BaBar Reconstruction Requirements Doc
Last updated April 12, 1997

Common Code   ==========================================================

Execution support (Framework, etc, including upgrade II for online et al)
Parameter and command support
Particle data table
Class libraries:
	Physics (CLHEP), mathmatics (CLHEP?), algorithmics (Tools.h++, STL?),
	geometry (???!!!), histo/ntuple, etc.  
AbsEnv load/store support
Geometry interfaces
QA/monitoring, constants generation
	Code for combining multiple jobs
	Automated checking
Error handling infrastucture, interface to production system
 
Every system  ==========================================================

QA histograms (a.k.a. "important data histograms"), summary module, criteria
Event display parts
Conditions database definitions, filling, etc
"Calibrations" both online and offline
Material model contributions
MC GHits -> DIgi (joint with sim), MC truth association, Aslund association
Error handling and reporting
Corrections for imperfect elements (hot, dead, off-gain, etc)
Standards conformance (esp. at level of outside interfaces, use of
	defined toolkit) (CVS, SRT, RGTs, CLHEP, Tools.h++, etc)

Tracking ===============================================================

Track output definition, outputs; associated service classes
	Structure for extrapolation through material
	Vertexing interfaces
Combined track finder(s)
Final track fit(s)
	Mass hypothesis support
	Error matrices within reason, including material
Cosmic tracking
Common segment interface definition
PId information via std interface (pending Art's studies)

SVT  ===================================================================

Hit information for tracking (position, errors, with detailed adjustment)
alignment to common coordinate system
corrections for imperfect elements
cluster reconstruction, residuals, derivatives
monitoring of efficiency, accuracy
stand-alone track finding
track segment finding
dE/dx (pending Gerry's studies), connection to common PID

DCH  ===================================================================

Hit information for tracking (position, errors, with detailed adjustment)
Segments: generation and interface
Bunch selection
dE/dx PID, connection to standard PID
Time to distance relationship
alignment and geometry, including wire positions, etc
"calibrations"
Standalone track finding, with/without Bfield
  Finder/fitter without an IP bias (for albedo, cosmics) - how 
  many cells to achieve 50% efficiency?

DRC  ===================================================================

Standard PID output & connection to standard tracks (>1?)
Photons or similar info for independant PID use
Angle histogram from methods that generate it
Bunch selection information
(??) FInding a conversion in front of an indicated EMC cluster

EMC  ===================================================================

Calibrated energy in each crystal
Positioned raw energy deposits
electron id, muon id information as probabilities (energy, eventually shape)
  Not complex bremstharlung searches, etc, at least not early
Track associations - for std form tracks 
How associate with IFR clusters (hadron shower, K0L search, etc)
Hit crystals nearest a line/track projection through EMC

IFR  ===================================================================

Standard PID output
Magnetic field map information in IFR
Implementation of track extrapolation through IFR iron
Bunch selection
1D, 2D, 3D clusters
muon id, K0L id summary information
hits nearest a line/track projection from EMC

Trg  ===================================================================

Segments, tracks as found
QA code
Bunch finding information
Scalars with livetime, luminosity info
Trigger time (via Fast Control interface)

Event Graphics =========================================================

"event display"
	online
	offline
	code debugging
calorimeter lego plot

Physics Interface  =====================================================

Track vertexing capabilities
Geneology of truth, combinations
Conversion of each reconstruction output to physics quantities, including errors
Primary, beam spot finding (based on algorithms from physics org)
Machine info (beam energies, directions)

"Other"  ================================================================

Sorting out all the bunch selection(s)
Overall PID (default combination)
BField map, time-dependent values in tracking volume

_Not_ reconstruction  ===================================================

Basic graphics, production system, database (applications of these
	are our requirements, as apparently are our debugging needs)
Neural network objects/libraries
"Paw-level" physics tools

MDC1 needs  =============================================================

List of required reconstruction objects
  (From reco in job, but all are in subset loadable from tape)

  	charged tracks
  	EMC clusters as photons
  	IFR clusters as 3D objects
  	charged-neutral associations
  	DIRC info on found tracks
  	dE/dx info desired (2 systems)
    
List of required (rudimentary) physics tools
	4vector calculations, errors
	electron id
	muon id
	spatial vertexing
	beam spot information
	K0s finder
	K0L desired
	sphericity, thrust, missing E, P
	particle ID combination interface, physics to provide implementation(s)

To be developed by physics during challenge:
	kinematic fitting desired (esp beam constraint)
	photon conversion finder
	brem finder for J/psi

.xdr-level persistence required, but tape handling / production is SEP