My take on this as a programmer with only 23 years on z/OS, and only 10
years before that:
My use of variable names and comments varies with the target audience but
not that much.
Whether I'm writing a sample that will have folks of nearly every skill
level look at OR I'm writing code that I will need to maintain, I tend to
have comments that describe the WHAT or WHY but use the instruction itself
to describe the HOW.
LA R3,encdata Output encrypted data
LLGTR R3,R3 64-bit clean
STG R3,CRYPV_OUTPUT_ADDR And stored into CRYPV
It might not be obvious why I cannot just LA/STG (yes, I KNOW I could use
LAE but I was young and naive when I wrote this 4 years ago :) ) but the
comments will explain the WHAT and WHY.
As far as I am concerned, it's more important for the source code to be
understandable than the listing. The listing will help in cases where the
source might be relying on USINGs that rely on other USINGs and the
assembler can straighten things out.
Eric Rossman, CISSP®
ICSF Cryptographic Security Development
z/OS Enabling Technologies
[email protected]
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
