Mumpster
http://mumpster.org/

Terry L. Wiechmann
http://mumpster.org/viewtopic.php?f=6&t=43		 Page 1 of 1
Author:  tlwiechmann [ Thu Nov 18, 2010 7:58 am ]
Post subject:  Terry L. Wiechmann
I've been a Mumpster since 1971. I spent a number of years programming on MUMPS systems starting with a variation of the original MGH MUMPS on a PDP-9. I started ESI Technology Inc. in 1984 after working for DEC for almost 10 years. Our primary focus was MUMPS and associated software training. When the IT world was shifting to OO, I started and funded the EsiObjects project. Jerry Goodnough and John McManamon implemented it.

While at home during non-winter months, I spend most of my time working in my wood shop, traveling with my wife. However, I go to Key Biscayne for the winter and unfortunately cannot take my shop with me so I revert to MUMPS/EsiObjects work.

So, between now and April, I will try to make a contribution to the Mumpster User Group where I can. I'm particularly interested in getting our Computer Based Instruction packages and EsiObjects working on existing MUMPS, particularly GT.M, MV1 and M3.
Author:  icekat [ Thu Nov 18, 2010 4:21 pm ]
Post subject:  Re: Terry L. Wiechmann
Welcome to Mumpster, Terry!
Author:  toad [ Tue Dec 07, 2010 1:58 pm ]
Post subject:  Re: Terry L. Wiechmann
When Terry says "MGH MUMPS," he's saying he got to write on a later version of the original MUMPS created by Massachusetts General Hospital. The man knows his MUMPS.

He also knows object-orientation (OO), an important branch of computer science. He and Jerry fought a long, hard battle to educate the MDC about the importance of object orientation and why we needed to add it to MUMPS. Their and John's EsiObjects is full of subtle wisdom about what works and what doesn't in OO, especially when combined with MUMPS.

Terry has this gruff, growly, flat-voiced manner of talking that disguises how passionate he is about OO, MUMPS, and teaching, but don't let him fool you. The man's a valuable resource. His help with Mumpster and Paideia is a great blessing for all you future mumpsters out there.

If we're lucky, Terry will start a thread in the MUMPS Implementations forum in which he will describe in detail what the version of MGH MUMPS he worked on was like. I am very, very, very curious about it, especially about how it differed from the MUMPS 1977, the first standard MUMPS.
Author:  tlwiechmann [ Tue Dec 07, 2010 5:19 pm ]
Post subject:  Re: Terry L. Wiechmann
Thanks for the kind words Rick. My growl is worst than my bite :-)

I walked into Meditech in 1971 and interviewed with Nick Johnson, Niel Papalardo and Curt Marble. I came from a couple of years programming on a Honeywell 200 batch system (yes, paper cards, paper tape, etc.) doing Book composition in the assembly language EasyCoder - sound familiar :-0 Once hired, I took over responsibility for 4 laboratory systems supported by some guy named George Timson who was heading for San Fran. What do I remember about the PDP-9 based MUMPS system I was working on - not much! Hell, I was just trying to keep my head above water working with the people working at Meditech in those days. I will say these things about it:

1) The laboratory systems were running on a PDP-9.
2) They provided 24/7 service to these customers over phone lines. The customers were Childrens Hospital (Boston), Cape Cod Hospital (Hyannis), the old Peter Bent Brigham Hospital (Boston), and Allegheny Hospital (Pittsburg). It had really good performance - but of course the 300 and 1200 baud lines made it look good on the other end.
3) The language was really simple but powerful.
4) Routine structure was a forced kind of structuring through parts (groups of lines) and lines within a part. It used decimal numbers as labels to group parts and lines. That is, the integer defined the part and the decimal identified the line position. Every line had a label. That is:

1.10 Set X=1
1.20 Set Y=2
etc.
If you wanted to insert a line, for example, "1.15 Write X" would place the line after the 1.10. You had to think ahead and sometimes restructure the entire routine if you did not. Now you know why the standard simply used line labels.

5) Of course, everything was done at the ">" prompt and with the editor ^%

6) Functionally, the language was very similar to subsequent MUMPS systems without all the added baggage.

It was a really great experience working with all those people. Of course, lets not forget we were all young and working in Cambridge in the 1970's - if you know what I mean?? Maybe that's why I can't remember anything :-))

Enough of that, let me know what I can do to help.

Terry
Author:  toad [ Tue Dec 07, 2010 8:14 pm ]
Post subject:  Re: Terry L. Wiechmann
Do you have any documentation on that MGH MUMPS (old pocket guide, reference sheets, etc.)?

That is, how much of a packrat are you (we're hoping for VERY)?
Author:  tlwiechmann [ Wed Dec 08, 2010 6:57 am ]
Post subject:  Re: Terry L. Wiechmann
There was no documentation when I joined Meditech. In fact, it was not permitted (company secrets and all that). You learned the language by looking at existing code and talking to the more senior programmers who had to learn the same way. It was rumored that all the documentation on the MUMPS systems were on a DEC tape locked in Neils desk. I doubt that it existed. Several years later at DEC, I looked at the Macro11 code for an integrated debugger written by Neil for MUMPS-11 under contract. There was not a line of documentation in it. Why write documentation when it's just as easy to read the code? Ya, right! I remember telling Neil that not having documentation was BS and that I was going to start writing some. He just looked at me and said OK.

No, I have nothing from those days. I'm the antitheses of a pack rat.
Author:  toad [ Mon Jan 03, 2011 9:53 am ]
Post subject:  Re: Terry L. Wiechmann
It's funny how many programmers think software is self-documenting, but of course it's not true. Software can at best document its own algorithm - what it's doing - but without comments no existing programming language can document why it does what it does. Without the "why" there's no way for a troubleshooter to know other than guessing whether the "what" is correct or not.

A great many bugs are introduced in the software world by programmers inserting "fixes" into code they don't understand, and at least half the time the fault lies squarely with the original authors for failing to document why their code exists and why it uses the algorithm it does.

Too many programmers foolishly assume the main reader of their code will be they themselves. It has become clearer to me over time that when an author is writing code, the proper audience to be writing for is specifically an inexperienced troubleshooter up in the wee hours of the morning trying frantically to fix the code while the hospital director is standing behind him or her growling impatiently.

I'm with you, Terry. Except under highly specialized (and mostly historical) circumstances, justifications for not commenting are inherently foolish, nothing more than rationalizations for laziness.
Page 1 of 1 All times are UTC - 8 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/