NEMUG Newsletter
August 2003, Articles
Board Meetings
NEMUG Executive Board Meetings are open to anyone to attend. Your presence and views are welcome at any time. The meetings are held on the second Monday of even-numbered months, at the office of Henry Elliott in Wellesley. We try to get started at 6:00 pm. Dates for the next few meetings are:
- 8/11/03
- 10/13/03
- 12/8/03
If you are planning to attend, please let a board member know in advance, as we provide food and eat as we work, and need to be sure we have enough for everyone. The complete address for the meetings is:
Henry Elliott and Company Inc.
One Washington St. Suite 208
Wellesley, MA 02481
July Meeting: "M Ontogeny and Some Crystal Bollixed Predictions"
Presenter: Bob Beckly
Bob Beckley, one of the early pioneers of medical computing using M offered us a 'fireside chat' style presentation on the early history of M. Bob (sitting to chat with us, rather than standing to lecture) reached back to the beginning of Mumps' at MGH's Lab of Computer Science in the 70's. Neil Pappalardo and Curt Marble were working for Dr. Octo Barnett using an early DEC mini (a PDP/1?) that wasn't so 'mini'. We heard how Dr. Barnet had forbid Neal and Curt from developing an OS / programming Language and which they interpreted as advice they chose not to take.
Bob described the heyday of MUMPS in the late 70's and 80's when there were a lot of vendors of M and M applications start-ups. and in particular how MIIS and later DataTree M was used by the Brigham and the Beth Israel hospitals, and the beginnings of InterSystems and IDS (now IDX).
DEC developed a version of MUMPS for its hardware very early on. There were long strings with lots of delimiters because of space and expense issues.
Bob recalled reading a book by Knuth in the early 70's that talked about having string subscripts and he thought how nice that would be. and shortly thereafter 'Standard MIIS' was came out with just that and all other versions of Mumps had them too. Bob and his colleagues began work on the HIS at Brigham/BI and in 4 years developed 30 systems including billing.
In the early 80's networking was in its infancy and minicomputer hard wired terminals were the way users were able to access on-line data.
In 1988 the VA issued a $4-6 billion contract (mostly for hardware) and MUMPS was on the map. IBM even began to offer it on their hardware. In the 80s the first MUMPS for UNIX was written and Dave Brown wrote MUMPS for MAC. Peter Beaman developed 'DataTree Mumps for the PC' and began working with the BWH to put together a Client Server Network of PC's that is still in use today (although it is finally being converted since it runs in a DOS window).
Many people in the audience participated in the discussion adding their own experiences from the early days of M computing. We discussed the consolidations in 90's and the impact of new hardware and the demise of the many M vendors, the MTA and the MDC.
Bob said much of what was happening in M during the 80's was also occurring elsewhere. At Brigham they were working on a thin net and suddenly Windows OS blew everything away. With the advent of the windows GUI environment, MUMPS looked "old fashioned".
Then, in the 90', the Web hit. The web can be great for M. M works great as a 'back-end' data server; it is simple and fast, easy to develop in. Because we had used up 2 hours talking about M's rich past, Bob's 'Crystal- Bolixed Predictions' were squeezed in to the last few minutes.
~ submitted by Heidi Pape Laird