8.1: Some information about LLMPS

posted Oct 19, 2014, 7:47 PM by Jeff Ogden   [ updated Oct 19, 2014, 7:48 PM ]
On Oct 19, 2014, at 9:36 PM, Jeff Ogden wrote:

I've been reading the Computer History Museum's 2009 Oral History Interview of Frank Belvin (http://archive.computerhistory.org/resources/access/text/2013/03/102702246-05-01-acc.pdf). Frank worked at MIT's Lincoln Laboratory and was responsible for the development of LLMPS. He is co-author together with Joel M. Winett of Lincoln Laboratory Multi-Programming Supervisor, Lincoln Manual 78 (ESD-TR-67-14). He went on to co-found IDC [Interactive Data Corporation].

Some background on the development of LLMPS from page 11:

My first task (we had gotten a small model System/360) was working with another fellow, Joel Winett, to develop a multi-programming system that we'd use on the small model 360 to replace all the 1401s that were in the batch shop. It was a typical batch shop, you come in with card decks, they get copied to tape, the tapes go over to the mainframe, you run the job, the mainframe produces output, either cards or on a tape, and then you take the tape back to the 1401 and do printout. We had three 1401s supporting this one 7094, and it was Jack's idea that we should cut our teeth on learning the 360 and do something useful by replacing all those 1401s; so Joel and I wrote a multi-programming operating system with the help of several people in our group who wrote various applications. We ended up replacing the 1401s and I remember a smooth running system.

I came across this on page 14:

And in fact, this little multiprogramming supervisor I mentioned that Joel and I wrote, we contributed that to the SHARE library, and it apparently got heavy use, not only within IBM, but also the University of Michigan picked it up, and it was a little, tiny inside part of the University of Michigan Multi Programming System (UMMPS), which then became the Michigan Terminal System. So we were heavily involved. I had Mike Alexander come out and visit us. We were very open. But that was before we became our own separate commercial company. I think competitiveness at the university and laboratory level is very cooperative, whereas competitiveness outside is your bread and butter, and so you think more carefully about it.

And here is another interesting statement from page 31 (the last page of the interview):

The most fun I've had other than writing LLMPS [Lincoln Laboratory Multi Programming Supervisor]] at Lincoln Laboratory, were the first years [at IDC], even including going through the layoff process, which was a very painful one.