1. Framework for reading CODA formatted events from ET and producing
histograms for online monitoring.

o  Incorporate new EVIO library in the form of a DANA DEventSource
   object.

o  Test it in an online environment during all available beam tests.


2. "Phase 1" reconstruction of simulated data from all detector
systems.

o  Tagger: Final object format for factory output should be designed
   first. This should be filled parametrically in the Geant simulation
	to allow development of the reconstruction code downstream.

o  UPV shower position/energy reconstruction. (not necessaily final
   algorithm) Final object format for factory output.

o  Start Counter position, time and covariance. Final object format
   for factory output.

o  CDC hit based and time based information output. This includes
   any z-position information, either by direct measurement (charge
	ratio) or rudimentary segment reconstruction of stereo layers.
	Final object format for factory output.

o  FDC hit based and time information. Package level track segment
   finding should be done. Mechanism for passing both locally resolved
	Left/Right ambiguities to tracking as well as ones that must
	be resolved globally. Final object format for factory output.

o  BCAL shower position/energy reconstruction. (not necessaily final
   algorithm) Final object format for factory output.

o  TOF position and time reconstruction. There may need to be an
   iterative mechanism here between tracking and TOF. Details for
	this should be completely worked out.
	Final object format for factory output.

o  Cherenkov: A decision on the exact type and form of the Cherenkov
   detector needs to be made before serious work can begin here.

o  FCAL shower position/energy reconstruction. (not necessaily final
   algorithm) Final object format for factory output.


3. "Phase 1" Tracking and particle ID. Output into format that can
by read by CMU PWA code.

o  Pattern recognition using hit-based information completed and
   optimized for Monte Carlo data.

o  Full tracking through an inhomogeneous magnetic field. The Kalman
   filter should be implemented at least in a rudimentary way.

o  Particle identification factory designed with well defined outputs.
   Not necessarily final algorithm, but at least capable of
	identifying probable Kaons.

4. "Phase 2" Tracking and particle ID.

o  Off beamline vertex reconstruction.

o  Improved particle ID through kinematic fitting.


5. Calibration/Parameters database designed and implemented

o  Database structure for holding both calibration constants and
   online information (some of it time indexed, some run-indexed)
	designed. Table formats documented.

o  Access method (Web/File, XML, security, ...) designed and tested
   for remote sites.

o  C++ API designed and implemented in DANA. Other APIs may also
   be needed (Perl, HTML, ...)


6. Geant4 simulation.