Hello to vintage assembly coders

[tradde]

I've got to be OCD, then. I can still recall IBM 1620 codes from memory.

Some days I can barely remember my name or where I live. I may remember many of the mnemonics, but the actual op codes have faded. You don't use something for 20+ it usually fades.

 

[tradde]

I think you misunderstood my question. I didn't ask about programming in assembly, but rather the raw binary (or decimal, depending on architecture) machine code. The alternative was symbolic programming (assembly). Consider Daniel McCracken's introduction to programming the IBM 1401.

Can you look at a hex dump of 8086 memory and identify subroutines, ISRs, etc.? That's machine code.

I used to be really good at reading IBM core dumps. And following stack traces and all. Found many problems this way. Actually many years later found a really weird problem just in this way. I had not seen a core dump for years, but for this I grabbed it and checked, and checked.
Turns out a string had been assigned a specific size. In the code they tried to increase the size by 1 to insert a character. Sometimes it worked, sometimes not. Turns out if the string was at the end of a page it would overflow on the expansion and cause a page fault. For years they just ignored it and re-ran that job and it would work. I was pretty happy to fix that problem.

 

[Chuck(G)]

In the early days of the 5150, you were advised to read the binary created by MASM 1.0. It had a bit of notoriety for generating the wrong code.