Re: Can fortran clear the 3270 screen?

Rick Fochtman <[email protected]> Mon, 07 Jun 2010 19:19:49 -0500
Newsgroups gmane.comp.emulators.turnkey-mvs
Message-ID <[email protected]>
  -------------------------------<snip>----------------------------

>
> > FORTRAN doesn't know anything about "screens" but a very simple
> > Assembler subroutine, callable from FORTRAN, can clear the screen
> > for you.
>
> If you use an external subroutine, which you would separately
> compile, and all it contained was /DATA/, then it is possible to
> include anything in that subroutine as "data", even functional
> machine code.
>
> As a TPUT is SVC 93, you would include what ever "orders" are
> required for a "clear screen" function, and then issue SVC 93.
>
> Everything would be coded in the appropriate source, possibly decimal
> which was converted to hexadecimal by the compiler.
>
> The MAIN would invoke the DATA control section as if it was a closed
> subroutine. MAIN cannot tell that the named subroutine is actually
> DATA, as it was compiled separately.
>
> The "usual suspect" prefix and suffix instructions would be required,
> as with all assembly subroutines, ending in a BR 14, of course.
>
> This is a "hole" in the FORT G and FORT H compilers which, AFAIK, has
> not been closed.
>
> You could, for example, assemble the desired routine, and also name
> it the same as the /DATA/ section, but this is not a requirement, and
> relatively simple assembly-type routines may be made by hand coding
> them, as described, as a FORTRAN external DATA area.
>
-------------------------------<unsnip>--------------------------------
That's all true, but it's sort of like trying to take tonsils out via 
the rectum. Why would you want to? 

Why re-invent the wheel?

Rick