Commanding HP 3000 Conservation

Paul Edwards is committed to conserving what the e3000 community has created. The owner of a Dallas-based consulting firm since 1980, Edwards has been a familiar face in the community for more than 25 years. His experience with system management and application development for HP 3000s includes posts with HP 3000 vendors and HP itself, a position on the Interex Board of Directors, and then on into training classrooms around the US teaching 3000 fundamentals and new system management skills for HP and third-party suppliers. As it turns out, Edwards will lead one of the first classes on the newest version of MPE/iX 7.0, teaching an Express 1 course at Tech Group University early next month.

But it’s not the newest developments of the HP 3000 which drew our attention to Edwards, or even the years of experience dating back to punched card days of the platform when he was a system engineer and district manager for HP. A few months ago the retired US Navy Commander began to talk up a concept for using the existing applications written for MPE as a springboard for reviving the system’s applications business. HP has long made reference to the thousands of programs once sold and supported for MPE, and pointed to their demise as a fact of life. Edwards refuses to accept the situation as an unchangeable fact, and wants the vendor and its business partners to see the value in those applications. As a long-time partner of HP and the community, Edwards was listening while HP was promoting its e-services message for the system. He believes providing applications on tap requires applications without risk: those proven in the field and with little to lose in trying the ASP model.

His message about conserving the system’s accomplishments seemed especially poignant this month, as the HP e3000 is turning 29 years old. We asked Edwards to outline why his conservation program might help the 3000 community, as well as his views on other matters of preservation for the platform.

You’ve mentioned an idea about using existing, under-developed HP 3000 applications as a means to help revive the community. What kinds of software seem to be the best prospects for the plan, and what sorts of resources would be needed?
It began with a friend of mine who used a medical collections package, and talked to me about setting up a separate company to do medical collections. He wondered if I might be able to take over the software package. Then there was a mini-warehouse application written by Chris Schaefer, also being discontinued.

It occurred to me there was a lot of software out there in the 3000 world where companies were leaving the platform to develop on Linux or whatever the OS of the day was. They were either shelving or trashing the software. I got to thinking about the software application development cycle. If you want to develop a package for one of these industries, you’re probably looking at two to four years of development. It’s hard to get venture capital for that kind of environment, because you’re betting on the come for a considerable period of time.
You could take these packages, and you can liberate them by paying a license fee or just rescuing it from the dumpster. Most of it is written in COBOL, and the report part of it has been tested for years. You could take that software and put a pretty GUI front end on it, and run it over the Internet or over a network.

Do you expect the venture would sell new HP 3000s, since HP says applications sell the platform?

The customer doesn’t have to buy a 3000 unless he’s a good enough size to own his own system. You could get into an ASP model with it and go after some fairly small companies as customers. Medical collection is one area, and that product could be modified for other kinds of businesses. You could expand the scope of the vertical package. The mini-warehouse package could be expanded to RV parks and marinas, where people are renting space, garages where people are renting space for cars.
There’s several products around like that, and if people look for them, we could pick them up at a pretty good price. HP must have some old application catalogs with these packages.

There are, of course, hurdles to be overcome in organizing such a business, right?
Someone could call these companies, see if they’re in business and go pick them off, and dust them off, and go run them. You could run this from an ASP model.

I’ve talked with the folks at the CSY [e3000] division about this, and folks at Client Systems also. Nobody seemed to have a lot of interest in it at this point. As I looked at it, the problem I saw in going out and doing it myself is the technical part is the easy part of it. The hard part is to go get somebody to run the finances, do the billing, the HR stuff, the overhead stuff. You also need salespeople to go to the trade shows for the vertical industries, to find customers to show them the product. I suggested that Client Systems could do some of that overhead, we could get technical people to do the overhaul, and the applications could be hosted on 3000s at Client Systems.

The idea would be to go out to the customers and sell them on the concept that all you need is a PC and a connection to the Internet, and we’d charge them by the transaction. Much like the Open Skies model. You’d need to figure out what that per-transaction cost is so you could make a profit.

What are the positives of the plan?
The hardware is available. The technical talent is available, and that talent can telecommute, so that’s not a problem. It seems like a good idea since you would take applications that had been tested, and have them out and ready for market in 60-90 days, instead of several years. With all the dot-coms going away, it should be something you could sell fairly easily if you had a good business plan for venture capital. A lot of those VC guys don’t have any place to put their venture capital these days. You have to get back to the basics.

You continue to work with the 9x7 platform as part of your consulting business. Can you make a business case for why HP should have continued to support 7.0 on those systems?
In 1984 I leased a Cadillac El Dorado, part of my philosophy that if you look successful, you’ll be successful. When I got through with the lease I’d had virtually no problems with the car, much like the HP 3000. I decided I’d buy it out of the lease. I didn’t want to go lease another one, because it was during one of the valleys of my consulting time. I kept it awhile longer, and replaced things like alternators, air conditioning compressors, brakes and tires.

My wife asked me why I didn’t go buy a new one. I said the car was in great shape, I knew its maintenance history and it was reliable. I got good support from the dealer, and owning it cost a little less than some cars. See, there’s all these parallels with HP 3000s. I figured if I spent less than the $5,000 a year lease fees on mechanical repairs, I was ahead. I continued to invest in it, and it’s immaculate today.

Applying this to 9x7 HP 3000s, first off, they’ll run for probably another 20 to 30 years without a problem. The only thing that will wear out on them would be disk or tape drives. My Cadillac salesman would see me for years and ask, “Aren’t you ready to trade that in?” I said until the car quits, or I can’t get parts for it anymore, why trade it in? There’s value in it for the dealer. He said the value is that I am bringing it back to them for service. Support money is coming in.

That’s a problem with HP. All of those support dollars around the 3000 don’t go to CSY. They’ve got a bad internal money model in HP for that. If it was the way it should be, where CSY got the support money, I think you’d see a lot of difference. Then there would not be so much of a push to get everybody onto the new boxes. You sell the product once, but you’ve got support that’s continuous. It’s easier to sell to existing customers than find new ones.

For the 927s, they’re giving them away. But the value is still there, and these things don’t break. Out of the five HP 3000s I’ve owned, none of them have caused me problems. You could support them very easily, but HP’s [9x7] choice will have them losing support dollars. The 9x7s are the majority of the installed base. I think a third of them are going to upgrade, a third will go to something else, and a third will keep them awhile and then do something else. HP’s only making money on a third of these customers. I can see why HP’s doing it, but it’s a problem because they’re cutting off a loyal customer base. The new HP is too commodity minded, in my opinion.

You’ve taught MPE skills for a long time in the community, both for HP and on your own. Why is enrollment declining for HP’s MPE courses?
It’s declining because HP didn’t teach classes for five years, and there was a pent-up demand for training when they started again. I was one of the people who re-did classes for HP, the system manager class for 6.0. That pent-up demand was met, and the customer base started shrinking when the pent-up demand was met. There’s a sustainable amount of training that needs to go on, but it’s smaller.

Second, most of the application providers don’t encourage their customers to go to HP’s training, because the providers offer their own training on the 3000. The training on their products is good, but the training on MPE is not, according to their customers, because they don’t have long-term MPE people teaching.

Another reason is HP’s marvelous marketing, or lack thereof. HP says the classes are on their Web site and listed in their catalog. They’re not going out to the customer base and letting them know, doing mailings.

What’s the advantage of taking training from HP versus a third-party outlet?
HP has good training centers and a set class that they give, so everybody teaches from the same material. It’s standardized. People who teach from the outside can sometimes not teach at the level of HP’s instructors. HP does quality assurance on their courses.

On the other hand, outside instructors have more real-world experience than the instructors from HP, because they’re never outside HP, consulting at customer sites.

The problem is that CSY doesn’t own the training business at HP. The graphics and material on most of the slides are not up to date. A proposal was made to CSY that would allow Client Systems to outsource the classes, which got lots of positive response from CSY, but negative response from the HP Education Group. They wanted to continue to teach in-house. But, they have had a major cut back in the number of classes offered and the number of teaching sites recently. They have also cancelled some classes due to low enrollment. This proposal could have done a lot to improve the quality of the classes and made them more available to the HP e3000 customers and resellers’ technical staff.

What’s the continuing value of HP’s supporting DTCs on the 3000?
DTCs are real important for printers and terminals, and there are still a lot of customers in the manufacturing community that use terminals. They’re also important for modems to dial into systems, but because of the Internet, most of those requirements are going away.

Do the current level of services offered to 3000 developers through the CSPP and its progeny meet the market’s needs?
When you look at the current program, it’s very Unix-, Linux- and Web-oriented on the Web site. When you look at it from the 3000, it’s a stepchild, like everything else. The development software bundle is very valuable, and if we could figure out how to put it together with the applications we talked about, you’d find more 3000 developers using it.

The MPE release schedule is backward. Customers receive the latest releases way before the developers do. This causes problems for the developers trying to troubleshoot problems in the latest releases. Beta releases are sent out to prevent problems in the releases for the customers, so the developers should have that opportunity, too. At the first of November, I still don’t have the 7.0 Express 1 release that was distributed to customers the first of September.

What’s the role of a vendor-specific user group in a heterogeneous computing world?
You can’t have a platform-specific user group anymore, because of the shrinking installed base of MPE and the multiple operating systems in use at most customer sites today. Rather than the control DEC and IBM have exerted over their user groups, I think it’s great that Interex has independence from HP to represent the customer base properly and provide advocacy for all issues.

Does Open Source development have a significant role in preserving the future of MPE?
At some point if HP decides to dump the 3000 and MPE, then it would be valuable. I’m not sure that based on the internals of the system — internal, HP-only compilers used on the operating system — it would be otherwise. I don’t think the Shared Source project that Interex and HP have done works at all. I don’t think the merger’s going to cause MPE to go away, with the large 3000 customers in the installed base. I really applaud CSY for setting up the Invent3k development machine for developers that don’t have their own systems.

You’re one of many HP shareholders who watched your stock value drop in the Carly Fiorina era. Does the proposed merger make sense to you?
I don’t like the merger, and if I get a shareholder proxy on it, I will vote against it. HP is trying to gain market share without respect to employees and traditions. They’re becoming too commodity-driven. All of HP other than CSY is commodity-driven. I don’t think Carly has done the things she needs to do to unify the company. I had a real problem with her not being at the conference of all the HP users the last two years. Maybe it’s time to think about another Wall Street Journal ad campaign.

Should HP be in the business of preserving its technology investments, or leading its customers using the 3000 to more modern technology?
Unix has lost its luster, and it’s not really a scalable transaction processing system. I think that MPE has a capability to run businesses very efficiently. Look at the AS/400; it’s more successful than MPE, in large part because of its applications, which we talked about earlier. CSY has done the right thing in putting architecture standards on the 3000. It’s the best fully-featured open-system operating system on the market. MPE and the 3000 have a niche that could be widened. I think HP should be looking at that option. HP could talk about how open it is, and how it’s the best operating system for business transactions. That’s why they should be spending some of their money on advertising.