If memory serves me right I think that machine you are referring to was something like IBM 4200 or 8402. It was a dark gray, pepbble finish card shuffler and punch with soft wired logic boards. Programs were changed by interchanging pre-wired boards. I did some early work on a system like that for a large photo studio in Chicago. The system was a fairly comprehensive job cost and billing system. this company would do photography for many newspaper sale pages and magazines, plus it did several large catalog companies.

I replaced this system with a 32K, 10mb monster in the mid 70's. Once again the Job cost, billing A/R A/P PR GL were written in Assembler. The 1 mhz system ran like a scared deer. Data entry was performed on either a teletype like console or if it was available a CRT. Debugging was performed using sets of switches and indicator lights. As the machine ran the banks of lights tracked the current address location, content byte value at the location, we could modify any address value and change the "PC" program counter to change the next executed instruction. It was easy to patch these programs using the toggle switches by modifying the instruction at any location. I remember that an unconditional branch was a 38h, followed by the 16bit destination. Using the control panel we could single phase, single step, stop at location. Many times in lieu of source code change / recompliling / linking it was easier to click in an address, set the step swith to halt at addr, load the program, press start again, and then when the system halted a second time key in the "patch". finally when time permitted you changed the source, and went thru the various incantations and rituals to generate and install.

VFP is nirvanna.

Glenn


>>Who else out there has any good stories of when programming had the additional challenge of limited (and I mean really limited!) capacities of the tools they were working with.
>
>Me! Although you probably know them...
>The first machines that I programmed (payroll, cost estimation applications, etc.) had:
>- a working storage of 15 registers holding numbers only (I believe the maximum was 15 digits, including the decimal point)
>- the main programming device in form of a 9"x6" board with holes and gilded contacts. Programming was done by sticking copper cables in holes, from one hole to another, using a tool like a screwdriver. The result was truly a spaghetti of cables.
>- the machine was the size of a desk and weighted over 100 kg.
>Getting a machine that would read and punch paper tape was a big step forward!
>And all this was not that far in the past - 1978.