 |
 |
> > In point of fact, I'll remind you that Digital/Compaq(/HP) > > Fortran for Alpha does indeed store .false. as 0 and .true. > > as ~0 (or -1 for those still thinking in two's complement), If you do a TRANSFER((.TRUE.),(X)) and print it out, you get -1 if X is INTEGER and -(HUGE(X)) if X is real (-1.7014117E+38). (For .FALSE., you get 0 and 0.0000000E+00). Also, it doesn't have IEEE format, but another format (designed for compatibility with the VAX format) and ALPHA, at least with VMS, is little-endian. Fortunately, none of this matters if one uses standard Fortran. That's one purpose of a standard: transparency with regard to the underlying architecture. > > and uses even/odd as the testing method. And they are easily > > converted to - and ~0/-1; as another poster said, these are > > just bit patterns and they take on "value" when you impose an > > interpretation on them. They aren't the right values or the > > wrong values for anything, until you declare the consistent > > context which imposes an interpretation. What's your context > > to claim they are the wrong values? > > I think you'll find that the even/odd testing goes back to the Vaxs (or > earlier) where INTEGERs could be used interchangeably with LOGICALs, and > status codes could be coded as odd numbers for failure and even for > success. So you would use "IF (status)" to test for success, where > status could be any one of many possible values treated as either .true. > or .false.. If you're thinking of VMS status codes, even are bad and odd are good. In particular, 0 is warning, 2 is error and 4 is fatal error; 1 is success, 3 is informational. (These are actually the 3 lowest-order bits, known as $SEVERITY. $STATUS is 32 bits and codes for many other things as well. > The proper definition of a logical data type is a big plus for Fortran. > Program by the standard and you never have to worry about the underlying > implementation. It really does not matter. Right.