The update in the current version was made on cross-sections and some experimental improvement on the jam interaction model.

• hA cross-secion: (h=p,n,pi,K...). The absolue value at 200 GeV is normalized to the values given in the PDG book.
  (For high A (such as W, Pb..), the correction is ~0 %; For low A's such as C, N, etc the correction is + ~2.5%. Exception is Oxigen for which it amounts to +5.5%; why Oxigen is exception is not known).
  Note: the values in PDG has been changing with the time.
  
  
• Heavy ion inelasic collision cross-sections ( AA' collision ) were also updated so that the user can choose either of old values or new ones. New one is made by normalizing the value at 5 GeV/n (K.E/n) to Shen's calculations (1998) used in PHITS. The correction is order of 10% but depends on the combination of AA'. Shen's result is based on some experimental data below few GeV/n which have normally large errors; at present we cannot judge it is good or not.
  The new values at high energies are close to the QGSJET-II.
  Note: If each interaction model provids the cross-section (like QGSJET-II), we use it.
  
  
  • The parameter:
    AAXsec=0 (default) uses old cross-secitons.
    AAXsec=1 is for new normalization. AAXsec must be given in "param" file ($Hparm).
  
  
• For heavy ion collisions, Jam interaction code generates spectator nucleons as indepenent nucleons so we cannot get heavy ions from hA' or AA' collisions. That is, for hA', the target A' breaks into A' nucleons. For AA' the same is true for target A' and projectile A. Such treatment is not harmuful for the target in our applications. But for the projectile, it will have some effect. This feature has been remedied in PHITS code. However, PHITS is intended to be used at low energies, and we found that, if we use it at high energies, it takes very long time in treating spectators (For 100 GeV/n Fe, normally spectator treatment needs 1 min or sometimes 5 min or so).
  So for the time being, we introduced a very crude method of forming heavy ions from the spectators. (The energy conservation holds very well and we expect it is not so bad. We apply a method used in Sibyll code for heavy ions A<=56; for heavier ions, we employ data generated by DPMJET-III). The effect is rather large; as compared to the case of old one (i.e, all spectators go into nucleons), the flux for large A projectile (Fe) drops by 5 % or more for W, PWO targets.
  
  
  • The parameter
    JamFragment=0 (old treatment, i.e, spectator go into nucleons).
    JamFragment=1 (default. new treatment).
    JamFragment must be given in "param" file. ($Hparm).
  
  
• AutoEmin=1 is now default in Epics9.13. AutoEmin=2 gives little bit larger lateral spread (1~2% level) than AutoEmin=1 or 0.( 0 and 1 give statistically the same spread, but 1 should have wider appliabilty).
  
• The awk command (the latest version 1.3.5) seems to have a bug (?) and is unable to process a regular expression like [a-z][A-Z]?* which has been correctly handled by v.1.3.3, 1.3.1. ( [A-Z]) is neglected in v.1.3.5). This regular expression is used in the fordpmjet command (in Epics). The workaround is to use [a-zA-Z] instead and implemented in the present versions. At present, it is not clear that similar expressions are used in other places or not.