Andrea DiMaio

Yesterday I read a most amazing news: German banks as well as a few companies (Symantec being one) were hit by something incredibly similar to the Millennium (or Y2K) Bug, i.e. the inability of many computer programs to deal with dates past January 1st 2000, due to the old tradition of using two digits only to represent years in date fields.

I found ironic that something like this happens ten years later. I spent about three years of my career, both when I was with the European Commission and later with Gartner, advising enterprises about how to solve the Y2K problem and manage its risks. I still remember that, as Gartner had taken quite a high profile on this topic, analysts covering this were on duty throughout New Year’s eve, waiting to report about possible incidents and to help clients where needed. I do vividly remember that I was given a satellite phone to make sure I could still be reachable in case wired and wireless phones had any issues. I also remember that I was on my terrace, holding this heavy satellite phone, few minutes after midnight, while fireworks were exploding quite close to my terrace: actually there was no Y2K risk I can remember, but I felt like a reporter in a war zone, with all these explosions around me.

I could not find any detail yet about the root cause of this belated date problem and whether it is related in any way to corrective actions taken to solve the Y2K problem. However this is a useful reminder that IT professionals do not always learn from their own mistakes and do not always keep risk management as well as “boundary management” (the set of precautions to be taken to prepare for a major date change event, such as a decade, a century or a millennium) into as much account as they should.

Maybe these were totally new systems, maybe programmers and project managers were of a new generation, and never took their predecessors too seriously, as technology now does not impose constraints like using only six characters for a date (memory and processing power are so cheap). After all, the Y2K meltdown that some scaremongers predicted never happened: but most likely it didn’t because of the work and effort of so many people who took corrective actions.

We have been through this also for the introduction of the new European currency, which is being joined by more and more countries: how many applications were patched rather than redesigned to move from the old to the new currency? How many little bugs are still around and could hunt us back when we least expect them?

One thing is for sure: as my colleague Nick Jones says, we have to accept that software is inherently imperfect and act accordingly. Failures are always behind the corner.

2 Comments »

Category: Uncategorized     Tags: risk management

Share