The looming day of reckoning

A digital Day of Reckoning is approaching, especially in government. Its arrival may be facilitated by a hurricane, or maybe a terrorist attack, or plausibly even a flu pandemic. But its arrival is much more likely to come from a fatal car accident, or a heart attack, or even a series of simple retirements.

That coming Day of Reckoning will cause governments to experience catastrophic shutdowns. Data will be lost, perhaps permanently. People will suffer. Lives will be threatened.

Now you may ask: Surely you aren't saying that a retirement could cause a catastrophic Day of Reckoning?

Don't call me Shirley. And this nightmare scenario is easy to comprehend, even predict, if you are a government CIO overseeing a monumentally ancient mainframe or midrange system that desperately needs modernization. And I am betting that in state capitals across America, and in our nation's Capitol, precious few resources are slotted toward the replacement or refreshment of these obsolete systems.

I was sitting down the other night, preparing to watch the film Forbidden Planet on my HD-DVD player (I have both HD-DVD and Blu-Ray players, so I am format-agnostic or "purple" in the lexicon of Hi-Def). But I digress. Anyway, I was at the part where Walter Pidgeon shows a pre-Drebin Leslie Nielsen the workings of the Krell, the super-race that once inhabited planet Altair-4. The Krell had died over 100,000 years ago, yet their subterranean machines continued to run themselves.

Do you detect a theme here? Government decision-makers act as if we are the Krell, master designers of supermachines that will run unabated for eons.

It's time to proclaim: We ain't the Krell. Far from it: Here in sunny Florida, for example, I estimate that we are ten years from absolute chaos, as mainframe systems that have lived way beyond what their creators intended will begin shutting down due to lack of maintenance, overexertion, lack of capacity, or simply because there was some type of limitation in the scalability of the solution.

Why? Because the people who wrote, maintained, and serviced these solutions are retiring (and dying) in frightfully high numbers. These people almost all started out in other fields and were eventually drawn to mainframe programming. But they wrote in older machine language that now is as rare as Mayan to duplicate. It is easier to find a Latin scholar in your city than it is a programmer immersed in these older languages. It is even easier to find an ambulatory World War II veteran.

And despite protestations from IBM and others that a new vanguard of COBOL and IMS programmers are being manufactured in leading universities as I write this blog -- forgive me if my father was from Missouri (he was) and I inherited his skepticism.

Compounding this is the failure of some very high-profile, and high-dollar, system conversions. We have had our fair share in this state, and I am sure your state can offer similar examples. But these failures should not serve as anything but a cautionary tale that we must perform better project management and, even before that, better requirements gathering before embarking on future conversions.

I will also now recognize my old friend, The Elephant In the Room. And that is that in order for these conversions to be ultimately successful, government must learn to do what the private sector learns every day. That is: Government must learn to reform its business processes and become more digital. It must re-learn how it does its business, and must move away from archaic business models.

There is nothing harder in society than for government to reinvent itself. But that, at the end of the day, is why so many government conversions fail - because the requirements gathering process reveals and discloses human business processes that more closely resemble the 19th century than the 21st. Trying to duplicate those wagon-wheel business processes digitally is madness.

I was the CIO at the Florida Department of Corrections for nearly six years. We began reimagining how we might transform the Offender-Based Information System (OBIS) from my first week on the job. Parts of OBIS are old enough today to run for Congress, and I am not exaggerating.

We could see benefits such as drastic streamlining of business processes, elimination of massive re-keying of data, shortened reporting from harried and overworked probation officers, quicker and on-the-fly aggregation of inmate and probationer data, and the virtual elimination of paper. But the price tag - somewhere between $40 million and $60 million, depending on which alphabet-soup consulting firm prepared its analysis - scared off any serious look at conversion. And now, with the national economy tanking and Florida's economy suffering from the double-whammy of usurious insurance premiums and confiscatory property tax rates - well, let's just say that since OBIS runs today, and there are very legitimate and pressing issues such as paying for insulin for the diabetic poor and equally heart-wrenching issues - OBIS transformation will get a pass for a few years.

This digital business process conversion also meant something else - and that is the potential for reduction of the workforce. To government employees, nothing could be scarier than that prospect. Never mind that these employees would be probably repurposed, and that tens of billions of dollars in taxpayer funds could be saved by these realignments and analog-to-digital business process conversions. Lost Full-Time Equivalent (FTE) positions is blasphemy in some circles of government. I fought the law and the law won. Bureaucracy is an amazing thing, no matter what sector you work in.

Eventually, the economy will rebound, and government coffers will fill. So when revenues start to pick up, leaders must return to this issue with gusto. They must look at both sides of this expensive coin. Heads: They must reform government and rethink everything regarding its very nature, structure and purpose. Tails: Then they must decide to digitize as much as possible. I don't mean information: I mean process.

Finally, they must fund quickly and turn the IT people and business process experts loose to develop those replacement applications - with all appropriate oversight and project management.

Otherwise, the headlines will proclaim the meltdown of this state's welfare system and that state's inmate system; the return to paper calculations of inmate sentences (in Florida today, literally hundreds of sentences have to be hand-calculated, because new laws and new sentences circumvent the ability of the obsolete OBIS system to calculate them); and people won't get checks because the mainframe solution broke, and there was no one available to repair it.

Kind of like Humpty-Dumpty. Now on whose political watch such calamities will occur is not possible to predict. But I will bet that person is already somewhere in government today, possibly thinking about being Governor, maybe even President. And that person will be reeling politically from decisions that were never made by their predecessors. These people would do well to read this blog, print it and tuck it away in their Bibles.

Just remember: We Ain't The Krell.

Computerworld's IT Salary Survey 2017 results
Shop Tech Products at Amazon