A few career details (still under construction)
Tab Machines (watch this space)
Super-Analyst During the first 15 or so years of data processing, the tasks of problem analysis and solution design were separated from programming and installation. Analysts usually had a prgramming background, but concentrated on design rather than doing the whole project. More recent times have seen the usual merging of those tasks into programmer/analyst positions. There are advantages to each arrangement, but my own preference was for the separated positions. I believe the specialization allowed for better results. When combined positions are used, the temptation is to design for ones own programming skills, which does not necessarily lead to the best design. I did the cartoon below during a time when I was leading a project, doing design that was being programmed by others, in a language I was not skilled in - FORTRAN. The cartoon illustrates that, even with language ignorance, the analyst can very specifically identify and define a program bug.	Click on the thumbnail below to enter the exciting world of Super-Analyst.
The manual, hand-held Bit-Winnow... to beat massive computers In the early 70's, I designed a computer system to match high school graduates with colleges, based on criteria set by the colleges, and preferences and qualifications submitted by the students. The application is a great example of a BAD centralized computer application. Quite often, the students' results were NO matches or POOR matches, because their criteria were too limited. That meant that if the student wanted to make another try, the turnaround time might well be a month... getting the poor results, submitting new information, then waiting for new results (that might still be poor). Understand that this was still well before the advent of personal computers that would eventually allow results to be seen quickly, and allow another attempt immediately. In an illustration of general-purpose ANALYSIS, I designed a simple physical device that would solve the college matching problem in a manner far superior to massive central computer processing... by putting it in the hands of the students themselves, and allowing them to easily make many selection attempts. The Bit-Winnow could have been cheaply produced for an application, such as college selection, and distributed to high school counseling offices, putting the "intelligence" and "processing" where it was needed. The device used logic that was similar to an older punched-card system (KeySort Cards), but was unique enough to have been patented. Within the next few years, personal computers could have done what I designed Bit-Winnow to do, so it's probably good that I didn't try to implement it commercially.	Click on the thumbnail below to view patent drawings for Bit-Winnow.