Network Computingwww.networkcomputing.com

Well, being topical with a five-week lead time is a tricky thing. I'd love to give you my knee-jerk reactions to what must now be the well-understood--or at least well-debated--purchase of Digital by Compaq. The questions that remain in my mind could easily be answered by the time you read this. Ah, who cares? I've got one foot in my mouth and there's plenty of room for the other!

Reminiscences of DEC Toward the beginning of the year, I reminisced about t he good ol' days when I studied electrical engineering. DEC, as Digital was then called, was in its heyday and Compaq was nothing more than an interesting anomaly in the new PC market. Compaq made IBM clones, oftentimes better than IBM did--and frequently just as expensive. DEC was an engineer's company. Its operating system and compilers were the choice of engineers writing scientific programs in every important scientific field. DEC's extensions to FORTRAN would be copied in other compilers for computers from the PC to the fastest Crays.

DEC did all right in business applications, too, but it was science and engineering where it made its mark--and it was a market that IBM happily gave DEC. The other business computing folk like Sperry, Data General and Wang also happily let DEC own this market. Only the likes of HP, and eventually Apollo and Sun, would challenge DEC.

DEC's operating system, VMS, and its hardware architecture, the VAX, were very closely tied together. DEC thrived by providing an inte grated system, including hardware, architecture, OS, compilers, peripherals, service and support. You could get it all from DEC. And, just as the elite in the business computing world most often chose IBM, the elite in the scientific world chose DEC.

Like an aircraft carrier being buzzed by PT boats, DEC was out-maneuvered and outwitted by the newcomers to the scientific computing market. Sun and Apollo introduced workstations long before DEC. They embraced RISC architectures long before DEC too, and when DEC began to lose the performance battle by 5-to-1 and 10-to-1 margins, the scientific and engineering communities abandoned DEC like the proverbial rats from a sinking ship. The company shrank mercilessly, and like a crew hopelessly bailing against the sea, the remaining DEC employees attempted to resurrect the company under the slightly modified moniker Digital.

A Company Grows in Houston Meanwhile, Compaq grew. It made good, solid machines that it supported well and it left the software to Microsoft. Marketing, support and good manufacturing: That's what built Compaq.

So what does this imply for the new combined company? Well, if your company relies on DEC Unix, OpenVMS or any other piece of software from Digital, it's probably a good idea to think about migrating to something--anything--else. From the nonstop Tandem machines to the most basic desktop system sold by Compaq, it's likely that the software will come from Redmond.

If you've been frustrated by the lack of worldwide support from Compaq, you'll likely be pleased, eventually, with the new combined company. Digital still has one of the best support organizations in the world, and that's a significant asset.

The questions revolve more around what Compaq will do with the Alpha architecture and other chip development programs. My guess is that Compaq will attempt to sustain Alpha's development. I seriously doubt that Compaq yet understands what that means. Whether a company that's used to operating in a slim margin industry has the staying power to continue development of its own processor is a tough call.

Waiting Impatiently? Try HP. Meantime, while Compaq digests its Digital purchase, look closely at HP. It's the company that looks most like what Compaq is trying to be. And if the Compaq of a couple years from today looks good to you, why not go with HP now and avoid the uncertainty of the wait?

Art Wittmann can be reached at awittmann@nwc.com.