When I started studying mathematics at the Universitaet Mainz the most fascinating institution was the Center for Data Processing which owned a Honeywell-Bull DPS-8 mainframe running Multics. Since I have been crazy about large computers since my first contact with a Nixdorf 8870 when I was 14, I had to gain access to this machine - no matter what I had to do for this.

My friend Ingo and I decided to subscribe to a one week course entitled "Introduction to Multics". Since the library of the center for data processing at our university sold photocopies of Multics manuals for a few DM each, we had already bought whatever we could and read through the manuals. So when the course started, it was pretty boring, since Multics was such a wonderful and user friendly system - if you ever were stuck, you had all the necessary information just at your fingertips - who needs lectures when one can practice on a new operating system? :-)

As Ingo was always interested in numerical models of thermal conduction and things like that, one day he wrote a small FORTRAN program which calculated the thermal flow in a steel plate, when suddenly our course advisor had a look at his screen. Since the class was just learning how to use the ted editor, it was obvious that he was doing something totally different, which did not seem as a good idea when he was asked to leave his terminal and follow to the offices in the back of the building. After a long time - as it seemed to me, in reality only a few minutes had passed - he returned and told me that they suspected that the only reason for us to enroll to the introductory course had been to gain access to the Multics machine (which was right :-) ) and to steal computing time (which was definitely wrong).

After this accident we had to figure out something new not to loose access to our newly discovered toy. :-) So we decided to ask one of my mathematics lecturers if he saw a way for us to get access to the machine in a more official way. As it turned out, he was very (!) polite and helpful and suggested to create a real "project" with only Ingo an me as participants so we could use the machine on an official base without having to worry about bureaucratic administrators and operators.

Since at this time I was constructing a processor (which eventually led to the NICE architecture) he setup two project account for us with the aim to support the processor development. (Thank you! After so many years, I still think quite often of "my" Multics machine and all the wonderful nights we spent at the computer center programming FORTRAN, PASCAL, PL/I, doing useful and useless, but always interesting things on this wonderful machine!)

As you can guess, I really fell in love with the Mainz Multics machine, which can be seen here (from left to right are SCU, CPU, SCU, CPU and IOM):

During this time I made some photos of the Multics machine (I am sorry for the poor quality and will scan the original images as soon as I have some spare time to do this):

• IOM:
  • Front doors opened.
  • Close-up.
  
  
• Freestanding MSP.
  
• CPU-cabinet:
  • Front:
    • Full front view.
    • Close-up of upper half.
  • Back:
    • Three
    • picturesof the
    • backplane.
  
  
• Tape drive.
  
• DPU(Diagnostic-Processor-Unit).
  
• DPU(Frontend-Network-Processor).
  
• A look under the false floor:
  • Picture 1.
  • Picture 2.
  • Picture 3.