Last update September 3, 2002 for minor link changes
This is just a sample of the kinds of programming work I did while I was still a student. It is not a complete list.
In my Freshman year at University Of Houston, I had a part time work study job as a remote job entry (RJE) clerk in the School of Business at the University. I would run the card readers and printers, monitor backup, and other minor stuff. I can still leaf through a fanfold stack of paper fairly quickly.
In 1982, while a Sophmore, I got a job as a Cooperative Education Programmer at IBM Federal Systems in Houston, Texas. My whole point of going to University was to get a job in the space program, so it seemed like a good chance. Besides, I made as much in one semester as I would have had in a whole year of Loans, Grants, and Work Study Jobs. Since I was older (24) than the average Coop, I was given more responsible assignments over the time I was a Coop.
My first position was for the proposal team for the Space Operations Processing Center (SOPC) that was to have been built in California to launch Space Shuttles in polar orbits. I had one of the first IBM PCs on my desk, (one of the first 3 in the building). It has 2 single sided disk drives, and "expanded" memory (128K). I wrote programs in Pascal and Fortran. I remember one to keep track of documents for the proposal. I was really surprised how many IBM employees had never used a computer. I also used a program on a VM mainframe to specify requirements and check their consistency.
In the summer of 1983 I worked in a department that was doing some research for a possible future Space Station. They were doing some analysis of the guidance programs from Skylab that were written in FORTRAN H (IV). I worked on a program in PL/I to parse the compiled listing and create a "database" of cross references across modules. This was when I began using MVS, ISPF, JCL and Clist as well as PL/I. I only was able to ask about 5 minutes of questions every few days, so I learned to read the manuals. I worked with a mentor, Charlie.
I did so well in the Summer of 1983, that in Spring 1984, I worked for a 3rd line manager on the Space Shuttle Software project. I reported to him or his staff department for the next couple years. At that time, there were several 2nd line organizations that each had their own TSO/ISPF setup on the MVS mainframes. I was working with the lead programmer on an effort to consolidate across the organization to have one common setup, so that if someone created a helpful utility program, it would be available to everyone.
The effort was called the "Panels Working Group". It was a first attempt to have some configuration control of the backroom tools for creating software for the Shuttle. Work had already been in progress to create a system for the Flight Software, simulators, and the Configuration Management system itself. I worked on a program with Charlie to allow programmers to schedule programs, ISPF components, Clists, etc. for a daily "build", or copy into the production libraries. I also ran the program to do the daily build.
I maintained these programs until they were finally replaced with an Oracle based tool in the early 1990s. I maintained the ISPF setup off and on until I left in 1997.
Charlie and I wrote an application to read ISPF panels and generate includes for Script documents to aid in creating user documentation. It was a preprocessor that would read the SGML (and later Bookmaster) document, and panel library. When our special tags were encountered, the panel source would be read, specified variables would have their values filled in, and the source would be ready to run through the normal Script process.
I also worked on some project resource planning tools. At that time, you could not go into a store and buy "Schedule Publisher", or "Microsoft Project". So, we wrote out own. Charlie created the main engine and graphics procedures, and I added error checking, boundary checking to be sure where to put titles and legends, and the ISPF interface. It would draw schedules with milestones and dependencies based on textural data input on a panel. I set up the storage methods. We did not have low cost commercial databases that did not require DBAs, so I helped design a minimal database for this application. We had a couple of these kinds of applications, the one I mentioned here was still used in late 1997 when I left, although there were plans to convert to "Schedule Publisher".
IBM periodically does a survey of their employees to check how things are going, and how good management is doing. At one time, it was a very big effort to collect and tabulate the data. Some people in Federal Systems Division decided to automate the process. I think it was sometime in 1984, that I wrote a prototype program to show that it was possible. It read a table (from SQL Plus, predecessor of DB2) and tabulated the results. The results were keyed centrally in at that time, and in later years were captured from a survey electronically. There were rules about how results could be reported, to ensure anonymity of individual responses. If there were enough answers, a percentage distribution of responses could be reported, if the number for a question went below a certain number, only the percentages of Favorable, Unfavorable, and Neutral could be reported. If there were fewer than another number, the distribution could only be viewed as a combination with other departments.
The program was made a corporate wide product, still with parts of the algorithms based on my design, and ownership was transferred from Federal Systems. It is still used today.
I also wrote some subroutines for a program that would use a SQL Plus database to send reminders to managers to do evaluations for their employees.
Back to Detailed History
Al's Home Page is at http://www.alprosser19.com/
Comments to email@example.com