Phoenix, US
|
Ian:
I'm pretty sure this was an IBM product. In those days, late 1950's to early 1960's, all of the mainframe manufacturers (IBM, Univac, General Electric, and others) were moving as fast as they could to come up with large capacity storage systems: meaning a whopping twenty million characters or more.
I still have copies of the programmer's manual for the 305 RAMAC, including photos of all the components. I was given the task of preparing all of the sectors on our machine, which involved rewriting each of them. During this multiple-hour run one night, the unit failed, and the service technician sent out to fix it was an IBM employee I got to know very well over the next year or so.
I remember seeing a large Univac mainframe in their main office in downtown Detroit while attending IBM 650 programming school in 1959. As I recall, it was all magnetic tape drives.
IBM also had an external disc drive for the vacuum tube 650 that preceded our 305 RAMAC, but our 650 was not so equipped.
Following the 305 RAMAC, in about 1961 or so, we migrated to a GE 225 which included a GE newly designed and manufactured disc storage unit that was about the size of the RAMAC unit, but had only 16 large discs with a separate read/write head for each disc, and held approximately 20 megabytes. I wrote the I/O routine for that unit, as the GE Computer Department did not supply one with it. To minimize track to track head travel time, I organized the sectors in stripes, like is done today.
One day I received a call from the machine operator who told me he was getting increasing numbers of error/retry messages from my disc I/O routine. Long story short, we had experienced a head crash, and the inside surfaces of the curved glass panels enclosing the discs were covered with the red oxide material used in early systems. Once the first head crashed, others followed as the fine red powder worked its way through the entire stack. We were down for several days while four or five new discs were shipped to us from Phoenix.
Once the new discs arrived, I helped the service tech install them. Our first step was to remove all unaffected discs, take them out in a parking lot and wash them with soap and water. Then we waxed them all, including the new ones, before installing them in the drive. To my great surprise, the system came up and worked perfectly for several more years.
Ironically, just prior to the head crash, I completed a disc-to-tape backup routine and gave it to the operator with instructions for periodic use. For some reason, it was never run and we had to rebuild all the files from scratch.
In my observation, programming stopped being fun when operating systems appeared on the scene. At that point, we lost our sense of intimacy with the system as all the "fun stuff" was done by the operating system. Prior to operating systems, a programmer could tell where the maching was in the code by watching the patterns in the square yards of blinking lights on the panel.
Those were the good old days of computing.
Thanks for your comments Ian, you brought back a lot of great memories.
Regards,
HBB in Phoenix, Arizona
Nikonian Team Member
Photography is a journey with no conceivable destination.
|
