The following corrections have been made in the programs currently present at
this ftp site.  If you have previously downloaded these programs, please make
the necessary changes and recompile the programs.  Corrections effective 
9/18/97.

NLSL has been updated on 12/3/97 to correct matrix overflow problems that
may affect some computers.  The new programs are significantly changed 
from the previous version and hence a file of CHANGES is not present.  It
is suggested if you use these programs you download the new files.

***********************

./NL2DR/xshft.f

In the NL2DR directory, the routine xshft.f had an incorrect line 
defining b:

c  Find ordinate and abcissa for extremum of a parabola through ixmx and 
      if (ixmx.gt.1 .and. ixmx.lt.npt) then
         a = ans(ixmx)
         b = (ans(ixmx+1)-ans(ixmx-1))
         c = 0.5d0*(ans(ixmx+1)+ans(ixmx-1))-a

./NL2DC/xshft.f

Here the correct code is used:

c  Find ordinate and abcissa for extremum of a parabola through ixmx and 
      if (ixmx.gt.1 .and. ixmx.lt.npt) then
         a = ans(ixmx)
         b = 0.5d0*(ans(ixmx+1)-ans(ixmx-1))
         c = 0.5d0*(ans(ixmx+1)+ans(ixmx-1))-a

This change has been made in all files as of 11/6/97.

*********************

In the following two directories the term wtol wass used to reduce the 
matrix size by discarding small contributions.  Unfortunately in some
simulations, this can result in errors in the simulations.  Replace
this section of code with with:

	wtol=1.0D-9

The errors were:

skeeve>find . -name pcheck.f -print -exec grep wtol {} \;
./NL2DR/pcheck.f
        wtol=5.0D-4
        wtol=1.0D-4
        wtol=2.0D-5
        wtol=8.0D-6
        wtol=4.0D-6
        wtol=1.0D-6
        wtol=1.0D-7
./NL2DC/pcheck.f
        wtol=5.0D-2
        wtol=1.0D-2
        wtol=2.0D-3
        wtol=8.0D-4
        wtol=4.0D-4
        wtol=1.0D-4
        wtol=1.0D-5

This change has been made in all files as of 11/6/97.

******************************

Noted on 7/16/97:

In the files:
./NL2DR/BASIS/cd2km.f
./NL2DR/cd2km.f
./NL2DC/cd2km.f

The code now reads:

      if (dabs(ang).lt.rndoff) then 
         sa=0.0D0
         ca=1.0D0
      else if (dabs(180.0D0-ang).lt.rndoff) then
         sa=0.0D0
         ca=-1.0D0
      else if (dabs(90.0D0-ang).lt.rndoff) then
         sa=1.0D0
         ca=0.0D0

Note the necessary minus sign for cos(180).

