The Challenge of Internationalization

This talk was delivered by Richard Ishida at the meeting to celebrate the 10th anniversary of the World Wide Web Consortium in Europe, in June 2005.

Additional commentary will be added shortly to convey the messages on the slides.

Please send any comments to ishida@w3.org.

One of the things to note on this slide is that English is just another language.

Another is that only a few years ago it would have been extremely unusual to see all these scripts correctly rendered on the same page. Initially this was a problem related to character encoding.

Not so long ago, the most widely used character encoding was ASCII. This was a character encoding based on a one-byte-per-character approach, where one byte contained 7-bits. The 7-bits allowed for a grid of 128 possible numbers, each of which could be used to represent a different character. Some numbers were associated with control codes, and the graphical characters in the set comprised the English alphabet and a number of symbols.

For Western Europe additional accented characters were needed. When 8-bit bytes became more prevalent, one byte could express 256 different characters. This was better, but still not good enough.

To support scripts such as Greek we resorted to multiple code pages. These had to be swapped in and out to deal with text in multiple scripts, but could not be combined on a single Web page.

All of this was academic to people in the Far East. This slide shows the size of some of the character sets that had to be encoded for this region. Each country requires the availability of thousands of characters at a time. This led to the creation of 'double-byte' encodings. Even so, you could not usually represent more than one of the Far Eastern languages per encoding.