psi, etc.) (7d)
      d. make new Patchy PAM and convert to CMZ (3d)
      e. test new program for variety of configurations (7d)

   7. backend scaler based monitoring [87d]
      a. determine what needs monitoring
         - study E772 and E789 experience and problems (14d)
         - study plans for E866 luminosity, beam, target, etc. monitors (14d)
         - compile list of things to watch (14d)
      b. implement BOR/EOS backend analysis scheme similar to that for
         main data backend analsysis
         - get data from hoist as for event data (14d)
         - scaler archiver, i.e. put scalers in disk files (3d)
         - SCAN program to monitor in real time (14d)
         - programs for studying all monitors for stability, cross
           comparisons, and correlations (ntuple based) (14d)

B. database for "runsheet" information [22d]
   1. decide what it should include, e.g. run number, tape number, time,
      log page, beam characteristics, events to tape, etc. (2d)
   2. decide whether programmed access is required for input and for
      examination. (2d)
   3. find out cost effectiveness of standard offerings (e.g. ORACLE,
      CERN-HEPDB, ExCel, etc.), their capabilities, and their potential
      accessibility to the whole collaboration (14d)
   4. choose a database and install it (4d)
   
C. magnet field maps [56d]
   1. decide what is needed for E866 configurations, e.g. do we need to
      make new measurements with an new SM0 in place? (7d)
   2. make new measurements if necessary (21d)
   3. analyze measurements (14d)
   4. construct and check new field maps (14d)

D. UNIX-farm development [>28d]
   1. construct and check farm analysis program (14d)
   2. measure IO limit on a test system (14d)
   3. work with FNAL-CD on farm system evolution and improvements towards
      a production system that suits our needs, e.g. try FDDI instead of
      ethernet, etc. (indefinite)