Commentary--The story of how Microsoft used its monopoly in operating systems to acquire a dominant position in office applications and browsers has often been told. But there's another Microsoft monopoly that's rarely mentioned, even though most of us see it every day--right in front of our eyes. Microsoft's fonts are used to display most Web pages on the planet. Even Linux and Mac users, who often have fled Windows to avoid dependence on Microsoft, read most of their content using Microsoft fonts.
Microsoft's font monopoly is due to the "Core fonts for the Web" program it launched in 1996. About 10 font families--including familiar names like Arial, Georgia, Verdana and Times New Roman--were made available "for free to the Web community, on all platforms" as Microsoft told the World Wide Web Consortium (W3C) at the time. The fonts have served us well. They've improved both aesthetics and interoperability on the Web, and they look good in a wide range of sizes. Unfortunately, Microsoft decided to close the project in 2002. The fonts are still available for anyone to use, but not to change. It is illegal to add support for more nonwestern scripts.
The time has come to break the Microsoft monopoly on fonts. This is easier than it sounds. There are thousands of font families on the Web--I call them Web fonts--that are freely available for anyone to use. One such font family is "Goodfish," which I will use as an example. Goodfish is an elegant serifed font designed by Ray Larabie in 2000. It comes in four variants (regular, italic, bold, bold italic), which are encoded as four TrueType files. When zipped up, the files take up about 100k of memory. That's about the same file size as a photograph.
Ray Larabie has generously allowed Goodfish and other fonts to be used by anyone, for free. And there are many other font designers on the Web. Some of their fonts look weird, and some only cover the English alphabet. Some are only suitable for print or for use in headings. Still, these fonts represent a huge untapped typographic resource for the Web.
In order to use Web fonts, browsers must be modified to start looking for TrueType files outside of the local machine. CSS2 style sheets can already refer to Web fonts, so there's no need for a new standard.
Browsers are already pretty good at finding information on the Web, and adding fonts to the list should be simple. There are plenty of benefits waiting on the other side: Font designers will find an outlet for their creativity, users will get visually richer content, and nonwestern scripts can easily be added. Also, Web page designers can often use Web fonts instead of images to get their designs across.
Some will claim that Web fonts cannot be supported due to the risk of piracy. It is feared that people will put fonts on the Web without having permission to do so. As we know, this sometimes happens with images. Unlike images, however, TrueType fonts carry with them information about permissible use. The font will know whether it's OK for it to be installed on a remote system. Only fonts that allow this should be considered for use by browsers. And even if the font is installable, browsers should not install it so that other applications can use the font. The font should be used only by the browser, and perhaps to display only pages that actively request that font.
Just as the visual appearance of the Web changed dramatically when images were introduced by Mosaic in 1993, the Web can change yet again if browsers start supporting Web fonts. I believe it will benefit everyone on the Web.
Except, perhaps, the monopolist.
biography
Hakon Wium Lie is chief technology officer of Opera Software.
