Hi Norman (mostly),
just a heads up on some fixes I've applied to make-prm-par.c.
I'm playing around with gfortran and g95, just to see how bad things
really are (quite is the quick answer), and came across a problem with
setting the VAL__BADD value. There's currently no recognisable HEX format
(they use X'...' or Z'...' not our '...'X, but these BOZ constants aren't
allowed for anything other than INTEGERs anyway...), so the PRM value is
being set to -1.79769313486231571E+308. Now g95 and gfortran both complain
that this overflows, so I thought it must be down to them (being rubbish),
but then I remembered some Fortran and realised that this should really be
-1.79769313486231571D+308, "E" is for real value constants. I've added a
patch for that.
Now for an older point reborn. Those numbers should really be
-1.7976931348623157D+308, that extra "1" is redundant, so I've also
changed that like we decided for prm_par.h, I've also done the same for
single precision numbers.
Cheers,
Peter.
|