What Changed Over 30 Years: Bespoke
August 19, 2014
I arrived here in the community of my career when gas was $1.15 a gallon in the US, the Dow was at 1,200, a new truck sold for $8,995, the Cold War Olympics featured no Soviet atheletes in LA, and Stevie Wonder had a top hit on the record charts. Because there were still records being sold for pop hits, along with cassettes. Nary a CD could be bought. The Mac was brand new and still didn't sport a hard drive. Those fellows to the right were right in style with warm-up suits that you're likely to see in a senior's happy hour cafeteria line today.
There were thousands of applications in the Hewlett-Packard software catalog of 1984. It wasn't a new idea to collate and curate them, either. MB Foster had one of the first compendiums of HP 3000 software, several years before it occured to HP to offer products the vendor did not make (or buy up, then sell back). But in the month when I entered this market, during that August you were at least as likely to find custom, bespoke software running a corporation as any Commercial Off The Shelf package.
People built what they needed. The bespoken software was often created with the help of fourth generation langauges, so Speedware and Cognos' Powerhouse were big players during 1984. Not the biggest of the 3000 vendors, in terms of customer size. Unless you counted several thousand MANMAN sites, all running the Quiz reporting tools that ASK Computer included with the MRP package. Back in those says, Enterprise Resource Planning hadn't been conceived.
Because so much of the community's software was customized, being well-versed in IMAGE/3000 -- not yet TurboIMAGE, let alone IMAGE/SQL -- was a key skill. Mastery of the database was more attainable if you had a database management utility. Adager was most widely installed, with Bradmark just getting off the ground in 1984. I nearly crashed my reputation with Adager and co-founder Alfredo Rego, less than a month after I began my career in the community.
The problem was a lack of MPE and IMAGE experience. Since I didn't understand the technology first-hand, I felt compelled to contribute to the effort of the HP Chronicle. Not by writing an article, but instead closely red-pen editing the writing of Rego. I didn't know yet that anything he shared with a publication -- his technical treatise was a big win for us at the HP Chronicle -- had already been polished and optimized. A writer well-steeped in mastery of his subject can insist an article be published with no changes. In the publishing business, stet means to ignore a change. I'd have been helped if someone had grabbed my inked-up printout of Rego's paper and marked "stet all changes" on the front. He had a legitimate beef.
Instead, we ran it and then I got to enjoy a rare thrill -- having my corrections corrected by the author, live in front of a local user group audience. Writers forming the troika of big independent vendors -- Bob Green at Robelle, Eugene Volokh at VEsoft, and Rego -- certainly had earned stet-all-changes. Their software became crucial in managing a 3000 that was gasping for new horsepower. Creating and maintaining customized software was a popular way to get the most out of the six-figure HP 3000s, already at the end of the line at the top but still more than two years away from getting a refresh.
But so much of what was running neither HP or third-party. It was custom-crafted. And that year could get a new level of support, via phone in the US out of Santa Clara, Calif. and from Atlanta.
In my offices, the 3000 was limited to an amber terminal emulator screen, representing time on a system down at Futura Press, where the newspaper was printed monthly. We never saw any SEs unless we were at a conference -- where they gave talks. We never installed an HP 3000.
It was an era where PCs were on the rise, but not being much trusted in the Data Processing departments. The financial forces started to carry the day with PCs and MS-DOS, but the established MIS sector analysts figured that PCs would saturate the market quickly enough. One $400,000 study reported "Early PC peak forecasted," where SRI International predicted PC growth tapering off after 1986. "Average annual growth will be only 5.4 percent in the 1986-1990 period."
Customization -- the bespoke nature of database designs -- was supposed to be holding back more PC growth. "Some companies find that the file structures within their corporate databse do not lend themselves to easy access by PCs." Personal computers were supposed to work unconnected to the databases like IMAGE, the experts figured. Then software like Data Express arrived to change all of that connectivity between PC spreadsheets and minicomputer databases. IMAGE could use what Lotus 1-2-3 wrought/
IMAGE adjustments, management and optimization were so popular that we had a pristine copy of the IMAGE/3000 Handbook in our office -- though it was more for my education than any operational use. The book was 330 generous sized pages, plus index, written by Bob Green, David Greer, Alfredo Rego, Fred White, and Dennis and Amy Heidner. "The book sold itself," said Green, "and since the price was $50 each and we paid for the printing, our editor Marguirete Russell had a nice extra income for the next few years."