Dave thought I should of spoken up at the meeting more. Well, I don't have as much to contribute as the speakers did, because my background meandered from *user* (as in 'loser' :-) ) to whatever I am today, but I thought I might write up some of what I can contribute. Some of this may duplicate things I rattled on about to James, the new member at Coco's, so if he is reading this, please bear with me, some of it may be new. Sorry if this is overloaded with personal war stories, but that's what I've got to contribute. The earliest connection to computing I can remember started with a child's book on mathematics I got for Christmas. Somewhere in the slim book was a brief discussion and some pictures of slide rules. I thought that was cool, so even though I didn't understand logarithms, I took some pieces of cardboard and by trial and error, made something that could roughly multiply integers. After showing this to an aunt, she bought me me a 6" slide rule, I think I ended up using in high school some years later. Next encounter was a crude toy mechanical 'computer' from Edmund Scientific (or Edmund Novelty as one friend prefers to call them. :-) ). I don't think I ever figured out how it worked, only had a couple of bits of information it could perform the crudest operations on. This was perhaps a forshadowing of things to come, since my degrees are in Mechanical Engineering. At Victoria Jr. College (TX), I took the introductory FORTRAN Programming class from Herb Sergent, who taught the engineering classes there for quite some time. Mark Mellis's pile of books comes in here in triggering memories for me. I also have a copy (somewhere here, I couldn't find it tonight) of Elliot Organick's MULTICS book. Part of the reason I picked it up from an MIT mail order book sale was that the text book at Victoria College for the FORTRAN class was 'A FORTRAN IV Primer', by Elliot I. Organick, which I've got my copy in my mits at this moment. While I never met Organick, he spent some time, perhaps a year at the central campus of University of Houston, my eventual alma mater (though from another campus). (Some people seem to have been destined by their names, like the axe murderer named Gore, Organick was of course a chemical engineer originally.) Paper clipped in my copy Organick's F IV book are: 1) 'A FORTRAN sort', a letter to 'Computer Languages' magazine, giving source for a Quicksort algorithm for FORTRAN 2) Electronic Design News 'Design Ideas', June 9, 1988, 'Fortran routine manipulates bits' 3) 'Elliott I. Organick 1925-1985', March 1986, Vol. 29, Number 3, Communications of the ACM One of the advantages of taking programing at the Jr. College was that unlike at the large universities, we actually got to operate the computer ourselves - putting in the start card, loading our own card decks into the reader. (At say U. of TX, TX A&M, U Houston central campus, they left off the card deck at the center, and picked up the cards and printouts later). I aced the class, probably because I was the only person in the class to bother reading the book, and understood the difference between equality and assignment. A very key point at the time. I don't recall the model of IBM we used VC, but it was probably obsolete at that time already. Somewhere around here, I may still have some manual for the thing. Also got my copy of the little Fortran book by Alexander and Messer that Mark was waving around here in my hands. We got to use some analog computers in electronics labs at TX A&M when I was there, and I did some of the dropping off of card decks at the Computing center while there. I still have the WatFOR/WatFIVE books they pushed at TAMU on the bookshelves. But I also made the mistake of taking the introductory electronics class in a 6 week summer session there, where most of the other students were returning armed forces electronics techs. :-) (I later made it a point to get and read the military electronics books reprinted by Dover.) Some time along this time, I bought a TI scientific calculator when they came out with one, actually ordering it direct from TI as soon as finding out about it. While in grad school at U Houston 'Victoria Campus' (or 'Center' or whatever you want to call it at that time. This was a major political point of contention, believe me. Eventually aspects made it into the Wall Street Journal. There are no politics as bad and nasty as academic campus politics. Stalinist show trials or the Night of Long Knives would be nothing compared to what these people would do if they thought they could get away with it.) Did a lot of studying of statistical mechanics. The campus never got enough people intrested to have a class on the subject. One book by, I believe ~'Knudson', had an appendix dealing with personal correspondence the author had with George N. Hatsoupolos. <tangent> Professor of Mechanical Engineering at MIT, Hatsoupolos wrote a large book, his magnum opus, 'Generalized Thermodynamics', summing up a lot things, I believe including many papers he had published in Institute of Radio Engineers, a predecessor of IEEE, on the thermodynamics of electron clouds. (I still haven't obtained a copy of this book. :-( ) Hatsopoulos then went on to found the Thermal Electron Corp., perhaps still a fairly large company, and became rich and successful. I think he was even briefly on NPR once, and made it into Forbes magazine. </tangent> In this private correspondence Hatsopoulos identified entropy as being proportional to the energy needed to sort particles by their (quantum) energy levels. Notice the use of the word 'sort'. While at UHVC, I was probably the last ME major to graduate without taking any Finite Element Method classes. Dr. Hedgecoxe had been the only full professor of ME *and* Mathematics in the whole UH system, and was a close friend with the local chancellor, Dr. Traylor, who was a topologist. Ended up covering a lot of Analysis, variational calculus and Raleigh/Ritz method stuff in his H's classes. He lived two houses down the street from my family. His wife had bought him one of the CompuCorp. programmable calculators for Christmas one year, and he put it to great use, grinding through tons of problems at night while watching TV. These things were huge for calculators, someone discribed them as 'not shirt pocket, but coat pocket sized.'. Only slightly smaller than a desktop adding machine. Sometime at this point in this narrative, I bought the TI calculator with the little magnetic strip cards, the printer/power cradle etc. Eventually, I ended up at Rockwell, where people signed up on a waiting list to get TSO terminals. I was amazed to actually be able to sit at an interactive terminal, oblivious to the fact that IBM 370 was probably already obsolete. They were still using them when I left in 1993, I think after Boeing finally acquired this division some one forced them to move to UNIX workstations. Rockwell apparently considered the Space Shuttle division something of a money cow, and didn't seem to want ot spend any money on it, being a 'cost plus' project. Also, since accounting needed the IBMs anyway, we were actually subsidizing this much more important department by paying for their swag CPU cycles, in effect an ironic precursor of SETI on your home PC. :-) The most important thing anyone showed me at Rockwell was when John stepped me through how to compile, link and run programs, each step by hand. This got tedious after a while, and I was primed for something called CLISTS when I noticed them in mentioned in the online documentation a lot. John also gave me a copy of the TSO command reference and said everything I needed to know was in it. Around this time, when the price hit $100, I picked up a Commodore VIC-20, and eventually bought some memory expansion cards at some kind of clearence sale in Cerritos. (Thumb rule: when any new technology drops to $100, there is no longer any reason not to jump in.) At a Toys 'R Us in Bell (Bell Gardens? off the LB Fwy.) I bought cartridge of Tom Zimmer's VicForth, which along with Ashly and Fernandez's 'Teach Yourself JCL' are what I credit with pulling away from the rest of the people at in our dept. at Rockwell in computer savey. Forth was great for learning about manipulating bits and bytes, stacks, strings, and linked lists. Eventually, the plant computer club, having contacts with other Rockwell plants, had a buy on RS65C02 chips, a CMOS reduced power consumption, expanded instructions set version of the 6502, and I put on in my VIC-20, where it still is. Eventually, I made a bizare mutation of Bill Ragsdale's Forth 6502 assembler for the thing, and had 3 Forth decompilers, macro processors and numous debugging/developement tools I could load into the VIC-20 with 32K of RAM *all* at the same time. While digging around, I found a copy of the notes for Bill Kizner's Automata Theory class at Rockwell. The real suprise for me from that class was that computers could be used for things besides calculating numbers! A very important paradigm shift. Before the meeting, Mark was amused when I mentioned that Jim Cook, founder/president of the Rockwell computer club and 6502/68000 expert bought out, at the Rockwell Surplus store, the remaining stock of Rockwell's foray into the abandoned magnetic bubble memory cards. Jim jokingly said he was getting ready to make sure his data would withstand thermonuclear attack. At some point along the way, I ended up working on CDC's in the department, and was once it's biggest user of CDC. While using this, I was learning from Tom, who was a CS graduate from U. of TX, and had quickly moved from being another IBM user to CDC wizard while consulting at Rockwell. As I recall he implimented an electronic mail system as an execise in NOS scripting, had everyone using a set of scripts he'd built up over time. Returning back to the IBM side of things, the A&F's JCL book finally cracked the mysteries of the TSO commands wide open for me, finally understanding most of what was in the manual John gave me. After occasionally doing some things that dropped peoples jaws in amazement (perhaps some people are easily amazed? ;-) ) I was at some point described as a 'CLIST specialist'. (I've forgotten almost all about that by now.) Some one said that Rockwell was the largest user of IBM mainframes on the West Coast at the time, but when I tried to track this down, I couldn't find anyone that would definitly stand behind the statement. Lesson on the need for archives/version control. They needed the results of a special version of the program the department/group revolved around. Only problems were that: 1) they'd lost the source code for this special version, and had only an executable. 2) this version produced only text output, with nothing in the special binary formated 'direct access' file format that we used for processing data in. After several weeks of argument, between the lead engineer for the study using the program (he said we needed a special program to reorganize the data) and the group supervisor (who said such a program would be impossible to write, and even if it could be it would take to long to make it operational.), I was approached on the feasibility of writing the program. After being shown the problem, my gut feeling was about 2 days. Using something I vaguely remembered from 'The Mythical Man Month' by Frederick Brookes (one of the books Mark was showing at the meeting) I figured three times 2 days would be an accurate estimate, and told them a week, which was acceptable. One day was spent on red tape getting an accounting number for this project, and after a day and half of working on it, I showed the results to the lead engineer. He found a problem with data signs, and explained how it had to be. Half an hour later I showed up with correct plots, and he was sputtering in amazement and vindication. Supervisor and his loyalists were not amused. We went to a presentation by the Information Services folks of Seal Beach, where they presented on the 'final update' IBM would supply for CLIST. That of course got my attention. They made passing mention of what IBM would recommend in future if you had any problems with CLIST - Rexx. Naturally, I went down to the plant library and ordered what I could on from IBM on Rexx, and read it when it showed up. After trying something roughly equivalent to 'Hello World', (I wasn't connected up with the hip mainstream to know that's what it was supposed to be....) I left a note in my supervisors office one morning that in the future, no more CLISTS, instead Rexx. This of course, got his attention, and I was asked some questions about the meaning of this. By this time, I'd moved from being the dept.'s token ME to 'computer guy', and was picked to 'port' our department's program over to the 'experimental' Cray installation. Part of this was taking the class on UNICOS I think it was called, Cray's version of UNIX. I don't think I will ever forget the name of Etain Serber, the fellow who taught the class. At the rates Seal Beach was charging, the Cray would be very economical, so I charged ahead. Richard, who was the #1 computer guy in the building, had worked out some JCL/Bourne shell scripts to do most of what we needed on the Cray (compile FORTRAN program, run FORTRAN program), so my first adventures on UNIX are probably unique in that I didn't get to do anything interactively, it was all as 'batch' background jobs. (Back to the spirit if not the physical reality of the card decks dropped off at the University Computing center. :-) ) The Cray essentually used the mainframes, CDC and IBM, as intelligent terminals. Because of bureaucratic delays about the time we got the Cray port running, Seal Beach announced that since no one had used the Cray enough, they were raising the rates on us so that it would not be 'cost effective'. Shortly before being layed off at Rockwell I remember we had a briefing on some new tools IBM had ported to the 370's. One of them was a program called 'make'. Some how the guy who volunteered an example program for the briefing decided he didn't see the point in make, and refused to go along with the demo. I felt somewhat cheated by this and let my feelings be known, as politly as possible. Anyway, when I was terminated at the Rock, I felt like it was time to move on anyway, and at the time thought it would turn out for the better. While working in Phoenix, I finally got to use UNIX interactively and get paid for it. The world did not end, Western Civilization continued. I met Bill and Esther Schindler while there, at the monthly OS/2 user groups meetings, where they are the driving force. (Someone should get their stories down sometime, but they may not want to go on the record. Bill had consulted at a now large large software co. when it was in NM, and Esther, a jouralist, had interviewed the widow/ex-wife of a onetime major player in microcomputer OS field, and both had some juicy inside stories, that contradict some popular myths, probably spread by the now large software co..) Toward the end of my year working in Phoenix, I took a the opportunity to attend their user group's conference, which was held jointly with the Rexx Language Association's, and there met/heard Mike Colishaw, IBM fellow, creator of Rexx, and expert on computer decimal arithmetic. One of Colishaw's talks touched on work he was doing with Sun/Java on a decimal arithmetic classes for Java. (Dave made mention, at least at Coco's, on how the original Eniac used decimal, not binary arithmetic.) This kind of brings my ramblings to a full circle in some sense, though hopefully I may add to this in the future.