2038 effect: Is a new Y2K on the horizon?

The Y2K bug and the fear that the world would be paralysed by the unleashing of a multitude of computer problems could have an aftershock relatively soon.

Some are already pointing to 2038 Effect and the possibility of digital system failures in thirteen years’ time.

This concern originated from the fact that many programs and operating systems are based on the C programming language with a 32-bit timer. Its maximum counting capacity will be reached in 2038, specifically on 19 January.

This is what the University of Design, Innovation and Technology (UDIT) warns, coinciding with the celebration of International Mathematics Day.

‘If not corrected before that day, these 32-bit systems will stop interpreting dates correctly,’ they warn, ’which could affect critical infrastructures, from financial services to electronic devices and telecommunications networks’.

UDIT’s technology coordinator, Sandra Garrido, explains that the problem is related to the Unix time system.

‘Unix time is a reference system that measures time continuously by means of a 32-bit counter, which stores the seconds elapsed since 1 January 1970,’ she explains.

‘Although it has been a standard for decades, its storage capacity is limited to 2,147,483,647 seconds from that starting point,’ she explains, ’which is equivalent to about 78 years.

And, ‘although hardware and systems have evolved, many Unix-based devices, such as those operating under macOS,’ he notes, ‘still use this structure’.

‘This means that, on 19 January 2038, the counter will reach its limit by resetting itself,’ Garrido summarises, ’so many systems could misinterpret the date as 13 December 1901, causing all kinds of failures’.

The solution to avoid technological chaos is mathematical, including concepts such as modular arithmetic and the theory of numerical representation.

For example, the 2038 Effect could be stopped before it started by migrating devices to 64-bit architectures, capable of storing 9,223,372,036,854,775,807 seconds.

‘This transition is not just a hardware change,’ says Sandra Garrido, ’but requires a profound redesign of the algorithms that manage time in operating systems and critical software’.