Version 2.4

MCFAST with forward and central tracking

The version v2.4 release of MCFAST contains a complete rewrite of the
tracking code from v1.4, but is very similar to v2.3.  The techniques are 
similar to those used in v1.4.1 which
was based on the program trackerr. More complete documentation on the 
fitting method used in v2.4 can be found in ~bphys/docs/avery.

The user needs to run the following script to set up mcfast on fsgi02.fnal.gov.

~bphyslib/setup_files/mcenv.fnalu v2.4    (csh)
~bphyslib/setup_files/mcenv.sh.fnalu v2.4 (sh)

This will setup the proper paths to the mcfast libraries.  If 
you are using mcfast at another site, be sure to change the $BPHYS
path name in mcenv and check the file setup.$SITE_NAME.  External
libraries are set up in ~bphyslib/setup.$UNAME. 

The new version can be found in ~bphyslib/mcfast/v2.4.  The following
directories contain 

mcfast/v2.4/Explorer   -- SGI 3-D graphics for mcfast
mcfast/v2.4/db         -- templates for geometry structures
mcfast/v2.4/doc        -- documentation for this release
mcfast/v2.4/example    -- Sample Makefiles, geometry and command files
mcfast/v2.4/inc        -- include files can be found in the subdirectories     
mcfast/v2.4/lib.$UNAME -- mcfast libraries
mcfast/v2.4/src        -- source code can be found in subdirectories

mcfast/v2.4/example    -- *.cmd  command files for control of mcfast
                          *.db   geometry file
                          *.fil  list of data files
                          Makefile
                          stable.dec  -- table for mcfast decays 
                                        (set make_decays .TRUE. in .cmd file) 

mcfast/v2.4/inc/event  -- include files and structure definitions for
                          generator particles and tracks
               /geom   -- include files and structure definitions for
                          detector geometry
               /trig   -- include files for the trigger              

mcfast/v2.4/src/datafile- data file management
               /dbin_mcfast -- routines for DBin interface
               /dst    -- XDR filters and structure
               /gamma  -- photon reconstruction routines 
               /gen    -- general routines for mcfast
               /geom   -- read in and set up detector geometry
               /graph   - interface to explorer routines
               /io     -- xdr routines for i/o of structures
               /muon   -- muon tracking - no bfield
               /shower -- em shower code
               /track  -- tracking code
               /trig   -- trigger code
               /user   -- default user routines
               /util   -- various utility routines
               /vertex -- vertex fitting


User routines:  see ~bphyslib/mcfast/v2.4/src/user
usr_analysis.F        -- user calls to mcfast tracking and analysis

usr_before_trigger.F  -- call to initiate trace of particles through 
                         the detector volume. Hit generation
                         and decays in flight are also done here.
                         
usr_command.F         -- called when "user" typed to the program
                         during execution
         
usr_filter.F          -- user can filter events to save analysis time

usr_generate_event.F  -- controls the generation of internal events 

usr_hbinit.F          -- initialize hbook output files

usr_hbsummary.F       -- close hbook files

usr_hsinit.F          -- initialize histoscope

usr_init.F            -- user initialization 

usr_mcfio_out.F       -- user routine for calling output from  MCFIO package 
     
usr_summary.F

IMPORTANT NOTE TO USERS OF MCFAST:
************************************************************************
1. All floating variables except those that are geometry related should
be declared with the DFLOAT directive, rather than REAL or DOUBLE
PRECISION. DFLOAT is a C-preprocessor flag which currently translates to
"double precision" but can be changed in the future. DFLOAT is defined
in the file dev2/inc/event/const.inc. Simply do the following in your
programs:

#include "const.inc"   !Defines constants, units and default float type

      DFLOAT x, y, z   !DFLOAT must be upper case

The C-preprocessor will automatically translate DFLOAT to the string
"double precision", so be careful not to have too many variables on the
line lest you exceed the 72 character limit when DFLOAT is translated.
Almost all variables defined in dev2/event/*.inc are defined as DFLOAT,
as you can see for yourself. I have not changed the variables in the
geometry area yet because of time.

When you use ucopy and vzero with variables in a way that works when
DFLOAT is set either to "real" or "double precision", simply change

	call ucopy(a_array, b_array, 10)

  to    call ucopy(a_array, b_array, 10*FLOAT_WORD)

where FLOAT_WORD is set to the number of standard machine words in a
floating point word (1 for real and 2 for double precision). This
parameter is defined in const.inc.

************************************************************************
2. const.inc also contains definitions of units, so that there are no
more implicit assumptions made about units in the program. For example,

      unit_meter = 100.     !Length measured in centimeters
      unit_sec = 1.E9       !Time measured in nsec

For example, if you want to define 100 microns in our defined units
without knowing what those units are, simply use

	1.E-4 * unit_meter

mcfast uses standard units of cm, nsec, GeV and Tesla, but they can be
changed to anything else by making the appropriate changes in const.inc
and recompiling everything.