Tuesday, May 21, 2013

My Life With Technology - Chapter 2.1


by Bill Holmes 

After spending a few weeks helping to get the remote data center up and running in Dalton, Georgia I was on the move again. I had a couple of weeks back in Atlanta and then I was off to Valdosta, Georgia. That was a move from one end of the state to the other. Dalton is on the Tennessee border and Valdosta is straight down Interstate 75 on the Florida border. In 1970 Valdosta had a population of approximately 30,000. There were an additional 25,000 in the county. I lived in Valdosta close to ten years and went through several technology and life changes. Let's get started with getting started.

I only made two trips to Valdosta before the permanent move. Although I'd driven by it many times while going between Florida and Atlanta, I never stopped except maybe for gas on the interstate. The first time I went with my boss to meet the data center landlord and a few of the officers at First National Bank of Valdosta. They were going to be our first and largest customer. I much later learned that the meeting was actually my audition. If the bank leaders didn't like me they had the option to ask for another data center manager. I guess I passed the test. My second trip from Atlanta to Valdosta was primarily to find a place to live. My third trip was the final move. 

I found an apartment that was a big upgrade from my studio in Atlanta. My new place had vaulted ceilings (actually an uninsulated roof), a separate bedroom, a real kitchen and an area for a dining room set. There was one problem, I didn't have any furniture. I had been living in a furnished apartment since college. I owned a small TV, radio, some kitchen stuff, towels, bedding and clothes. Everything I owned easily fit into my luxurious yellow 1968 Buick Opel which was just slightly bigger than a VW Beetle. To get started I bought a card table and two lawn chairs for my dining room set, one real chair for the living room, a couple of TV trays and a lamp. My parents had a folding bed that they gave me. I also did what every self respecting single guy did during that time. I acquired a couple of cement blocks and some 1 x 10 or 1 x 12 boards and made a TV stand. As you can see I was living the high life.

For the data center we leased the second floor of what had been the FNB Valdosta building. The bank had built a new building two blocks down the street that had a parking lot and drive-in teller windows. The first floor of our building was occupied by a stock brokerage. The landlord was the manager of that brokerage. Our space was about twice as big as we needed but I think there had been some good old boy dealings involved plus it was probably cheap. Second floor office space in downtown Valdosta was not in high demand. At one time our space had been the bookkeeping department, the employee break room, a bathroom, some storage and the board room for the bank. The old board room had wood paneling and plush gold carpeting. We didn't even use that room at first. I had a few weeks to get the office ready. We had to have electrical and HVAC work done. The bank had recommended vendors  for those jobs and we followed those recommendations. We had to have the data line and regular business telephone line installed. I set up an account with the local office supply store. First National Bank of Atlanta (FNB) sent us surplus office furniture and a keypunch machine. The IBM computer equipment and the specialized supporting equipment, supplies and forms were ordered through FNB's purchasing department. I also had to interview and hire the staff and a janitorial service.   

This building had an elevator but it was old, small, not maintained and not included in our lease. The stairs were narrow and had a 90˚ bend. Somehow we had to get big, bulky, heavy and expensive equipment up to the second floor. The solution was to remove one of the windows and use a crane. The windows weren't wide enough, so part of the wall was also removed and a new window frame built. Of course there were utility poles and lines between the street and the building. The equipment would have to be hoisted over the power lines and then lowered to the window height. There were five pieces of IBM computer equipment. The 360/20 CPU, 1403 printer, 2501 card reader and 1419 check sorter that was shipped in two pieces. There were also a couple of big boxes for the cables. The CPU and sorter parts were the biggest and heaviest. To do this crane job we hired Bubba and Sons Incompetent Crane Service on the recommendation of the Valdosta bank and/or the landlord. The computer equipment was delivered to our site on a flat bed truck. The Bubbas attached a rectangle metal platform with railings on three sides to the crane hook. When the platform was brought to the window, the side without the railing was rested on the window sill, the platform tied off and the equipment rolled into the building onto a small wooden ramp that had been built. They got the card reader and printer in without too many problems although there were some close shaves with the utility lines. The fun began when they loaded up the CPU. They got it to the window opening OK and tied off the platform. When they got the unit part way into the building the platform kicked back toward the street. Luckily it stopped before the CPU went crashing to the sidewalk. We now had a 360/20 balancing between the window sill, a small ledge and the platform. Oh, did I mention that Bubba, Jr. rode the platform up and the CPU had his leg pinned. Of course he didn't have a safety harness on. After much discussion and cussing the Bubbas decided to get another smaller crane to raise up the front of the platform. By this time we had a pretty big crowd watching this drama. We also had the police and fire departments on the scene. The building was at the intersection of two major streets in Valdosta. Amazingly the Bubbas got the platform righted and eventually got all the equipment inside the building. That small ledge was clad in some kind of metal sheeting which was now dented and crinkled. The 360/20 had a bent cover panel and some gouges in the bottom frame. The IBM Customer Engineer (CE), that's what IBM called their hardware maintenance folks, had already called his bosses to find out how to file the proper paperwork for a smashed CPU and order a replacement. What should have taken a couple of hours wound up being an all day affair. I wish I had pictures but this was way before cell phone cameras. I'm sure there are photos somewhere. I looked at the building on Google Maps Street View but all evidence of our adventure is gone. The utilities are now underground, the windows and building facade have been renovated. 

Even after we got all the equipment inside we weren't home free. It took the CE a couple of days to get everything installed and tested. As it turned out it all worked OK in spite of the bumpy ride. No harm, no foul. Bubba, Jr. had a bruised leg but no serious damage except maybe to his pride.

The data line installation went smoother than it had in Dalton. It was all Bell/AT&T so there was less finger pointing. We got a real data phone (grey) rather than one modified on site. We still had the 4800 baud microwave oven size modem that we supplied.

In addition to normal office supplies and equipment, like adding machines, we had several pieces of specialized computer and check processing support equipment.

Carriage Control Punch & Tape

Carriage Tape in Printer
Printer Ribbon
Impact printers like the IBM 1403 used carriage control tapes to determine the length of the forms. The tape is a loop of reinforced paper with holes punched in it that indicated where the printer should stop after a carriage skip, as opposed to a carriage space. Stock fanfold paper (green bar) might only have a top of form punch in the tape, typically channel one. Other forms might have total and subtotal lines that could be skipped to. There were 12 channels on the tape, so a form could have 11 different skips plus top of form. The "skip to channel n" was under program control. Carriage tapes were typically the length of the form although for some very short forms we would double or triple the form length and duplicate the punches. The tapes had to be punched, cut to length and the ends glued together to form a loop. There was a special piece of equipment to punch the holes. There was special IBM carriage tape glue too, probably very inexpensive, but any strong glue that didn't get brittle or tape would work. We only had about four or five different forms so only about four or five different carriage tapes. Large data centers, like FNB Atlanta, may have hundreds. These printers also used special print/ink ribbons. The ribbons were about 15 inches wide and I don't know how long. The ribbon was continually moving and reversed direction when it reached one end. They were pretty expensive and you had your choice of black ink.

Check Tray
We also needed check trays and a check tray cart/rack to put them on. Our check trays were plastic with a sliding metal partition to hold the checks tight when the tray was less than full. They were similar to punch card or mail trays. The rack was placed next to the check reader/sorter so we could easily move the trays around as we processed and sorted the checks. Our cart only had three shelves, not the six in the picture below.
Check Tray Rack/Cart

We had a small manual desktop MICR printer so that we could make batch separators and end of run items.  Batch tickets (separators) were placed between each batch of items. The batch tickets were bright yellow and oversized. They had a special routing number preprinted on them that indicated to the sorter program to pull batch totals. We would print the dollar amount of the batch on the ticket. The sorter program would calculate the items and compare it to the batch ticket total. The difference would be printed, either plus or minus. If it was zero the batch was in balance. If not we figured it out. The end of run tickets signaled the program to pull totals for the last batch, total run and end the program.

There was one other diabolical piece of equipment for check processing and that was the jogger. A check jogger is a high powered vibrating rack. It was used to get the checks ready for the reader/sorter. It would get all the bottom right edges, the leading edge, of the items lined up so the sorter could feed them more easily. The compartments were approximately handful size. The area where you put the checks was wood to reduce the static electricity and maybe reduce the noise a little. I call it diabolical because it was so noisy. It even had a dial so you could adjust the intensity of the vibration and noise. It may have even had a This Is Spinal Tap 11 on the dial. The joggers, check sorters, printers and computer hum are the reason I now have tinnitus and hearing lose.

We also had a forms decollator which removed the carbon paper, remember that, from multi-part continuous forms. We printed most of the reports on multi-part forms. The main account balance report (trial balance) was printed on four part forms. We removed the carbon paper before we sent the trial balance to the customer banks. Other reports were printed on two or more part forms. A couple of forms we left intact, depending on the bank, because they would make notations on them. We had a simple two part decollator so if it was more than a two part form we had to make multiple passes. There were fancy decollators that could separate up to six part forms. They were great when things went well, but usually at least one carbon paper ply would break or one copy of the form wouldn't stack correctly. You would have black hands and maybe a smudge or two on your face after decollating a few boxes of forms.


Eventually we got all the equipment in place, the data line working and a staff hired. We made some test runs with fake documents and some FNB Valdosta items. The staffing was three people, including me, on second shift (3:00 PM to 11:00 PM), and one more on third shift (11:00 PM to 7:00 AM). More on that later.

There was another part to the startup of our data center, that was to get the account numbers, balances, service charges, fees and other information loaded into our systems in Atlanta. That was either done manually or programmatically in Atlanta. It depended on whether the bank had been processed by another provider or was still manual.

I realize that this chapter doesn't include a lot of new 1970 technology but it does include some adventures and points out some supporting equipment. Every piece of equipment wasn't transistorized or program controlled but they were all important.

Next chapter we'll fire up the new Valdosta data center for real. We'll get going on the journey that took me through the next ten years of my technology career. While I never again almost dropped a computer from the second floor, I assure you there were other exciting adventures. We will be stationed in Valdosta for several years but there will be many technology and other changes.


No comments:

Post a Comment