Bob Green has punched his ticket for a fresh HP 3000 adventure. The 53-year-old founder of Robelle Solutions Technology — his first name makes up the first three letters of the company’s name — is back at the helm of his 24-year-old firm, one of the oldest in the HP 3000 market. He was ensconced on a beach in Anguilla, an island in the British West Indies, when he and Robelle business partner David Greer parted ways late last year. Green had left Greer to the daily management of a company which has 14,000 products installed, sticking to product development on an island so remote it’s more than one travel day away from Robelle’s British Columbia headquarters. Then at the crossroads of becoming a datacenter services and solutions provider, Robelle pulled back to its original course. Greer left to pursue his technology solutions interests and sold his shares to Green. The founder’s second adventure began, as president and owner of a company once again firmly devoted to creating HP 3000 software.

At about the same time Green took over the management of the Robelle Web site, transforming it into a place where new information related to the HP 3000 has been posted every business day since December. The tradition of publication is strong at Robelle, a company whose engineers have delivered countless papers to the 3000 community over the last three decades. Green was named to the Interex Hall of Fame in 1987, an award which was given until 1992 to one Interex member a year who provided outstanding service to the HP user group community. He’s one of four authors whose name is on The IMAGE Handbook, along with Dennis Heidner, Alfredo Rego and Fred White. Robelle published the book through Wordware, and it remains a teaching text about the database which made the 3000 a success.

The products Green invented, Suprtool and Qedit, remain at the head of Robelle’s lineup after more than 20 years in the field. The company counts more than 6,000 HP servers around the world running its products. Green came to HP in 1969 with a Stanford philosophy degree and began documenting what the engineers were doing in creating the HP 3000. In a sense he is still doing that with the company’s Web site, tracking what’s developing for the server. We asked Green why he’d want to return to close contact with a software company that serves a mature market after so much accomplishment and achieving an engineer’s dream: programming on the beach.

Your company has made a major change in management. What made you want to step up to the daily management and increased travel now?

I realized that I could visualize going forward with Robelle for years to come. I invented the company’s products over 20 years ago and still get a thrill when someone tells me how great they are. That is what I work for.

Robelle’s got a lot of technical expertise that could be attractive in a services offering. But now the company is going to be sticking with HP 3000 software as its main offering. Why is services less appealing to your business model than product?

Although we decided against the push into datacenter management and refocused on software, in many ways we are a services business and have been for the last 10 years. If you look at the lifetime revenue from a typical Robelle customer, only a small fraction of it comes from the software license fees. Most of it comes from maintenance fees. We plan to focus on retaining that service income, and supplement it with consulting services related to our core competencies. For example, when you buy Suprtool, your productivity would improve much more quickly if you bought some consulting time from us to convert your most important batch processing tasks to take advantage of Suprtool.

My attitude was that in the past, when we’ve gone away from our core business, we haven’t done as well as when we’ve stuck with our core business. I didn’t think it was a good investment for my money.

We like to please our customers. Ten years ago they came to us and asked for Qedit and Suprtool for HP-UX. At the time it didn’t seem like that big an investment, but we had to buy a bunch of new machines, learn a lot of new skills, expand the staff to be twice as big as we were. What we found was the only people we could sell to were the people who knew our product and were moving to HP-UX. A small group, really. What I learned was the success of our marketing in our core business was being in the right place at the right time. We just happened to be interested in the things that people were buying. When we tried to follow the marketplace into something we weren’t personally interested in, we never made any money on it. All we did was make the company more complicated, and spread our resources thinner.

When I started Robelle the goal was not to worry about sales volume, only profitability. In the early days we would turn business away, when someone would send us a 40-page purchase order. We would send it back. We were going to focus on the most profitable customers. If you required a 40-page PO to buy Qedit, gosh, you were going to be a pain to deal with.

Is there then a move away from Unix at Robelle?

We’re not writing off our Unix customers, but we’re not as important to them as we are to our 3000 customers.

The elusive goal of being a star in the Unix world — we’re not ever going to attain that. We know enough about Unix that it’s not that much extra work now to support Unix for our products. What it takes to be a star in the Unix market is different, and we don’t have it. That’s what I thought about datacenter management.

There’s a pretty prominent Web presence for the 3000 out there in 3kworld. Why start posting daily at the Robelle site when there’s a clearinghouse available?

I am a supporter of 3kworld and have donated my own time to help them make it better. My biggest suggestion was that someone who knows the 3000 has to edit the home page and decide what to highlight on it. 3kworld listened and now has Chris Gauthier sifting through the postings and highlighting the most interesting. And we repost our material on the 3kworld clearinghouse after we publish it on our own Web site.

The purpose of our Robelle web site is to do the best possible job for users of Qedit and Suprtool by bringing to their attention information that will help and entertain them, regardless of whether we wrote the material or not. Daily updates are something that I tried in the Caribbean with my Anguilla news and I found that people loved them. If someone sends you a tip on using Suprtool and you post it on your home page the same day, they are amazed and very pleased. If I post their tip on 3kworld, buried three clicks down from the home page, I am not having the same impact on my customer.

What we are doing is building a community of Robelle users that help each other with their problems. I see this as a subset of the larger 3kworld universe, which covers people who have never heard of Qedit and Suprtool.

Your own personal Web pages show the potential for the Internet in its purest sense: Home page really means knowing something about your home. Did retreating to the Caribbean help you recharge, and would you recommend such a retreat for others who’ve been in this business as long as you?

First, I didn’t see it as a retreat, I saw it as an adventure. Could I function as a programmer and remain part of the Robelle/3000 community even though I now live on a tropical island? On the one hand, the answer is no, I couldn’t. I didn’t have anyone in Anguilla to help me solve the problems of getting set up. I couldn’t even boot up my HP-UX and MPE server by myself because I had delegated those jobs years earlier. It was very humbling.

On the other hand, the answer is yes. As soon as I admitted I needed help and built a local support system around myself, I could be almost as productive as I was in Canada, using the power of the Internet to keep in touch.

I spent two of the last six years in Anguilla working half for Robelle and half on local projects: a computer club, Web pages, Internet news, an encrypted money project, construction, lots more. When I left Canada I was emotionally involved with all the staff. Sometimes I was too wrapped up in it. Having been away for six years I could have a more grown-up attitude. They’re still my friends and I like them, but I made new friends, too. Maybe I was a little burned out when I moved to Anguilla, but I had a lot of hours off. Plenty of time to get recharged. When I looked at the possibility of five years from now, still being the guy maintaining Qedit and answering questions on Suprtool, it actually felt real good. This is something I could do right into retirement.

When the time came to decide whether I wanted to sell my share in Robelle and relax, or buy my partner’s share and put all my energy into revitalizing it, I found I came back to the 3000 world and Robelle with a new perspective. I could appreciate the uniqueness and value of it in a way that I had lost after being immersed in it for 20 years.

I have now visited other computer communities like the Linux world and the encryption world and the MS Windows world. I can see that there is something wonderful about a computer community where we take reliability for granted — as in the encryption world; where we share information freely — as in the Linux world; and where we are doing productive work that pays well — as in the MS Windows world.

You’ve got one of the longest histories connected to the HP 3000, a history pre-dating the system’s first release. What kind of advantages and unreasonable expectations do you encounter in your relationships with 3000 customers new and old?

If you work on the same topic long enough and do a few good things year after year on a consistent basis, you will become well-known. When I started working at Hewlett-Packard I was only 19 years old. My entire working career has been spent with HP, with HP users, and as an HP vendor. Mostly this is an advantage. The only downside is that old time compatriots sometimes look to me for solutions to impossible problems (how do I stop my boss from converting to Windows NT?). I don’t have all the answers — I am just a person who documents and communicates the few answers that I do learn, mostly learned from other people, by the way.

One of the things I enjoy about your background is your ancient Greek philosophy degree. How did that schooling help you in the world of technical accomplishments?

When I went to work for HP the computer division was in Palo Alto, and they were working on the 4Kb Fortran compiler for the 2116. They had about seven programmers, and I was just the go-fer for them. Then they gave me a project to covert an Algol program from the Burroughs 5000. HP offered me a full time job when I graduated with my philosophy degree, so they made me into a technical writer. That’s when I learned that I didn’t know how to write. It was the worst manual ever written — and we have testimonials to that.

Succeeding at school builds skills that you can use at succeeding at other tasks. I don’t, however, feel that school is the only way you can achieve those skills. Any structured environment where you are moved ahead in an organized way to tackle ever more complex topics will do it. The advantage of university is that it prepares you to tackle intellectual challenges, even if they have nothing to do with your degree topic. However, university does not necessarily prepare you for any practical skills (see my comments on writing skills), so it is not a cure-all.

Your work on the 3000 began at HP before the system was even released. Was there any expectation the product would last 30 years, or was the goal just to build something that worked reliably?

They started designing the new machines in 1971. As the software went on, they needed people to work on the documentation, so I got drafted. I was like a documentation coordinator, but I did get to see all the different parts of the system that way. The only other person who got to see all the parts was the product manager.

They had brought in a really strong team [of designers] and compiler writers, and I wasn’t in their category at all. I was just a guy one year out of college. But luckily, none of them could write. Isn’t that always the case? So there was plenty of opportunity.

By the time we did this the computer division was still a very small part of HP. I don’t think anyone at HP in the 1970s ever considered how long the product would last. They were just building another voltmeter. It should be better than the last one and should come in under budget. Remember, the computer industry was new. The creators of the 3000 went into it with a spirit of adventure: Can we put the best ideas of mainframes and big multiprogramming systems into a little minicomputer and blow people’s minds? Not, do we have a business plan for 30 years of computing reliability?

And speaking of reliability, that was not the number one priority in the beginning. In fact the original 3000s were extremely unreliable, because the MPE software ran on underpowered hardware. I think the trauma of the original crashes and customer rejections was what created the drive for reliability that was attained a couple of years later with the Series II.

What is ironic is that HP had to go through the same lesson again ten years later with the release of the RISC versions of the 3000.

One of your keenest interests seems to be in writing, from the SMUG books and IMAGE Handbook to the Web pages you’re maintaining daily now for Robelle. How has writing about the 3000 changed for you over the years?

I always wanted to be a writer, but believe it or not, I was originally a very bad writer. I learned to write at university, where everything had to be convoluted and in the passive voice: “The new program was written, on the main, with an eye toward the possible reduction of unacceptable inventory levels” instead of “The new program reduces inventory levels.” It was while I was a technical writer at HP that a professional technical editor whipped me into shape and taught me to write clean, concise prose.

The main difference now is the Internet. Writing for the Internet means being even more concise. You need to make your point in three sentences, then you can link to an expanded description, which is optional from the reader’s point of few. So you need to write small, but with the ability to link; you can expand an article into something much larger and deeper than is possible with print (i.e. lots more examples, links to background resources, etc.)

How do you want to change the way 3000 customers interact with a programming editor? How does such a product find a wider audience than those technically adept enough to cut and maintain code?

One of the reasons we wrote Qedit for Windows was to broaden our customer base. Only a small subset of people in an e3000 installation are interested in learning a large set of commands in order to edit a few files. But most of the people know how to use Windows editors. So what we did was put a standard windows front end on our Qedit server editor. We now have many at user sites editing Suprtool scripts with Qedit for Windows, plus data extract files, configuration files, and more. So Qedit isn’t just for cutting code anymore.

HP believes there’s two audiences in its customer base, those super-informed and cutting-edge, and those still unaware of the 3000’s advances. CSY is searching for a way to connect directly with the second group. What would you recommend?

The only way to reach them is through their primary interests: their business community and their application software. Specific manufacturer platforms are not as important to people as they used to be — it is the software that builds on top to solve the problems of their industry that is important.

Robelle has stretched its abilities in working with platforms other than the 3000. CSY now says that it succeeds even when something else is sold other than a 3000, so long as it’s an HP solution. How does that strategy sit with you?

I agree with CSY. The solution to the customer is the most important value. So our goal is to solve as many problems as possible with the products that we sell, then communicate that information to as many people as possible. If we get our little slice of the giant IT world, I am happy.

What made you realize happiness could lie in returning to something you had done for years?

Looking back on things like the products I wrote that never went anywhere, and the diversification efforts that weren’t profitable, I realize that I was very lucky. Why should I fight it? This is something that I am good at which I now feel I have the energy to do again. There’s nothing else in my life that’s as likely to be as rewarding. It’s kind of surprising.