In May 2000, Ron Bowles (former CSIRAC engineer) returned from Adelaide with some material donated by Frank Hirst (formerly Head of the Computation laboratory).
Of particular interest was a listing of the Interprogram source code of a program, together with a corresponding 5-hole program tape of the same program, as produced by Interprogram. No data for input to either program, or output from them was available up to that time.

The 5-hole interprogram obtained from Frank Hirst was the only known example of a 5-hole Interprogram tape, which made it of considerable importance.

By examining the Interprogram source code, it was possible to create some data, which was added to the end of the Interprogram source code. (The program asked for data to be input, so data had to be created in the hope that it would work). The Interprogram T754 was run under the CSIRACEM (emulator program) to process Frank Hirst's Interprogram source code. Results that looked sensible were produced.

Up to this time, the Interprogram T755, which was known to be able to produce a 5-hole program as output, would not run as it contained errors due to having many corrections (masking tapes over holes) on the paper tape, and also was later found to contain a number of logical errors and could never have run.

Ron Bowles and John Spencer laboriously read by hand the entire 5-hole program tape (from Frank Hirst) as Flexowriter code, and compared notes as they did this. Ron was able to explain in detail how the bootstrap at the start of the tape worked. After converting the Flexowriter code to ASCII, and modifying the CSIRACEM (emulator program), it was eventually possible to run the 5-hole program in standalone fashion so that it produced the same output for the same data as the Interprogram. After considerable modification and correction, and with Ron Bowles' help, the Interprogram T755 was able to produce a 5-hole program tape output, which agreed with Frank Hirst's tape.

Without the Interprogram source and corresponding 5-hole program obtained from Frank Hirst, it is unlikely that T755 could ever have been made to work.

More Information