25 Years: Build an emulator, so they'll stay
MPE/iX networking flaw has workarounds and a fix

25 Years: Java promise yields Go e! app

Let's Go e Enhydra app
It's November of 2000, close to a year past the harrowing Y2K milestone. The HP 3000 is now renamed the HPe3000, adding a letter to remind customers and prospects that the 36-year-old server is ready for the Web.

HP Europe is running a "Let's Go e!" conference. The event is so multilingual that a set of translator booths sits at the back of an Amsterdam hotel conference room. The presentations will convince customers from France, Spain, Italy, Germany, and the UK that tge 3000s in their datacenters can connect data with remote customers.

Amsterdam translation booths
I'm in the audience and look back to see a UN translator setup worthy of a scene in Judgment at Nuremburg. In real time, the genuine capabilities of a Java-driven app are being demonstrated. It's a proud moment for people like me who invested in the future of the 3000 world.

In a way, the conference is multi-lingual for technology, too. Java made its debut in commercial markets just a few years earlier. In that room we're being told that MPE/iX can speak Java right alongside Unix and Windows NT. It's an important point, that similarity with an open Unix environment, or the omnipresent Windows. The 3000 deserves a seat at the table, HP believes. It's especially important in Europe, where they've had a tough year selling against Unix. HP-UX and Sun Solaris are well dug-in across the continent.

An IT manager from Dornier, which makes custom looms for the fabrics sector, explains how their Enhydra web app server built upon Java/iX runs as fast anything. An outside team built them the app for Windows NT, then moved it to the 3000. At the time, that would've been a 3000 before the ultimate generation. 

Not especially fast compared to what would be announced four months later: PCI-based 3000s of the A-Class and N-Class. Still, for Dornier's business clients, fast enough.

Java earned a reputation over the next year or so as being significantly slower on MPE/iX than open system implementations. In almost one year's time, HP decided the ecosystem of the 3000 didn't have a strong future. Despite the translation magic in that Amsterdam meeting room, the place the e3000 was going to go was away from HP's futures.

Comments