A great session, with lots of technical meat. The only session I went to all week where it was standing room only.

Bob Shannon talked about the subsystem interface, its uses and history. He gave a clear and concise overview of the control blocks involved, and a comprehensive overview of the users of the SSI. He knows his stuff, he can talk in front of a crowd, and his sense of humor is frighteningly close to my own.

I've been slowly assembling a (small) bibliography on the subsystem interface for an article I've been meaning to write. Bob's bibliography is a superset of my own, and he goes back 15 years or so (not that 15-year-old documentation is of that much value to anyone in this business). It reflects a LOT of research in a topic that doesn't have all that much published material.

While he didn't say much about the subsystem interface itself that I wasn't aware of, he talked at length about using some of MVS' internal routines to build the subsystem control blocks. I had been rolling my own control blocks, and the advantage to his method is that you are more release independent.

The handouts will be valuable resources for a couple of years, easy.

He surprised me when he said that IBM has formally reserved a set of subsystem function codes for customer use! This interface used to be treated by IBM as an internal interface. It was undocumented and red flags were planted everywhere: use at own risk! This has changed only recently with the advent of operations automation tools that get into the subsystem interface.