This  is  a little wordy, but I'm hoping someone with
Fortran/COBOL/Segmenter expertise can point me in the right direction. I
must be missing something obvious here. We recently put a new version of
a very old  G/L application into place. The application is made up of
about 180 Fortran/66(CM) routines, and about 10-15 COBOLII/85(CM), all
compiled into one USL,  and PREP'd into one PROG. I wrote most of the
COBOL stuff, and a some of the Fortran. About 60 Fortran routines were
modified, about 4-5 new Fortran programs were written, and about 4-5 new
Cobol programs were written, all for Y2K, and all of them are working
great.

 Here's where the problem starts, we are now in the process of
converting this application from FORTRAN 66/V to 77/V, still in CM. The
COBOL stuff is already ansi 85, so I anticipate no changes to them, they
work.

I ran the MIGF77 utility, and made many mod's to get all 180 Fortran/77
routines to compile into a USL. As soon as they compiled clean, I  got a

CODE SEGMENT OVERFLOW.

All the routines had a $OPTION SEGMENT statement on them, and all 180
routines went into about 10-15 different segments. So, I removed all
these SEGMENT statements, thinking that it would work like COBOL, and
each routine would go into it's own segment. Well I was wrong, they all
went into one segment called "SEG' ". Then I wrote a neat little command
file to insert into the first line of each source file, the SEGMENT
statement containing the name of the source file as the segment name.

Now each routine is in it's own segment, however now I get a ton of
errors at PREP time.

"ERROR #45 ACTUAL PARAMETERS INCOMPATIBLE WITH FORMAL PARAMETERS".

And this is really strange, two of these problems relate to the COBOL
routines that I haven't changed AT ALL!  They are getting some error
like CALLED WITH WRONG NUMBER OF PARAMETERS calling intrinsics, ASCII,
and COMMAND, and I know the parms are correct.

Is there a difference in the way items are stored in the 66/V to 77/V,
that would affect the way COBOL is passing parms to intrinsics?


I would really appreciate any and all clues, to point me in the right
direction!
Please send any replies back to the list, because I won't be able to
check mail at this address until late Thursday night.

Thanks in Advance,

Michael Anderson
mailto:[log in to unmask]
mailto:[log in to unmask]
mailto:[log in to unmask]