[TNO-logo]
Museum logo

The Control Data 6400 (CDC 6400) (part 2)

Anecdote: miniskirt problem

In 1978, a new and much faster printer was installed. That printer had less "dancing" characters on the printed line and made much less noise. This new 580-printer had an automated "table" (stacker) that folded the output paper stream. As usual when installing new equipment, a small model of the printer was moved around the "floorplan" until a place was found that looked good both from operational view while reguarding all technical aspects as maximum cable lengths. Before installation, holes were made in the tiles that comprised the elevated computer room floor and the required power and signal cables were prepared (op photo with both printers). Only one small detail was overlooked. The "door" that hid the output stacker of the printer had its hinges hanging at the other side as the old printer had.

Because computers were a "hype" and very interesting at that time, one of the (only limited number) of secretaries really wanted to become a "computer operator". After a short course, she worked half-days as an operator. Being young and following the latest fashion, she liked to wear miniskirts. The, in that time mainly male, Laboratory workers, including many military at the laboratory stationed people, like that too....
It turned out that the removal of paper from a line printer is ergonomically a less obvious task for someone wearing a miniskirt. To decrease the fast growing "social meeting" at the input-/output desk, a fast decision was taken to turn the line printer 90 degrees in order to reduce the "view" for the laboratory personnel. That her co-workers, the operators and system programmers, got 'better working condition' at the same time was something that she never noticed.

System software

The Scope 3.4 operating system, later renamed into Network Operating System/Batch Environment or NOS/BE, was delivered in the form of source code on a number of reels of magnetic tape. Compilers and run-time libraries were delivered in source code as well. Systems programming at the Physiscs Laboratory (PhL) regarded it as a sport to be the first in the world to have a new version ("level") of the operating system in "production". Aside of the advantage of new features, sometimes urgently required by the users, the disadvantage was of course that we ran into all the small system bugs related to the new software. The diversity of these problems and system errors required a fast solution. The alternative was going back to the previous level of the operating system. In order to solve the problems fast, special skills were required to analyse problems fast, to built a hypothesis, to generate code to solve the error and to verify the correctness of the solution during the "happy hour" for systems programming (between 17.30 and 18.30 hours).

The Laboratory had made the choise for the 63-character set, like nearly all of its college computer centers in The Netherlands. In the US, Control Data only tested NOS/BE in the 64 character set mode. Obviously, programmers overlooked the 63-character set particularities. Thus less good developed code or code by "new" programmers resulted in not-working modules or compilers at almost each new release. We, at the Physics Laboratory corrected these problems and made each time a lot of noise about the failures in coding practice. Problems were reported using the Problem Reporting System-mechanism (PSR). With a two week interval, a set of microfiches arrived that included all reported problems - and solutions - collected throughout the world. With each release level, the system programmers anxiously waited a couple of weeks for the microfiches in order to figure out whether we set the mark with the new problems found or that our collegue of the University of Arizona had beated us with testing the new release....

Year

Our PSR reports

Incl. code (solution)

Year

Our PSR reports

Incl. code (solution)

1975

67

31

1983

69

48

1976

56

48

1984

36

20

1977

56

45

1985

231

54

1978

70

44

1986

216

21

1979

134

49

1987

252

14

1980

86

75

1988

201

4

1981

63

51

1989

136

0

1982

69

29

1990

78

0

The software problems as reported by the Laboratory (site code PLTN).

Apart from correction code for system errors, many additional feature code was developed. Much coding effort went into developments that eased the work of the console operator. As the standard NOS/BE operating system sometimes required information displayed at two or three screens, we tried to depict all required information on one single screen. Just to avoid typing errors, most commands that required the full job name of seven characters were replaced by commands that required the two or three digit "ordinal" number instead. In other cases, we supplied automatically the remaining characters. In this aspect, the Laboratory was years ahead of what we call nowadays an ergonomic workplace.

Anecdote: "The Director received a strange letter.."

The Programming section of the Computer group designed and maintained a number of programs for the Armed Forces. For exercises in the command and control bunkers of the Royal Netherlands Army a program was designed to generate randomly so-called tri-grams. Tri-grams depicted certain communication channel settings. To keep the work of the draft soldiers "dienstplichtigen" 'interesting', the "randomness" of the parameters was varied in such a way that the maximunm number of (dirty) three-letter words was reached in the deliverables. The soldiers liked that until at one day a General held inspection at the command and control communication bunker. "New tri-gram ?" "K..T".... The general: "Who the hell said that ?!"

"General, this list says 'K..T", as you can see." The Director of the Physics Laboratory was then surprised by a letter of the General stating his thanks for the "real random tri-gram list (!)", but "could it be less random next time?". Explicitly, the General stated which three-letter words like "k..t" and "l..l" had to be removed from the list next time. Probably, this has been the only letter to TNO ever stating dirty words!
Of course, it bacame a sport for the programmer to generate lists with other "offensive" tri-grams for his "customers", the soldiers!



Museum logo