IBM Mainframe Forum

Hello!
So, I've been trying to understand this code:



I've tried several Cobol compilers, and I didn't get any sentence that made sense, so, since I'm a newb in the area, I'm certainly doing something wrong.
Could someone walkthrough how to get what it prints out? Do I need a different machine to run the code?
Thanks in advance!

if you have tried several COBOL compilers then you are probably not running on a mainframe and that is what you need to do, compile and execute. On a mainframe the character set is coded differently from PC data. The source says IBM 370 so you have to run on an IBM mainframe.

Well, I do not have access to an IBM mainframe. Is it easier to try to emulate it, in this case, instead of understanding the code?
I'll try that!
Thank you.

What do you mean by 'emulate'? Run on mvs 3.4 on a Hercules emulator? That would do it as it uses the EBCDIC encoding and not ASCII.

These users thanked the author NicC for the post:

Nitele (Mon Jul 30, 2018 6:43 am)

I think the code is there to help you understand: A byte is a byte is a byte. With most architectures, a byte consists of 8 bits. So a byte can hold 256 different combinations of the 8 bits. What any of those combinations means depends on how you look at the byte. Is it a single character? If so, what code page? Is it a one byte binary number? If so, signed or unsigned? And so on.

This code defines 13 single precision floating point variables (COMP-1), each one needs 4 bytes to store a floting point number. Each variable is initialized. The compile needs to transform the haman readable number to the internal format. E.g. the first number -.24093455E+04 is transformed to the four bytes x'C3', x'96', x'95', and x'87', or x'C3969587'.

The code then tells the compiler that you want interpret the 13 * 4 bytes in a different way, i.e. as a character string. So the 4 bytes above are lokked at as the first 4 bytes of a 52 byte character string. But what code page is to be used? Since this is about mainframe, the code page can be expected to be an EBCDIC code page (which one exactly is unknown). This is what the DISPLAY statement does, it looks at the bytes as characters and simply writes them out.

If you want to run this program on a ASCII platform, and still get a meaningful text being displayed, you need to initialize the float variables with values that correspond to the text in ASCII. Another option is to run the program as given, save the output in a file, and then use the editor of choice to display the content as EBCDIC data.

HTH

These users thanked the author phunsoft for the post:

Nitele (Mon Jul 30, 2018 6:43 am)

I tried using Hercules yesterday all night. It seems I'm too much of a newb in the subject to make anything work, but I bet it is suposed to be the easier solution, indeed. hahahahaha
I tried to follow your tips today, phunsoft. I got some results, based on another similar problem in this forum!

Thanks for the help, guys.

What is the result you got?

Since it's a challenge at hacker.org, I'd be spoiling it for anyone who searched it if I posted the text!
But I got something similar to "Congratulations, the answer is blablabla". Hahah