By SARA INÉS CALDERÓN (TechCrunch)
The Internet doesn’t like me — or, at least, it doesn’t care much for my name. My first name consists of two words and I have accent marks in both my first and last names, which seems to complicate my online life considerably. When trying to purchase an airline ticket or sign up for an email newsletter, I’m never completely certain whether it will go through, or how my name will come out of the transaction, but I’m usually pretty sure it won’t be right.
Turns out it’s not me. Rather it’s the archaic remnants of how computers came to be programmed in the U.S. during the mid-twentieth century. When standards for exchanging data between computers such as ASCII were created, the workforce consisted mostly of white males. They created programs still widely used today recognizing certain symbols and names — accent marks are not typically among them.
“There is no good standard for representing names on the Internet. It’s almost a vestigial remnant of our programatic past; programs were highly limited to things that were in use in America on typewriters,” said Dan Lynn, CTO of FullContact, a cloud address book and contact management provider.
In other words, Lynn said, there were no accents because the creators of these systems didn’t see a need for them. Nowadays, programs that don’t recognize accents or two-word names are in use by choice. Supporting multiple character sets may be more costly, because you have to support that input with databases, but ultimately Lynn said the internationalization of characters is on the rise.
Internationalization, or i18n, is an entire field dedicated to making software appropriate to language and culture, according to Professor Rob Miller, who teaches human-computer interaction and software engineering at the Massachusetts Institute of Technology. More parochial views of software development, wherein just first and last names with no special characters are used, are increasingly rare.
But, Miller explained, when older systems interact with newer ones (say, when purchasing an airline ticket), older systems may not recognize special characters, putting a damper on the whole transaction. When I try to purchase an airline ticket as “Sara Inés Calderón,” the airline company might be able to recognize those symbols, but the outdated communication protocols of other agencies — federal regulators, law enforcement, customs and whoever else — may not.
“I think software is slowly conforming to us. It’s just inconceivable that English will be the sole winner of the Internet,” Miller said, noting that bigger tech companies such as Apple, Microsoft, and Google are keenly focused on internationalization. As these companies incorporate more diverse programming, others will follow, and eventually because the Internet is so global, older systems will be replaced, too.
It’s not technically that difficult to change, said Coram Bryant, the Director of Analytics at Motion Math. Web interfaces already support full character sets, he said, building databases with multiple fields for names (say, two first names) is not difficult — it just takes some design. Technically this is not an astronomical hurdle, it’s more an issue of who’s doing the programming, Bryant said.
Miller of MIT believes that, ultimately, technology will conform to our changing culture. We’re always negotiating between our evolving culture and software programs, he said, but change is only a matter of time. Which may mean I might be able to purchase an airline ticket with my real name any day now.