[Simh] pdp7/decsys bug?

[Daniel Baum]

>
> The OP can test this by doing "DEPOSIT STOP_INST 1" from the "sim>" prompt
> before running the failing program.  This will cause SIMH to halt if an
> unimplemented instruction is executed (the default is to keep running with
> potentially erroneous results).
>
>                                      -- Dave
>

>
> The OP can test this by doing "DEPOSIT STOP_INST 1" from the "sim>" prompt
> before running the failing program.  This will cause SIMH to halt if an
> unimplemented instruction is executed (the default is to keep running with
> potentially erroneous results).
>

Hi all,

This is both an answer to the previous posts on this thread, and a report of 
some more progress in getting things to work on simh.

I tried what Dave suggested. Decsys does not appear to be trying to use an 
unimplemented instruction. Therefore I still have not managed to solve the 
Decsys Fortran problem.

To remind everyone what I'm talking about: Decsys Fortran does not seem to 
work properly with floating point values. It was suggested that it might be 
trying to call a floating point processor, but this does not seem to be the 
case.

On the other hand, TSS/8 seems to be calling unimplemented instructions all 
the time. It stopped almost as soon as I ran it with the above instruction. 
In the case of TSS/8, I have not been able to compile _anything_ with its 
Fortran. In this case, I suspect either a corrupt or missing file, but I 
don't know for sure because the error message is a single number that is not 
mentioned in the documentation.

Other things:

I have managed to compile simple programs in the following languages:

OS/8 Fortran II and Fortran IV.

Unix V7 Ratfor. A surprisingly pleasant language which is actually a Fortran 
preprocessor.

RT-11 Pascal. A rather good, but cumbersome, third-party compiler which I 
downloaded from the web. I installed it according to the instructions and it 
worked without problems. I tried RT-11 Algol, but had no success. I suspect 
it may be meant for an earlier version of RT-11 than we have available.

The RT-11 Fortran and Basic were distributed separately and do not seem to 
be available. (shock, horror!)

TOPS-10. I have only been playing with this for a short while, but I have so 
far been able to compile and run simple programs with its Fortran, Algol, 
Cobol-68, and Bliss compilers. This last is a language that I hadn't heard 
of until this week, and the documentation is truly atrocious, so there was a 
certain amount of luck and guesswork involved in getting it to print "Hello 
World". I will provide instructions if anyone is interested.

TOPS-10 Cobol-74 seems to have a problem with numerical values, probably in 
the PIC clause. It always substitutes an incorrect symbol for the last digit 
of the number. Other than that it seems to work. The same program runs 
correctly with Cobol-68. I thought the problem might be that it was 
targeting the wrong processor, so I recompiled it, but it didn't make any 
difference.

As always, I am not discounting the possibility that the fault is mine and 
not the software's.

That's it for now.

Daniel