Mark Cathcart has posted a nice article about the Unix System Services roll-out process. He provides a historical overview of the decisions involved in the revitalization of the mainframe back in the early 90’s.

Personally, I remember when we introduced USS at my company about 8 years ago. It was a hard but also a funny work. Lots of mainframe guys dealing with a subsystem under OS/390 that has a strange behaviour and terminology.

What the hell is a “sticky bit“?
Oh, I see, it’s an executable that resides on LPA or LNKLST.
Why this guys doesn’t speak more clear!

Can I manage the user access rights via TSS?
Yes but not. Do you know what are the file mode permissions bits?

I also remember when I personally had to deal with Rexx and USS. We needed to implement a change control system to manage the production environment changes. Executing OpenEdition (USS former name) services through the “Address Syscall” Rexx instruction under a TSO/ISPF environment on a OS/390 mainframe was very very funny (and painful!).

I remember specially when one of our sysprogs executed the “rm” instruction on the development root filesystem using the change control user, which was, logically, a superuser. Just imagine his face when he realized what he has done! After this destructive action, we created an “rm” alias to prevent these kind of incidents.

Snif, I miss these great moments.