I checked the clerical staff and found that one division chief
had two secretaries. One worked half her time on data processing
paperwork and the other half on the division chief's outside business
interests. The other spent full time on his outside business interests.
I abolished both secretarial positions and told him my secretary would
handle his paperwork in her spare time.

	I found that we had two individuals in that division chief's
organization who were on temporary promotions to GS-12 and that the 
Chief, Position and Pay management Branch, Civilian Personnel Division 
was extremely unhappy about the promotions and ready to cut them back 
to GS-11's. Both had GS-12 jobs lined up in Huntsville. I got with the 
Chief, Position and Pay Management and Chief, Manpower Office and worked
out an agreement to divide the Programming Division into three
programming divisions with a GS-12 over each one.

	The Chief, Programming Division came to my office and said he
worked for prestige, not money, and if I changed his division into three
divisions and he only had a third of the people he would go to a job he
already had lined up in Huntsville. I handed him my phone and told him
to call and arrange a transfer date and I would approve it. He called
and they asked how much notice I wanted and I told them none. He told
them he needed thirty days.

	I went to New Orleans to a programming school. Helen Saxon, a
computer operator, and Stan Magness, a budget analyst for the
comptroller, went with me. We learned almost nothing about programming
except it was a difficult job to do right. Tic Vinson, one of the new
programming division chiefs, called me while I was there and said the
division chief that was going to Huntsville had gone to the military
director for supply and maintenance, Major Elvington, and told him I was
ruining Data Processing and that the Major told him he would fire me. I
called Major Elvington and asked him about it and he said it was all
settled and that Colonel Baler agreed with him and I was going to have
to find another job.

	We rode back to Anniston in Stan's car and I got home at two
A.M. I called Colonel Baler at two o'clock in the morning and when he
answered the phone I told him, "Bob, you are a no good son of a bitch."
He asked who I was and then asked if I was drunk. I told him what Major
Elvington had told me and he said he had not even talked to Elvington
and if the depot was not big enough for me and Elvington both then Major
Elvington would have to get the hell out. He told me to go to sleep and
come see him Monday.

	I went to Elvington's office Monday morning after I talked to
Bob Baler and Elvington called the personnel officer in and they started
talking about jobs I could transfer to. I told them I was staying where
I was and they said I could not do that. I told Elvington that if he and
I could not get along then he would have to get his butt off the depot.
I said it loud enough for twenty or thirty people to hear. Elvington
continued to tell me I was through and I told him to go see Bob Baler
immediately or I would have the provost marshal come take him to his
office. Elvington left and was gone an hour. He came back in and did not
bat an eye when he said, "You and I have no problems, Tom. You run Data
Processing and I will run the rest of the directorate." He then asked if
Colonel Baler and I were friends and I told him we were bosom buddies and
he needed to remember it. The Major never gave me another problem while
he was there. I went back to Data Processing and told the departing
division chief to get on the phone and get an immediate transfer worked
out or he would not have a job to transfer from. He left for Huntsville
that week.

	I found that three of the programmers had transferred in from
the Management Division when the Data Processing organization was set up
and none of the three knew how to program and had made no efforts to
learn. I called them in and gave them thirty days to find another job.
Two went to Huntsville and one went to Virginia. I found we were very
low on people who actually knew how to program. I set up a trainee
program with six spaces I had eliminated in the directorate. The six
selected started out as GS-5 trainee programmers, with periodic
promotions guaranteed until they reached the GS-9 level, subject to
satisfactory performance. I selected Doc Brown for one of the jobs and
let the programming supervisors select the remainder. All six turned out
to be excellent programmers.

	When I first went to Data Processing we had an IBM 650
Tape/RAMAC system. The big problem was that the tape portion was called
a tape data selector and was a real monster to work with. The other
depots could handle cards only and we could handle tape. This made us
several times faster than the depots with the IBM 305 Card/RAMAC system.
It also meant we were not compatible with system wide procedures and had
to write all our own procedures and send them to the procedures agency.
This kept at least one analyst occupied almost full time. Another
problem was that we had to do a selling job to get the more advanced
computer and our people had promised to do the total data processing job
with 61 people and we really needed over a hundred.

	Equipment was controlled very closely by utilization and it did
not matter that a system would not run without a particular piece of
equipment, if it was not used a certain number of hours you had to write
a total justification monthly, and fight to keep it. We had an IBM 407
that was used about 24 hours a month and every month we were spending 40
hours in our fight to keep it. I called the Chief of Operations in and
told him to start showing 720 hours per month usage. Headquarters never
did ask the first question after that and the division chief thought I
was some smart cookie to think of that.

	Every month we got call after call from headquarters asking for
our equipment usage breakout in a manner we could not possibly break it
out. I called other depots and found Headquarters needed it for a
Congressional report and no one had it that way but they kept calling
many times every month. I sat down and made a spread sheet and put down
a sheet full of figures that I just made up. The next time Headquarters
called I told them I had them informally and read them off on the phone
and they copied them down. They called back the next day and had three
figures they thought were way out of line. They called off one of them
and I asked if they thought it was too high or too low. They said it was
at least ten times higher than they expected. The figure was something
like 146 and I told them they copied it down wrong as it should be 14.6.
We did the same thing on the other two figures and they were happy. Next
month they called back on just one figure. When I was fairly sure they
were happy with the figures I had a programmer make up a simple program
to just adjust the figures up or down a little each month and print them
out. I started mailing the report then. Headquarters called and said no
other depot was able to provide the information we did so they were just
multiplying our figures by the number of depots and sending it to
Congress and Congress was very happy with it. I was so glad I could help
out. I was telling lies to Congress before Ollie North ever thought
about it.

	The greatest problem was probably that we had a lot of supply
applications on computer and not too much else. There were hundreds of
processes that needed automating but the computer was in the Directorate
for Supply and most of the computer workload was naturally supply.
Another problem was that computer processing was programmed to handle
transactions the way people manually handled transactions. Finally,
there had been a trend toward selecting programmers and analysts from
the people who knew the applications on computer. This created two large
problems. One problem was that the people were not programmers or
computer analysts and never would be. The other problem was that the
people were frequently the most knowledgeable in their area of supply
and there were no qualified people to replace them in the functional
areas. They therefore tended to continue to spend most of their time in
the functional area. I recognized that we had to start selecting on the
basis of programming skills or potential.

	Finally, the equipment took so long to justify that it was
outdated before it was delivered. This is a common problem in
government. The approval cycle needed to be shortened but the people in
headquarters wanted to lengthen it to justify the number of people they
had on board working in the approval cycle.

	A good example of equipment problems is that we had a thirty six
megabyte hard drive that was hydraulically operated and weighed two full
tons. The raised flooring had to be braced again every two weeks to keep
the hard drive from falling through the floor. The movement of the
hydraulic actuators literally pounded the huge piece of equipment into
the floor. Those same vibrations also created disk errors. The equipment
was the finest of its day but its day had passed before we ever got it.

	I suppose it was inevitable with so many depots using similar
computers that a project like SPEED would be born. SPEED was the acronym
for System-wide Project for Electronic Equipment at Depots. Someone
decided the Army would save a fortune if the computers were standardized
and one central agency was designated to write the programs for all
depots. This would supposedly reduce the number of programmers required
overall.

	The Commander, Bob Baler, was notified that a meeting was being
held at Rossford Ordnance Depot in Ohio and he should attend, along with
his Director for Data Systems. I decided to go by train and Bob was
going to fly, so I left a day before Bob. The next day there was a two
inch snow and all flights in the South were cancelled and Bob was unable
to go. I was notified that I represented both the Commander and Director
for Data Systems.

	There were 175 people present and a presentation was made on
SPEED and the charts showed savings in the hundreds of millions of
dollars, mostly in programmer salary savings. The meeting was opened for
questions and I raised my hand and was acknowledged. I merely said that
I knew that figures did not lie but it was obvious that liars had been
figuring. A two star general ran from the back of the room and bounded
onto the stage and  said, "Sir, I will have you know that these are my
figures." I said something to the effect that 'now we know', followed by
a comment that I did not retract my statement. The general commented
that I was calling him a liar in front of 175 people and I said I had
not made a head count. He said he wanted to talk to my commander and I
told him Bob got snowed in. He said he had commanders from areas that
had ten feet of snow and my commander could not make it because of a two
inch snow. I said at least he was honest and I think our conversation
degenerated after that. The Data Processing people from Headquarters
were telling me to shut up and I told them I would speak my piece. There
were more than twenty depots at the time and the average programming
staff was about twenty. We had sixteen at Anniston. There was no central
programming group at the time. The plan was that the central agency
would be staffed with a maximum of sixty people and each depot would
lose sixteen programmers. There would be a savings of about 20 times 16
(320) minus 60 for the central staff, or a net savings of 260 people. It
is getting ahead of myself but eventually, under Project SPEED, the
average depot programming staff increased to thirty five and the central
group had a staff of about 300.

	There was a cocktail party that night and the general came over
to me and said, "You are the one who claims I am lying!" and I said, "I
am the one who knows your figures are a lie." He said he was going to
have Bob Baler straighten me out. All of the other depot data processing
people came over one by one and shook hands and said they agreed with me
but I would not be around very long.

	I left on the train during some of the worst winter weather I
ever saw. We got to Nashville, Tennessee and could go no further. We
stayed in the train overnight and the next morning there was four inches
of ice all over the floor and it was bitterly cold. I had less than a
dollar with me. I went to the dining car and found I had enough money
for one cup of coffee. There was another man at my table and he asked if
I was going to eat breakfast. I told him I had not cashed a check before
I left Ohio and was broke except for enough to get a cup of coffee. He
took out his billfold and handed me a twenty dollar bill and his
business card and told me to mail it to him when I got home. He just
about saved my life.

	I had a message waiting when I got back to work to come see Colonel
Baler. I went to his office and Bob glared at me, bit his cigar real
hard and said, "Tom, why in the hell did I send you to that meeting?" I
said, "Bob, you sent me because of my great tact." Bob mulled that one
over and came back with, "Tom, you got about as much tact as a meat
axe." We then talked the whole thing over. Bob said he had been
requested to fire me on the spot when I got home but he felt I had told
the truth as I saw it and there would never be a day when he fired
someone for telling it like they saw it. He said to get to work and he
would call the general and tell him to go to hell. The old man took a
beating for defending me and I am sure no commander I ever had after
that would have done it.

	Project SPEED involved visiting every computer manufacturer
submitting a bid on the standard equipment. It also involved depot
support. The central group was based on a small organization that had
been writing depot procedures and none of their people knew anything
about computers. Each depot was given a number of programmers they would
have to provide on temporary duty at Rossford Ordnance Depot to help
write programs. Anniston had to furnish eight programmers full time.
This was half of our staff. The depot had to pay the travel and per diem
cost. We asked for volunteers and got four. The other four had to be
selected, and periodically bring one or two home and rotate others to
Rossford. The effort took more than two years. Programming began without
even knowing the final equipment. It turned out that an IBM 1410 and IBM
1401 were selected.

	All programmers had agreed to travel as part of the acceptance
of the position. I had to keep bringing people home and rotating new
ones. We got to the point that all had been except one female programmer
with a four month old baby. I had travel orders prepared for her and she
refused to go. Her husband was six foot six inches tall and weighed more
than two hundred and fifty pounds. He came to my office and threatened
me and I ordered him out. Finally the female programmer went to
Huntsville and found jobs for both herself and her husband. She gave me
two weeks notice. Then a week before time to leave she came in and said
she had thought it over and decided I was right. She wanted to go on
travel and take her turn and then when she had done her job she would go
to Huntsville. That is what happened and I admired her very much for her
decision.

	We built a temporary computer room next to the old computer room
and moved our IBM 650 system into it and remodeled the old room for the
new IBM 1410/1401 systems. When it was our turn to convert, the central
group was prepared to send a large team in to help with the conversion.
I made a decision that we would do our own conversion without help and
we did it despite dire predictions about what would happen to us. We
made the smoothest conversion and in the shortest period they had seen.
We were on Project SPEED and having relatively few problems. We had the
best people in the system.

	It eventually became obvious that the IBM 1410/1401 systems
would not handle the applications the depots wanted to put on the
computer systems and it would be necessary to move to larger and more
modern systems. A new acronym was coined - Project SPEEDEX, which was
nothing more than System-wide Project for Electronic Equipment at Depots
- EXtended. The federal government had an anti-trust suit against IBM
and it was a safe bet we would not  get IBM equipment this time. The
equipment we ended up with was a Control Data Systems 3300, which was
really a communications system with a Business Data Processor front end.
It was obvious to even the most uninitiated that the CDC 3300 was a poor
choice for depot data processing.

	I made an attempt to get SPEEDEX deferred five years. I made a
speech to members of The Armed Force Management Association and my
speech was reproduced in The Defense Manager, a national publication
distributed at all levels of government and to every Congressman and
Senator. Within an hour of distribution of the Defense Manager, John
Gilbert, AMC Director of Management Information Systems (GS-17) was on
the phone with the Anniston Army Depot Commander, demanding I be fired
that day. At the same time, Ralph Tappen, Chief of the Logistics System
Support Center, (GS-15) was on the phone with Charles Heard, Anniston
Army Depot Civilian Executive Assistant, demanding I be fired that day.
Charles Heard had guts and backed me one hundred percent, to his own
personal detriment, and I not only survived but Charles Heard insisted
the Commander approve an Outstanding Performance Award and a Sustained
Superior Performance Award for me.

	Army Materiel Command sponsored a committee consisting of
representatives of AMC, the Logistics System Support Center and the
Director for Management Information Systems of each SPEEDEX depot. The
group elected a Chairman and I was elected Chairman two times. The
directors knew I would fight for what I thought was right. I had their
support and loyalty, and even better, their respect. When my second term
ended I was presented with a poem that the group felt expressed what I
stood for. This poem, "Never Give an Inch", is my proudest possession.

	One very bad problem with the CDC 3300 was the environment it
had to have to even have a chance to run. Temperature control had to be
maintained within plus or minus two degrees and humidity control had to
be maintained within plus or minus five percent. Power had to be almost
perfect. The equipment ran on DC current, requiring DC motor generators.
Steam had to be available to help control humidity. It cost each depot a
fortune to build a new computer room and utility room and it had to be
built by every depot as no depot had a computer room that would support
the Control Data Corporation requirements.

	Construction money was not available and every depot had to
classify large portions of the computer room as equipment. Even walls
had to be classified as equipment and screws were driven in instead of
nails to make it look like panels were demountable. It made liars of
every engineering department in the depot system.

	While all this construction was going on, the depots were once
again sending their programming staffs on temporary duty to do the
programming. Eventually the botched up systems were released to the
depots and the depots began to operate the new Control Data 3300
systems. We learned the difference between IBM and CDC immediately. When
there was a problem with IBM equipment an IBM systems technician used
the appropriate test equipment and within minutes or hours had the
problem identified and the equipment operational. The CDC technicians
were equipped with test gear that was strange to us, consisting
primarily of a comb, a plastic hammer, a hair dryer and a can of freon.
They would run a test program and one technician would sit at the
console while others took turns rubbing a comb across rows of cards to
see if there was a vibration problem. Then they would use the hair dryer
to blow hot air on the cards to see if there was a heat related problem.
If this failed, they would spray freon on cards to see if there was a
cold related problem. If none of this disclosed the problem they would
take plastic hammers and beat on the cards to see if there was a shock
related problem. Frequently, during the hammer phase the console
technician would yell and we would ask if that meant they had found a
bad card. The response was always that they had either found a bad card
or had made a card bad from the beating. If all else failed they would
do what they called 'turn up the margins.' This meant they would
increase voltage a certain percent to see if they could burn something
out completely, making it easy to spot by the smoke.