December 2004

Care for a Market With a Longer Lifespan

Steve Cooper and his business partners see growth at the tops of the computer market’s oldest trees. The president of Allegro Consultants, Cooper steers a company with 3000 roots that reach even deeper than RISC, the 1980s gamble that paid off to extend the life of the 3000. Allegro’s founders consulted for HP when it tested MPE for RISC, then helped to co-author Beyond RISC, the first book to detail the architecture that’s still driving more than 95 percent of 3000s running today.

Allegro has grown beyond its namesake business since those days of the late 1980s, developing products during the 1990s and now taking on first-line support for companies like Robelle and 3000 customers who are leaving HP support behind. Cooper has been at Allegro’s helm for the entire 20 years the company’s been in business, and he counts seven more years with HP 3000 consulting experience before that. This year is the 35th that Cooper has worked in computing, and he’s still making a splash in the market. Cooper was essential in starting the idea for the Resource 3000 consortium, a group now aiming to be a one-stop shop for 3000 services and products since HP’s leaving the field.

But even with four companies joining forces, Resource 3000 is still more the size of Allegro when compared to the size of HP. But Allegro has substituted for HP in its past, from contract development of IMAGE database features to “destructive testing” of MPE before HP released it on RISC hardware. At one point, some in HP were calling Allegro’s lab “Cupertino East,” an extension of 3000 development. When Allegro appeared on the Resource 3000 roster we wanted to check in with Cooper — who’s always been a realist about business prospects — about how much growth remains in the 3000 market. We spoke in late November by phone, just as Resource 3000 was hanging out its consortium shingle.

What makes you think that four HP 3000 companies can cooperate together in a consortium?

I don’t think that’s unusual at all. The 3000 community has had a spirit of cooperation from the beginning. We’ve always understood strength in numbers. In the past it was for advocacy. Now, to some degree, it’s for survival.

What led you to believe a consortium could make a difference to HP 3000 homesteaders?

We started by looking at why people chose the 3000 in the first place. Most of the attributes are still there: It’s reliable, secure, robust, still has rich third party tools and a solid database management system. The only thing that’s really going away is HP support. The consortium can fill the gap that’s left by HP’s exit. If that gap is filled, the 3000 can be a viable platform for years to come.

Beyond that, what the consortium can help do is revive the community spirit that used to exist. It’s hard to find kindred spirits at HP World meetings these days. A lot of people probably think they’re out there on their own; they’re not.

What downsides will you have to overcome at Allegro to participate in Resource 3000?

Allegro is not a marketing company, and we don’t want to learn or acquire that knowledge. But if you look at the other three companies in the consortium, they are our partners already. We developed De-frag/X that Lund markets. Ideal sells hardware and software support, and we’re their software escalation center. ORBiT markets our most successful products, Hourglass and Rosetta Store.

We share common experiences and acquaintances, and we all struggle with redefining ourselves after HP’s 2001 announcement. So far there’s been no issues at all.

So has the 3000 market entered a new era of cooperation to resolve those struggles?

There are a bunch of companies still in the 3000 market. The survival of each one is essential to the survival of all of them. We can’t be viewing each other as competition. The competition is every other platform that’s out there. Working together, convincing the customer that this is still a viable platform, is essential to ensure all of our livelihoods.

Where do you think the customers for your consortium are likely to come from — HP, or other 3000 product and support providers?

My guess is that HP’s always had 90 percent of the market sewn up in areas like support or backup products. Now, for the first time ever, that 90 percent is becoming available. Their provider is abandoning them. Rather than trying to steal each other’s customers, the remaining support providers should be looking to pick up those users left in the lurch.

Is the 3000 customer base ready to believe a small firm can deliver 3000 and MPE support as good as HP’s has been?

If they’re getting support from HP now, they’re already getting support from a small firm: the small HP 3000-aware group inside a large printer company. After all, our group has a much higher percentage of HP 3000-knowledgeable people.

In actuality, we have an impressive client list among the firms in the consortium; they’re already convinced of our ability to deliver “better“ support. I hope others will give us a chance to convince them. I encourage people to take us for a test drive.

How does a 3000 customer go about telling their management that support will come from a smaller firm?

It’s a matter of doing a risk-benefit analysis. If their 3000 applications are meeting their needs, there’s an excellent chance it will do so for years to come. The hardware’s still going to be there, the software’s still going to be running, and we’ll be there to answer the support phone.

Do you believe the 3000 market is in that bottom-right part of a bell curve?

I believe that’s where we are on the curve, but I‘m not sure the shape is so bell-like. We’ve got a few customers that still have Classic 3000s out there. They are incredibly successful applications for those companies. For one reason or another they’ve been unable to migrate to PA-RISC [3000s]. They live on, and I think that’s going to be true for a lot of 3000 sites as well. Hardware is an expense, but software is an investment. You want to be able to protect your investment and have it produce fruit as long as possible.

How much of a commitment in years is Allegro making to the 3000 as of today, and why?

Shortly after the HP announcement in 2001, both Allegro and Ideal came out and said we’re supporting our customers until Dec. 31, 2011. We figured 10 years would be enough time for our customers to take the next step. Believe it or not, a few companies have approached us and asked if that 2011 date is fixed, or if we’d consider supporting beyond that. 2012 is a long way off. I’m not sure I’m willing to commit beyond that, at the moment.

Do you believe that moving away from the 3000 is an eventuality for everybody using the platform?

At some point in the future the last 3000 will be turned off. Everyone will want to replace their systems or power them off. But people ought to be able to do that on their own time schedule, not one dictated by another company.

Does the 3000 market have to begin to grow again in order to survive?

Actually, no. I’ve always been a believer that small is beautiful. Allegro is a good example of that. There’s been plenty of opportunities for us to get much larger. Large companies abandon marketplaces when they’re ‘only’ generating a million dollars a year. That doesn’t have to be the case when your market is small.

It appears to us that not-so-bell-shaped curve is leveling off. The number of companies remaining on the platform will be enough for us to maintain the infrastructure.

What makes you think the curve is leveling off?

Just looking at customer lists and support renewals from the consortium companies. We certainly saw a trend develop which accelerated once HP made its announcement in 2001. Companies were getting off the 3000 because it wasn’t the operating system of the month. We’re not seeing them drop off anymore. To some degree, those who are going to leave have left. Those who are left are either in it for the long run, or for the time being, and are patiently replacing their systems with something new.

Should some team be getting started next year on taking over MPE/iX source, if HP will release it?

We don’t think it’s essential that anybody gets the source. We’re prepared to meet our commitments even without the source code. That said, I would love to see the release, and the sooner the better, to see HP announce some decisions that they’ve had under consideration since the end of 2001.

To a large degree, that doesn’t much matter anymore. HP has become more and more irrelevant to the 3000 community as time goes on. People are adapting to the lack of information from HP.

How are they adapting?

We see a growth business for the first time in many years. Customers are coming to us for support, saying they’re not going to be off the platform by HP’s deadline. They need support at the end of 2006 anyway, so they start developing a rapport with a support organization now, and not wait until then.

How long has it been since you’d seen growth in the 3000 business?

Since Y2K. It was a godsend to the computer industry. If I knew how to switch the millennium again, I’d do it. I have to assume there’s not that many new 3000 customers. They must be coming from somewhere. They must be coming from HP.

Are you getting new support customers who haven’t been HP support customers in a long time, the self-maintainers?

We’re seeing some of everything, but mostly we’re seeing people leaving HP support.

Can the 3000 market survive with a frozen MPE/iX, or will it need to be enhanced in a few years?

It can survive with MPE frozen, if it needs to. Just look at how many sites have not upgraded to MPE/iX 7.0 or 7.5. In many cases people have stable applications. They have no interest in seeing anything change from underneath them right now.

What is the one thing the 3000 market has which alternative markets do not?

The first thing that comes to mind is stability. I started in 1977 with the HP 3000. If you take a program compiled back then, object code from a Series II, and you put it on the latest and greatest HP 3000 N-Class today, that program will run. It’s an incredible story of compatibility. No other platform can make a statement like that. Companies have come to depend on that. If you’re going to spend a lot of money to develop a system, you want to be sure you’ll get your return on investment, and sometimes that’s five or 10 years or longer.

So the five years that HP felt was generous for migrations is short to some companies in this market?

Yes. We get to work in a lot of different industries. In the chip business, products sold today can be totally obsolete nine months from now. That’s true of the motion picture business, a place where we’ve got clients as well. Movies don’t have a long shelf life. Compare that to the forest products industry — where people are planting trees and doing experiments that no one’s going to know the results of for 70 or 80 years. Each company needs to follow its own timeline, not one dictated by HP.