A Note on the Type

Compared with the world of print, the Web is a typographically impoverished place. It was built to display the fonts that users already have loaded on their computers, which in practice means the 10 “core fonts for the Web” that Microsoft started bundling with its Windows operating system in 1996. A few of these fonts are admirable–Verdana and Georgia, for example, which were designed by Matthew Carter specifically for the computer screen. But as the Web grows more sophisticated, and as the need to improve on-screen legibility becomes more urgent, they are far from sufficient.

You may not have noticed the problem–in part because the Web is filled with flashy headlines and logos, which can be more visually diverse because designers display them as images. But significantly absent are the workhorse text faces that occupied the great type designers of the past five centuries, like Garamond, Caslon, and Baskerville. These typefaces were sturdy and legible, graced with small but essential touches that elicit a curious passion in designers and laypeople alike. Fonts speak to people, probably because they give form to their written words. People brag about their favorite font and complain about the awful one their boss makes them use. Corporations are believers, too: they regularly commission expensive new fonts to lend flattering associations to their brands, and they wish they could do the same online.

Web designers and type designers have been planning a better future for at least 11 years, ever since the technical foundation was laid for browsers to load fonts stored on remote servers rather than on a computer’s own hard drive. For the first time, this prospect feels tantalizingly close. As of this past June, with the release of Mozilla Firefox 3.5, all the major browsers finally support what’s called the @font-face rule, a way to use the Cascading Style Sheets language to designate a remote font. As designers try to develop a format that delivers the fonts efficiently and securely, visions of the future are taking shape. The question of which will be realized was debated at this summer’s ­TypeCon, a conference of type and graphic designers. The answer will be determined by decisions made over the coming months by browser makers, type designers, and Web designers, some in their capacity as representatives to the Web’s international standards organization, the W3C.

In one possible scenario, history repeats itself and we still don’t get new fonts on the Web. Twelve years ago, Microsoft released a format–called Embedded OpenType, or EOT–that was meant to launch the Web-font era. EOT, which was developed to embed fonts in Microsoft Office documents, worked as a “wrapper” for desktop font files. It compressed them for downloading, included instructions about how they might legally be used, and allowed for the fonts to be encrypted. But Netscape, then a formidable competitor to Microsoft’s Internet Explorer, chose a different format, leaving graphic designers with the choice of designing two versions of every Web page or leaving well enough alone. Web fonts never arrived.

Today, Microsoft is again promoting EOT at a time when its competitors have chosen another route. There would be two significant advantages to establishing EOT as the standard: first, the provisions for protecting and optimizing fonts, and second, the potential for swift change. Any solution that works immediately in Internet Explorer could eliminate years of waiting; IE users still represent about 65 percent of the market, and they take significantly longer to upgrade than users of other browsers. (Some are big corporations with conservative upgrade policies.) Microsoft has tried to answer objections from open-source advocates, including its largest competitor, Mozilla, by opening up the proprietary components of EOT. But with no sign so far that other browser makers are willing to adopt Microsoft’s standard and thus extend its dominance, a format impasse could spell another long dry period for Web fonts.

All the browsers but Microsoft’s, meanwhile, have embraced a technique called “naked” or “raw” font linking, which means uploading ordinary desktop fonts onto servers. What if Microsoft, in the interest of guarding its diminishing share of the browser market, abandoned EOT in favor of this standard? Though it would take a few years for IE users to upgrade, soon enough the Web would be typographically transformed. But at a cost: in this new world, fonts would have no protection from piracy. After all, when your browser downloads a remote font onto your computer for temporary use (as it would have to), that font functions like any other on your hard drive. Who could blame you for thinking it was your property, no matter what the license might say?

Open-source advocates are not particularly worried about this scenario; they see piracy as a challenge that can be addressed through legal action and educational campaigns. But type designers see a world of raw fonts as a nightmare that could do to them what file sharing did to the music industry. “You’re throwing up a desktop font on a Web server, and there’s no inherent protection for the font there,” Thomas ­Phinney, formerly of Adobe’s type group, said at ­TypeCon. “Unsurprisingly, this is freaking scary to type foundries.” They could dig in their heels and write licenses that forbid using their fonts on the Web, but the price of sitting out the Web-font revolution would probably be too high to accept.

With a bit more patience, though, a third option could emerge–one that type designers see as ideal. Faced with the limitations of EOT and the risks of raw font linking, type designers and software engineers have been compiling a wish list for a brand-new, built-from-scratch font format. This system would offer better file compression to accommodate low-memory environments like mobile phones, and it would have a flexible metadata structure to embed information about permissions, but without turning browsers into enforcers.

Such a format has actually been created: Web Open Font Format, or WOFF, the combined work of the type designers Tal ­Leming and Erik van Blokland and Mozilla’s ­Jonathan Kew. A petition supporting WOFF began circulating in July, and several dozen type foundries, including important ones such as Carter & Cone and Hoefler & Frere-Jones, have signed. Mozilla and others are already testing code for WOFF (and also a less ambitious but backward-compatible format called EOT Lite). This option seems like the best hope for type designers, but since it might take time for all the browsers to implement it and for users to upgrade, it depends on the Web community’s patience.

While it remains to be seen if this ideal can be achieved, a startup called Small Batch is poised to launch an alternative with a service called Typekit. Small Batch (and also the startup Kernest) is offering to be a middleman: it would look after security issues for type designers and browser-­compatibility issues for Web designers, who–instead of linking to fonts directly–would use JavaScript to link to Typekit. The company would create a new business model, transforming fonts from goods into a service. Rather than buying a font once and being able to use it indefinitely, Web designers (or their clients) would pay a recurring fee to buy access to one or more fonts in a library. (Typekit does offer some free fonts, too.) Users would lose access to the fonts when they stopped paying, but no fonts are really forever, anyway: formats change and type designers make improvements. In the service model, those improvements would be made along the way. Not only would fonts be kept up to date, but they’d be kept compatible with evolving platforms in an ever-­changing browser market. “Until there’s one format, one browser, and one operating system, there’s a lot we can do to help,” says Bryan Mason, one of Small Batch’s founders.

Finding a solution that balances protection and freedom could redeem the missteps of a generation ago, when U.S. copyright law declined to cover typeface designs. The argument at the time was that protecting them would stifle innovation by concentrating the most popular designs in the hands of a few major foundries. But in effect, this decision permitted rampant plagiarism and thus removed the incentive to invest in type development. “We lost several decades in which corporate research could have contributed to our understanding of typographic legibility, aesthetics, and ergonomics,” says Charles Bigelow, a distinguished type historian at Rochester Institute of Technology and one of the designers of the Mac OS X system font, Lucida Grande. Now copyright protection has been extended to fonts as software, and type design has made significant advances–for example, the text-rendering technology ClearType, a result of Microsoft’s support for research into legibility. Bringing more fonts to the Web could lead to further progress. One prominent champion of Web fonts, Håkon Wium Lie, described the significance to me this way: “Archeologists of the future will classify Web pages into pre-font and post-font eras.”

Joshua J. Friedman, a former editor at the ­Atlantic and Boston Review, is a writer based in New York City.