Where were you when you heard the news, the bolt of HP's notice about exiting the 3000 marketplace? Last week I passed the very spot where I first heard about the "end of life," as HP wanted to call it in November, 2001. On my recent 50th birthday vacation in Paris and Switzerland, I walked past the very phone booth in Lausanne's train station where my distraught partner was telling me from the US, "HP says the 3000 is going away. They're not going to make it anymore. They need to talk to you, before they announce."
I didn't shiver last week when we finally walked past that phone booth together. The weather this May was warm and sunny, even in the shadow of the Alps. But I could call up the chilling thoughts and questions I already had started to dream up, even before that trans-Atlantic call with my wife Abby was finished. On that night I was 45 years old. I'd covered the HP 3000 for 17 years. I knew that HP's five-year time-frame for getting customers off the 3000 was outlandish, knew it even before I hung up the phone.
It's been five and a half years, as of this week, since HP made the announcement to change your careers. I bought a new notebook the next day, when my son and I got back to Paris, and began to write questions for my interview. At the top of the first page I wrote the seminal query, the one that fueled 49 more:
Tell me why it's going away.
Even after all this time, some of those 50 questions I wrote in a fever of inquiry, roaring toward London on the under-the-Channel Eurostar train, remain unanswered. Some that HP's Winston Prather and Christine Martino did answer have fuzzy replies, even after half a decade. Things like open source or sharing of MPE code with third parties, or a delivery channel of HP-like 3000 services beyond 2008 — remain unresolved. When will third parties get HP's direct help for the homesteading customer? Things like that remain mysteries.
To this day, customers still wonder how long HP knew — before it began to tell its top application partners and bigger customers — that it would cancel its 3000 operations and development. "A few months before" was the answer I got in 2001, sitting in a London hotel doing an interview, two train rides away from that Lausanne platform. "Months" is the most real part of that answer.
Among the other questions that still must be nailed down, now by a team taking over for Dave Wilde, Mike Paivinen and Jeff Vance:
Will the customers and development community get access to HP's internal compilers, to make changes to MPE/iX? (HP hasn't gotten that specific about what it will release.)
What are HP's plans for its own 600 internal HP 3000 systems? (Here I was told those plans "were similar to those of major accounts; hundreds have a plan to migrate." But there are still HP 3000 systems running HP company functions this year, well beyond the 2006 deadline)
Are the PA-RISC customers in the HP 9000 customer base being given an obsolescence date as well? (Not at all, I heard. But by now the PA-RISC HP-UX server is being phased out, replaced with Integrity servers using Itanium.)
Three years ago HP committed to IA-64 [Itanium] for the 3000. What has happened in the market since then to change that commitment? (We heard of a decline of "the 3000's ecosystem." But low-growth product lines like the 3000, generating small revenue increases, had to go under CEO Carly Fiorina's plans. If anything changed in the market, it was HP's measure of an acceptable size of a business line.)
Will there be a planned reduction in Response Center staff trained in MPE? (The answer was no, but then HP offered two Enhanced Early Retirement programs, plus moved its MPE staff onto concurrent support duties for other operating environments. Not exactly planned, but a real staff reduction has taken place.)
What are the possibilities of having solution providers take over some parts of MPE source, like the spooler or ODBC? (HP didn't know then, and it doesn't know now how it will hand off MPE, or to who.)
Is there any possibility of reviewing this decision? (I was told no. But then HP extended its 2006 deadline about four years later. No reversal of the decision, but the vendor isn't out of the market yet, is it?)
What migration training do you have planned? (A Webcast series was promised beginning in February 2002. Not enough training emerged from HP until well into 2003. Third parties took up the slack, but the leadership from the vendor would have sparked more serious migration activity.)
Is this decision in the best interest of the 3000 owner, and if so, how? (HP said back then that it was in the customers' best interest, because HP felt it was risky to remain a 3000 customer. Ownership of a 3000 was most influenced by the vendor's leadership and plans, so HP's decision steeped whatever risk was already floating in the waters. For the customers already moving away, they saw it as in their best interest — a reason to invest in migration.)
Sitting in that hotel in Europe I asked, in a question near the end of 90 minutes of grilling, "What blueprint will HP use to close down a mission-critical business line like the 3000?" It was a trick question. I already knew the answer. There was no blueprint on that week in November. That's what you, your third parties, and HP have been developing during the last five-plus years. In 2001 I walked away from that train station phone booth knowing that drafting the blueprint would be the story — for years to come.
In an editorial I wrote in a London Internet cafe about the afterlife of the 3000, I said nobody would know what any afterlife might look like. And here we are, still 18 months away from that life-after-HP's 3000 business. There's plenty of stories still to tell and advice to pass on, I thought, passing that phone booth on our way back to the US — and preparing a May, 2007 issue to put onto the presses.