Mainframe near history IBM 3380 and 3880 docs
Looking for Johnson's pcc DG Nova and Eclipse 3471
tcc looks a little too 32BITish and a little too X86ish for my taste, after a very quick look. The reason I would start with...
Looking for Johnson's pcc DG Nova and Eclipse 3472
I can probably remember it. If I ever get that far. Besides that it's long dead :-) Multi user time sharing system. Apps programmed in interpreted basic. System written in buttembler. Slim, for several reasons...
As often as not for us, it was more like whirl, kerchunk, clatter-tick-tick-clatter, kerchunck, whirl, kerchunk, clatter-tick-tick-clatter, kerCRUNCH with lots of red lights.
The way it worked was that each strip had a pair of tabs on top denoting its position within the slot in the bin. When a given strip was desired, two sets of fingers came down and pulled the ones on either side of it away, leaving the selected strip sticking up. The read-write drum had a pair of fingers on its outer surface, and it was rotated such that the fingers came down and grabbed the strip. Then the drum reversed direction and spun for the read-write process. When reading and writing was done, the drum simply stopped and reversed direction with the idea that it would throw the strip back into the same slot whence it came. Sometimes it actually worked. When it did, the fingers on the drum released the strip and were retracted back into the surface of the read-write drum.
When it didn't work, the strip, with your precious data on it, was almost always as neatly pleated as one could hope for. There's a good possibility that some of the fingers got bent in this process as well, because the pleating resulted from the strip landing on the selection fingers instead of in the hole the selection fingers were trying to maintain.
History of Programming Languages III was AFC Book Listing, Big Update 3474
Eric Smith Actually looking at Dennis Ritchie's HOPL-II presentation I think the best interpretation is his conclusion on "how to succeed...
Another point: if any of the red lights came on, the operators *could not* get into the device to get what remained of the data out. Took a 2321-trained engineer to unlock the door, and he usually had to disbuttemble the picking and head mechanisms to get what remained of the crunched strip out (which usually had a very neat accordion pleat by this point), which explains why we had one either directly on site or available within half an hour 24-7 ...
In our case the "whirl, kerchunk, clatter-tick-tick-clatter, kerchunk" sequence was repeated *twice* for every access to a master record in our *batch* daily processing. We used to condense it as "pick-turn-pick-turn-pick-turn-pick-turn-pick-CRUNCH". Captures the pace at which we were accessing this device better.
Alt Folklore Computers Newsgroups