In The 21st Century, Computers Still Trip Over Human Names

You'd think that in this day and age of the Internet, self-driving cars, and wayward chatbot AI's, computers would no longer have problems dealing with human names, whatever their form. But as some people from around the world have discovered, to their amusement and exasperation, that is plainly not the case. Today's software, especially database software, are still ill-equipped to handle corner cases when it comes to names of real, not made up, people, causing no small amount of inconvenience to those and practically stalling the evolution of information systems.

Advertisement

Jennifer Null got her surname from her husband, who warned her well enough about possible problems she might inherit just by marrying him. Love, of course, conquers all. That is, except computer software. While seemingly innocuous, though odd, "null" is also the word used by databases to refer to something that is empty, causing forms on online software, like for booking plane tickets or even filing taxes, to have a fit, thinking that her surname is just blank. Jennifer would have to phone in to get things corrected.

"McKenzie" might be a more common name, but Patrick McKenzie, who is, ironically, a software programmer, isn't safe from such software oversight. In Japan, it is uncommon to have a name with that many characters, because Japanese names can be shortened to a few Katakana characters. Sadly, Japanese software are designed with papre forms in mind, limiting the number of boxes for characters way below McKenzie's name.

Advertisement

And then there's the case of Hawaiian local Janice, whose "Keihanaikukauakahihulihe'ekahaunaele" is both rare, odd, and extremely long. 36 characters in total, most systems, whether physical or computerized, do not account for that many characters.

These are often called edge or corner cases because they are exactly that. They represent the outliers, those outside the norms. As such, software engineers usually don't test for these cases, either because of a lack in resources or, like in the case of Japan, they are designed with certain conventions in mind.

The problem is that these conventions are pretty much local in nature. McKenzie's plight, for example, is caused by Japanese software not designed with Western names in mind. Jennifer's situation, on the other hand, as the bad luck of coinciding with a well-known and widely used keyword in software.

The situation today has improved but there's still a lot of work to be done. Technology is making the world a much smaller place and pretty soon those edge cases won't be as rare anymore. Considering how names are relatively more static and harder to change, the onus falls on software engineers to modify programs to take these into account. Then again, considering all the legacy software and critical systems that need to be changed and tested, it might actually be easier and faster to get you name changed in the end.

Advertisement

SOURCE: BBC

Recommended

Advertisement