Forty years later, ADPAC lives

“COBOL, RPG Bested By New Language?” That was the headline on page 1 of Computerworld’s inaugural edition on June 21, 1967.

The language in question, called ADPAC, was developed at Applied Data Systems Inc., a San Francisco-based company that this month marked its 40th anniversary. Peter Harris, the creator of ADPAC and founder of the company that subsequently changed its name to ADPAC Corp., calls himself “America’s senior programmer.” Harris, now 74 and the company’s chief technology officer, recently spoke with Computerworld U.S. about what may well be the oldest software company in the world still operating under its founding leadership.

CW: What ever happened to the ADPAC language?

It’s still widely used. We sold it to hundreds of companies, and we still have a number of them who have some major applications written in ADPAC.

CW: “A number” being how many?

I would say a hundred. Travelers, Prudential, Citibank Corp. — they all had it. I don’t know (offhand) which companies are still using it, but I know there are up to a hundred companies paying us annual renewals. We get close to US$1 million a year in annual renewals.

C: You would contend that ADPAC is superior to Cobol, no doubt. Everybody knows that. Technically, there’s no question about it. It has more features, functions; it’s easier to program in, easier to learn. Cobol was not very popular in the early days; it started to gain strength in the ’70s. To what do you attribute the fact that ADPAC never managed to unseat Cobol?

The government. There’s no question about it. The federal government said, “All programming for the government will be done in Cobol.”

CW: Your company did a lot of Y2k conversion work. At what point did you, as one of the early programming pioneers, realize that there was going to be a problem on Jan. 1, 2000?

I wrote letters (to clients) in 1968 and 1978 and 1988. All that time, I wrote letters, and I got absolutely no interest whatever. All of our (client) companies had ignored the Y2k issue. They were all acting like there was no change of century. I started to call them; I got absolutely no response — no reaction. I couldn’t convince anybody to do any work on it. I’m talking about Aetna, Travelers, Prudential, you name it — big corporations. Nobody did anything about it until 1998.

CW: So you foresaw this problem as early as 1968?

Well, you could see it, yeah. Here I am, thinking I’m in the forefront of software and data processing, and one by one, we’re picking up large national accounts, and I’d tell them about this. Zero interest. Even as early as the ’60s.

CW: As it turned out, Y2k had a very positive impact on your company as a revenue generator, didn’t it?

Oh yeah. We went from $1 million or $2 million to $10 million in 1999. Interestingly enough about Y2k, very little of it has actually been converted in databases. The majority of fixes to Y2k was done with what’s called windowing. What (technicians) would do is, where they needed a date, they would call in a small routine and modify the date right there and print it in the report. But they never changed the database. And that’s still true today.

A (similar) problem coming up is the UPC (Universal Product Code used for bar codes). It’s required to be changed by the first day of next year. And almost no company is moving on it. And I’m talking about large corporations like Target and people in the grocery business that depend upon that bar code. All of their applications are sitting there with eight- or 10-digit codes, and they’re going to have to increase it to 14 characters by the first of the year.

CW: How has programming changed as a career over the years?

At various times in the past 30 years, Cobol was supposed to die and people lost interest in it. Three or four years ago, it was all going to be Java and big companies were going to change all of their code over to Java. I kept telling everybody, “There ain’t nobody going to change over to Java.” It’s a crazy language. It’s more like C++. It’s difficult to program, difficult to maintain; who’s going to do it? Now, programmers, when they go to school, what do they learn? They learn C++. Where are they going to get a job? They can’t get a job until they can boast some Cobol experience.

The software employment situation is going to come back. The first people who will be hired will be people who know Cobol — not Java and not C++. The big mainframes are virtually 100 percent in Cobol. Some of them are in PL/1, and we have a translator from PL/1 to Cobol. We can’t sell it because PL/1 shops are fantastically in love with PL/1. It’s a personality issue. When you learn PL/1, you don’t want to learn or think about anything else.

CW: What is your view on the offshore outsourcing of programming work, and what does it mean for the future of programming as a career in the U.S.?

I’m not sure that big corporations want people trained in beautiful downtown India dialing into their mainframe in New York City. So I don’t think there’s going to be as much offshoring as a lot of people think. Big companies regard their software as very proprietary — very personal and very home-owned. Moving it and training people in India and that area I don’t think is their cup of tea.

CW: Does ADPAC send any programming work offshore?

No. We have very specialized people. Remember, practically all the services and programming we provide are done with automated tools. We have so many automated tools for processing mainframe Cobol. Here’s a crazy example: We just submitted a bid to the state of California. They will not accept fixed-price bids. All of our bids are fixed-price. Why? Because we use our software. We do in a day or a few hours what another company would take a month to do. We can’t bid that on an hourly basis, because we’d finish in three days what represents a month or two of work. So when we bid competitively, we have to bid the month or two of labor. We pick what we think is a winning price, bid it and complete the work in three days.

CW: How do you account for the fact that your company was never acquired?

We eventually will be acquired. We’re trying to position ourselves for that, now that we’re just starting to get profitable again (following a post-Y2k downturn). Before this, owning a language and owning a compiler and trying to sell it was weird to them. Everybody said, “It’s a Cobol world.” What we’re doing now in preparation for the sale of the company is writing a special translator that translates ADPAC into Cobol. We have a million lines of ADPAC code that maintains all of our systems — our Y2k system, our UPC system is all written in ADPAC. We have many special processes that we sell, but they’re all written in ADPAC. So if we want to sell the company, they’re not going to buy it if you need to have an ADPAC programmer. What they want is ordinary Cobol.

CW: When do you expect the company to be acquired?

I would say within the next two years. The thing of it is now we’re starting to be profitable again, and we like it. When we get the state of California contract, we’re going to be very profitable; one more contract, and we’re going to be filthy rich.

CW: Why did you never take the company public? Don’t you think you could have made a bundle when the dot-com bubble was intact?

Our board of directors, which are very senior people, were not dot-commers. They didn’t like the idea, and you had to prove to them that if you’re going to go public, you had to be a multimillion-dollar company. Our board never saw us that way this soon.

CW: What would you say is the single smartest thing you did during your career?

I’m successful because I’m patient. Patience is the key to success.

CW: What’s the one thing you’d be most inclined to change if you had it to do over?

We opened 12 branch offices 20 years ago to support services throughout America. I had to raise a million bucks to do it, and I had to bring in private capital to do it. Three years ago, I bought them all out. And I paid 2 million bucks to do it. I’m not so sure buying out the old stockholders was the right move, because I’d rather have the 2 million bucks.

I don’t know that there’s anything (else) I’d change. Suppose three years ago I sold the company for $6 million. What would I have done with it? I would have put it in the stock market. I could have lost my shirt. Because many of my friends did. People I paid that $2 million to took their money and bought stock. I have one guy on our board who absolutely went broke.

CW: What’s the biggest challenge you face now?

Having 40 years of these ups and downs, I am now facing a tremendous up. Usually, a company sells for four to six times its revenue. When we get to be a $5 million business, which we’ll start to approach later this year, that means we could sell the company for $20 million. What would I do, at my age, with all that money? I program 10 hours a day, six or seven days a week. I still work hard at it. So the challenge is, I’m going to face that opportunity. Am I smart enough to deal with it? I don’t know.

Related Download
3 reasons why Hyperconverged is the cost-efficient, simplified infrastructure for the modern data center Sponsor: Lenovo
3 reasons why Hyperconverged is the cost-efficient, simplified infrastructure for the modern data center
Find out how Hyperconverged systems can help you meet the challenges of the modern IT department. Click here to find out more.
Register Now