Strategy for STAF Transition
1: Get XXX_user into STAF as one large module
- Define IDL structures for ALL input and output. Remember to convert ntuples to STAF tables. Convert input from various random files to STAF tables and let STAF do the input.
2: Subdivide PAM into 2 PAMs
- The first PAM does all simulation after PISA (corresponding to the old PISORP 1) and outputs the RAW data tables and fictious tables
- The second PAM does all the offline processing based on the RAW and fictitious tables
3: Increase the level of sophistication in PAMs
- Change offline PAM into several modules only operating on RAW data and data derived from RAW data and create another set of PAMs evaluating the derived data based on the fictitious tables.
4: Start changing PAMs from Fortran to C/C++
- This task is easier, when the PAMs are not too large
- By changing one PAM at a time you will always have a working version of all your PAMs.