Tactical planning for the HP 3000's future is a current practice at shops like MagicAire. The company that manufactures mobile cooling units has a Series 939 that continues to run MANMAN and carefully-crafted applications. Ed Stein there has a need to think about something more pressing than getting his apps and utilities licensed for emulator use. He's thinking strategic.
Stein chooses to think about the end of the 3000's calendar days. He's interested in getting someone to fix the date issue that will arise at midnight on Dec. 31, 2027. The foresight is the first customer readiness we've seen that examines what can be done before that day arrives.
Developers and vendors have been talking about 2028, but not yet in explicit design language. Stein is the first customer who's doing the talking.
I am more concerned right now with the Year 2027 MPE issue. Not that we plan to be on MPE in that year—but if a fix is to be had, that fix needs to be done sooner than later, given the age and availability of the required expertise to develop a fix. There may be no one around in 2026 who knows how to fix it, in the event that in the worst case we are still on an HP 3000.
My company would look at paying for a fix now as insurance.
It's 10 years and five months away, but the end of 2027 is the deadline for regular date handing to stop working. It makes the challenge a Year 2027 issue if you consider Y2K to have been a Year 1999 issue. The most intense work always happens ahead of a deadline. If you're savvy, it's many years before a deadline.
There are the customers who will rely on the work, too. Now there's at least one who's putting near-term licensing in its appropriate rank: secondary to making sure there's a platform that can carry on. Keeping the dates working is like keeping the GPS satellites in orbit. We'd say keeping the street signs on the corners, but that's not the way we'll find our way in 2028. In lots of places, we won't need those signs in 2018.