Thursday, April 18, 2013

My Life With Technology - Chapter 1.1

by Bill Holmes

This is the second installment in my trip down technology memory lane. I had intended to write chapter 2.0, a time when I left the big city and big data center. Well, once I got started, I had many memories and made many notes that didn't make it into chapter 1.0 of this thread. So, instead of chapter 2.0 I give you chapter 1.1, a continuation of my days in Atlanta at First National Bank (FNB) in the main office.

I told you about the computer room equipment when I first started but there was a lot of stuff going on outside that room. There was a Tabulating (Tab) Department that processed several bank applications that were still on punch cards. Many people thought the Tab equipment was what computers were and looked like. TV programs in the 1960's often depicted a card sorter as a computer. All the Trust Department applications were processed in Tab, over a million cards on file. It made sense for the Trust applications to be some of the last converted to the computers. They were relatively low volume compared to checking, savings and installment loans. They were also relatively high dollar accounts and needed special handling. There were varying schedules for trust statements and reporting, much of it on demand. The Tab department was somewhere between the old manual personal method and the new computerized impersonal processing. The Tab guys weren't big fans of the young whippersnappers in the computer departments. They were all older than us, sometimes 20 or 30 years older. That doesn't seem that old now. They could see all the applications they used to run moving to the newfangled fancy computers. Since I mostly worked in Production Control back then I had to work with all the support departments. I'd take work into Tab that needed to be done quickly for another department that they didn't see as a priority. One solution was to do it myself. There was always a lot of idle equipment at night since there were only a couple of guys on the Tab night shift. Eventually the Tab guys got used to me and would actually do what I needed, help me do it or show me how to do it. Tabulating equipment proceeded electronic computers, those with vacuum tubes, transistors or printed circuits. It was electromechanical. The equipment was solely for the manipulation of punch cards and some of the original models dated back to the 1890's for use by the US Census Bureau. Between 1890 and 1950 the machines didn't change that much. They got faster and more reliable but the basic uses didn't change much.

IBM 084 Card Sorter
IBM 085 Collator
The main machines were a sorter (082 or 084), collator (085), reproducer (519) and tabulator (407). A sorter was a mechanical machine that could sort one column of a punch card into one of 13 pockets, 0 through 9, two zone punch pockets and a reject/no punch pocket. If you had a 10 digit account number you had to make 10 passes on the sorter, stating with the low order (units) digit to get the cards into numerical order. Alpha sorts took more passes. Collators could merge two decks of cards into one, kick out duplicates and a couple of other things. Useful for when you had new accounts to merge into a deck of older accounts or merge any two sorted decks together. A reproducer could duplicate (reproduce) all or part of a card plus a few other functions. The tabulator or accounting machine was the first computer. You could stick a handful of cards in it and it could add them up and print a report. They had programs to designate what was added, subtracted, multiplied or divided, what was printed and what was totaled and/or subtotaled.

IBM 407 Accounting Machine

IBM 519 Reproducer
The 407, 519 and 085 had crude "programs" that were large panels with changeable jumper wires (program instructions). I don't remember any 407 programmers at FNB, I guess all the main jobs had been programmed long ago and any ad hoc stuff could be handled by the Tab operators. They showed me a few basics of the "programming" language. I do remember walking through the programming department at the Prudential when I was a kid. Some of the programmers sat at what looked like draftsman desks and had panels and jumper wires piled up all around. Not what we think of as programmers today.
IBM 407 "Program"

I kind of liked the Tab department. The old salts who worked there were interesting. They were true application experts who knew everything about the jobs and equipment they ran. There was always the sound and smell of cards being manipulated. It was fun to put a big deck of cards into a sorter and see them fly into the 13 pockets. There was a rhythm and grace to handling large numbers of cards and the sounds of the machines. Tab was a dying department but they did play a vital role and even took on new tasks as we added new applications. More on that later.  

There was other old technology stuff in use. We did processing for many small banks. Those banks that were more than 100 miles or so outside of Atlanta would transmit their data to us via paper tape over voice telephone lines. The paper tape would be created at the originating bank by an attachment to a proof machine or a separate key to paper tape machine. Once created the tape would be transmitted to us in Atlanta. On our end was a machine connected to the phone network that received the paper tape data then we converted it to magnetic tape, a tiny reel of tape. I remember the noise and smell of the paper tape being created on our end. The paper to mag tape machines were made by Mohawk Data Systems and were like keypunch machines except they wrote to tape instead of cards and had a paper tape reader attached. Once on mag tape the data was processed by the computers. 

The banks we did processing for that were close to Atlanta would send the actual checks and deposit slips to us via courier. Those items were read by our Computerized Transit department on their check reader/sorters and transferred to mag tape for processing. I'll go into check processing in more detail in another chapter.
Our Delivery Zone

The farther away banks that sent their data by paper tape were processed on second or swing shift (4:00 PM to midnight). The closer banks that sent their checks to us were processed on third or graveyard (midnight to 8:00 AM) shift. At first each bank was processed separately on the IBM 1410. Later, after the application was rewritten, all banks for each shift could be processed together on the IBM 360/40. All the reports were sent back to the various banks by ground courier. The distant bank couriers picked up the output between 11:00 PM and 1:00 AM. The more local banks' reports were picked up between 4:00 and 6:00 AM. We could stretch those deadlines a little but the couriers had other customers too. We had great incentive to make the courier deadlines. If we missed them, we had to deliver the reports ourselves. That wasn't a big deal if we had to drive to Smyrna or Tucker, suburbs of Atlanta. The problems arose when we missed the courier for one of the far away banks. Those customers ranged from the Tennessee and South Carolina borders down to near the Florida border. From Toccoa to Bainbridge. If we missed one deadline we usually missed all of them. That's when we got in our cars and spread out over the state. Bainbridge was about a 250 mile drive one way. Of course we drove just a tad over the speed limit. Sometimes we'd get back to Atlanta just in time for our next shift. We almost lost a few folks who fell asleep at the wheel. I dozed off one time on I-75 but luckily veered off on an exit. The bumpy shoulder woke me up. On a few occasions we'd miss the deadlines again the next night and have to drive all over the state again. Thankfully we were young and indestructible way back then. Is it any wonder that many of us made more money in overtime than regular salary? We got mileage and meal money for those delivery trips too. I wish I could remember the mileage reimbursement rate and meal per diem. It wasn't much, but gas was really cheap until the 1973 oil embargo. You could probably get a Big Mac, fries and drink for $1.25. McDonald's didn't have a dollar menu back then because everything was less than a dollar.

In the computer room there were always changes. More applications were converted from cards,
Bus and Tag Cable
manual applications were automated, new applications and services were added by the bank and new technology was introduced. As equipment was added or replaced it required new cables. The power cables were 208V 60-Hz. The peripheral device control units were connected to the computer channels with a pair of cables called bus & tag. There were more cables from the control units to the devices, i.e. disks and tapes. These were big cables that weren't all that flexible. The cables circa 1970 were grey and were bigger and less flexible than the newer blue cables pictured. All the cables were under the raised floor which in retrospect was not raised enough. Often when old equipment was removed the cables were left under the floor. Some of this was laziness, but sometimes the cable was intertwined under and over other cables. Pulling a disconnected cable might dislodge one still connected to another device. After a while it became very difficult to add or subtract cables from certain parts of the computer room. There were many floor tiles that were floating on cables, rather than on the supports. If it was behind equipment the tiles might not be replaced. Probably not OSHA approved. Another problem was that the mass of cables restricted air flow and the missing floor tiles altered air flow so that was a constant concern.

The Culprit 
We once had an interesting mystery that plagued us for a while. We would get errors on tapes and disks. That wasn't unusual for the equipment and media back then, but we were getting more than the average number. We started gathering statistics. We put colored stickers on the media that had errors. Eventually somebody noticed that tapes and disks stored nearest the floor in the tape library had more errors. Why was that? It took some time to determine that. It turned out to be the industrial strength floor buffer that the janitor used. At least once a week he would buff the floor in the tape library bumping into the tapes and disks on the bottom racks. The giant electrical motor created enough of a magnetic field to cause the problems. The buffer was immediately banished from the library and eventually from the entire computer room. That's not the only problem we had with the janitors. Every now and then they would unplug something, knock something over or throw stuff out that wasn't trash. A couple of times they would wheel in their buckets and wet mops. Raised floors leak at the seams not to mention the cutouts and vents. Besides the data cables, the electrical plugs for the computers were under the raised floor. Over time the janitorial people were pretty much banned from the machine room.

IBM 2321 Data Cell - aka "Noodle Snatcher"
We had one very interesting piece of equipment for a short period of time. It was an IBM 2321 Data Cell. This thing was designed by the Rube Goldberg division at IBM. It was a big cabinet that held a carousel containing hundreds of strips of magnetic tape, each 2.25" x 13". These strips were grouped in cartridges that could be removed. The big carousel would rotate to get the correct cartridge to a read/write head, the proper tape strip was removed and pulled under the head. One 2321 unit held 400 or 800 MB of data. This beast was big, mechanical, noisy, slow and prone to breakdown. Other than that it was perfect. We didn't keep it for too long and I don't remember what we used it for. IBM never made any follow on models of this technology. It was fascinating to watch it in action, when it worked. Because it used strips of tape that were grabbed by mechanical arms it was nicknamed the "noodle snatcher".

Amazing times with amazing technology. Hard to believe some of this stuff was "state of the art", hard to believe some of it actually worked. I'm glad I don't have to wrestle with it anymore, but I'm also glad I went through it.

There will be at least a chapter 1.2 and maybe a 1.3 before I leave Atlanta and move on to chapter 2.0. Next I'll document our adventures with a brand new major application. One we weren't prepared for and had to learn on the fly. Stay tuned.



1 comment:

  1. This comment has been removed by a blog administrator.