The Pull of the Present

By this time twenty years ago, many of us were feeling relieved. We’d been hearing for months about the near-certain fallout from the “Y2K bug”: widespread computer system failures caused by the practice of shortening years to two digits instead of four (e.g., 99 rather than 1999.) But by mid-January, 2000, it was clear that all would be ok. Or so it seemed.

Some context, in case you weren’t around then. By the mid-1990s, computer systems were already essential parts of our infrastructure. Nobody knew how many of these computers had the bug or what would happen after 11:59 pm on December 31, 1999, when these systems would assume it was now January 1 of year zero. Would there be blackouts? Urban transport cancellations? Airplane collisions? The complexity of such infrastructure-level systems made the consequences impossible to predict. Governments and companies undertook massive and expensive projects to “fix” the problem. FORTRAN programmers suddenly found their skills in demand.

Then nothing happened. By the end of the first week of January 2000, it was clear that either the fixes had been successful or the potential downsides overblown. Those of us who’d been stressing out about the Y2K bug felt relieved and quickly forgot about it.

And yet, there was a clear lesson to be gleaned from the incident. We would’ve spared much trouble and treasure if systems designers had thought ahead to the turn of the millennium. (To be fair, some old computer systems used shortened dates due to technical constraints.) Some system designers may not have thought about long-term scenarios at all. But some did, yet proceeded anyway assuming — incorrectly, as it turned out — that their systems wouldn’t still be in use by the turn of the century.

Perhaps this attitude was forgivable fifty or forty years ago when advances in core technologies were more obvious, but it seems less responsible today. And yet, there are reports of a new batch of Y2K bug-like problems that have emerged as the clock turned on January 1, 2020. Consider this report in New Scientist:

Thousands of cash registers manufactured by Polish firm Novitus have been unable to print receipts due to a glitch in the register’s clock. The company is attempting to fix the machines.

WWE 2K20, a professional wrestling video game, also stopped working at midnight on 1 January 2020. Within 24 hours, the game’s developers, 2K, issued a downloadable fix.

Another piece of software, Splunk, which ironically looks for errors in computer systems, was found to be vulnerable to the Y2020 bug in November. The company rolled out a fix to users the same week – which include 92 of the Fortune 100, the top 100 companies in the US.

The problem? Instead of changing systems to understand four-digit dates, programmers hired to fix the original Y2K bug opted for a faster/cheaper approach called “windowing,” which kicked the can down the road. We’ve now come to the can.

The urgent needs of the present take precedence over the needs of the future. When your kitchen is on fire, you look for any means to put it out — and leave questions of sustainability or longevity out of the process. But when the emergency passes, only a fool carries on as before without investigating and remedying the causal factors.

(Alas, when dealing with complex systems, the emergency itself may not be apparent. We assume all’s well because we see no flame – even in the presence of smoke — and because we want all to be well. Significant changes are often expensive and sometimes painful; they require sacrificing our near-term comfort for the needs of other, distant people — perhaps some who aren’t even yet alive.)

New Scientist: A lazy fix 20 years ago means the Y2K bug is taking down computers now