This week marked the onset of cheaper air fares. August 23 was the official day for lower fares to be posted by airlines. A fare between Austin and Bejing -- the country where a dozen HP 3000s are driving a manufacturing corporation's operations — is down to $863.
So the impact of airline companies' failed disaster recoveries will recede. There are fewer people to strand when a system goes dark like the one did at Delta Airlines. The disaster was centered around a single building in the Atlanta power grid. It led to people tweeting “I’ll never fly your airline again.” This is just life in 2016. Get used to it: instant reviews, dashed off in the heat of anger and dismay. Tweets motivate spending millions to do good DR.
But the assumptions are that legacy systems are to blame. "Legacy systems stay on way too long," said one blogger who's had some software experience as well as work at Boeing. "Vendor agreements, support and maintenance — and the pain of switching and upgrading a system that’s by and large pretty reliable and so deeply integrated—are things few CTOs want to touch."
Southwest Airlines had no legacy systems at work during its high-season meltdown. The 3000s had been turned off. The plan was to save money by getting more modern. The disaster recovery was not high on the budget list. Customers don't care about IT budgets. They expect the best, even during a disaster. Plenty of 3000s have come through hell and high water.
That reliability doesn’t come out of thin air. The track record the server built during the advent of ticketless flight operations is one reason it still drives manufacturing in places like China and airports serving See's Candies. Celebrating the days of MPE glory won't return it to those places where DR has failed, though. Turning back only happens when a system fails upon installation. Once you're in, it's hard to turn something away at the gate.
Tim O'Neill — a 3000 manager who qualifies for Most Devoted IT Pro to MPE — was moved by our report of a 3000 doing more magic than SAP could at one company.
Yesterday a new user began using our one remaining MPE application. I could say that utilization is growing. If I were a little smarter, I would be able to rewrite the application in C (or some other new programming language on MPE) give the application a Windows look and feel, and then customers who are currently asking to manage their data in the way that only this application does might say, "I want a system like that." And Stromasys would sell them Charon systems.
Charon is ready. The new programming language, not so much; Java was a candidate, but not for long. In that magic alternative world, the thrill would be new MPE customers buying hardware. "I would think I were dreaming," O'Neill said. We can settle upon the dream of keeping MPE alive where it's working. There's still business out there.
One reason business is still in play is that people expect the best. Any company whose product could get you killed, or injured, or create a sleep-in-the-terminal kind of event, should be expected to put in enough DR to avoid a disaster that’s not an act of God. Nobody expects to read in the ticketing side of the Delta website that “You should know we don’t have millions to improve our disaster recovery. Have a good flight, and may the odds ever be in your favor.”