The computing environment in 1980 when I assumed the role of
director
consisted of the IBM 3033 mainframe system physically located in
the Math-Sciences
building. The Office of Academic Computing (OAC) ran the mainframe
in support
of all campus computing requirements. (There were two other
mainframes
on campus at the time: one located in Murphy Hall to support all
central
administrative activities and another located in the medical
center to
support those services as a completely separate operation.)
GSM housed the North-facility, which was a remote access location for OAC users. Access was available primary in batch mode (via keypunch or "dumb terminals) or interactive mode (via Decwriter terminals ). There were about 20 key punch machines, a card sorter, a card reader, a line printer (printing on green bar paper), and a large set of bins where output was sorted. There were a couple dozen "card" filing cabinets where individuals stored their computer programs, which may have been on hundreds of punched cards. There were about six or eight IBM 3330 video display terminals (green screen with white images). These were dumb terminals, used for online editing and submittal of programs, making debugging faster and easier.
Furthermore, there was a set of about 20 Decwriter terminals. These interactive terminal were installed in the mid-1970s at the request of GSM to provide support for the only interactive program then available on the mainframe, APL. The Decwriters were like a selectrics typewriter which enabled to you see your input on paper as you typed it in, and the computer in turn produced the output on the paper. The Decwriters were linked to the IBM 3090 by acoustic couplers. The acoustic coupler was connected to the Decwriters. Next to each coupler was a telephone. To make the connection between the Decwriter and the IBM mainframe, you dialed a phone number, and when the computer "answered" you pushed the phone hand piece into the acoustic coupler. When finished with your session (for which the clock was running remember), you disconnected your phone from the acoustic coupler and hung it back up.
OAC's funding was a dollar allocation from the Chancellor. However, it did not cover all of OACís costs. They were dependent on grant dollars obtained in the form of recharges to researchers. OAC converted the Chancellorís allocation into internal UCLA dollars which were issued to departments and school as RU (resource units). In principle, if one ran out of RUs, you needed to purchase them at the prevailing rate. (I do not know how that was established.) However the rate was set, that was what was charged to all grant recipients who used OAC resources. (the government A21 rule which says government must receive the best rate meant that OAC was motivated to make the rate as high as possible to capture the maximum grant dollar, which then put them well out of the reach of the academic department and users who did not have any grant funds and were totally dependent upon the RUs system.)
A major responsibility of GSM computing director during that period was managing the Schoolís RU allocation. GSMCS had several hundred accounts on the IBM 3090, with about 100 faculty, 100 doctoral student and the remainder student accounts. During that period, GSM offered a couple of different computer programming classes (APL, PL/I, and some simulation classes which used ???). all the students enrolled in these classes required individual accounts. APL students used the Decwriters while the PL/1 students used the punched card machines and submitted their program in batch mode.
The final line on your printout, whether it was from the batch job or on the Decwriter, was an accounting statement indicating how much computing resource you were allocated, and how much you used. (There were always three numbers: CPU cycles used, amount of disk storage, and amount of paper!)
The student were motivated to keep their sessions as short as possible as they were issued RUs and when they ran out, they had to go through a series of hoops to get more. It didnít matter if they used the online or batch submittal of programs, the system accounted for their use.
The mainframe was almost exclusively used for computational
activities
either user written programs in general purpose programming
languages
like APL, PL/1, or FORTRAN, or statistical languages like SAS,
SPSS, or
BMDP. The only "user" applications which ran at that time
was the
early versions of Orion, used by library staff to assist with the
back
office functions of the library. During the early 1980s the
first
user terminals to access Orion as an online catalog
appeared. One
terminal was located in the library. During 1981, a few 3270
terminals
were located in the North Campus facility, allowing for online
programming
input and editing. One could actually save a program online,
recall
and edit it. If you submitted the program and there were
errors,
the error messages appeared corresponding to the place where the
error
occurred and processing interrupted, making debugging faster.