Jun232009A Few Thoughts on TypeKit
If you haven’t heard about it yet, TypeKit is a soon-to-be-released service that will allow Web designers to use fonts of their choosing without worrying whether a visitor has the same font installed. Normally, when we specify a font using CSS, the visitor viewing the Web page must also have that font installed in order to see it. In other words, if I want to use “Warnock Pro” on a page, then anyone visiting the page must also have “Warnock Pro” installed (if they don’t they’ll just see a back up font like “Times”).
Browser vendors have various solutions to this problem. Microsoft’s Embedded OpenType format (http://msdn.microsoft.com/en-us/library/ms533034.aspx) has been around for years, but only works in Internet Explorer (and involves a craze-complicated procedure to create the proper EOT file). The other solution, the @font-face CSS rule is supported by Safari, Chrome, and soon by Firefox 3.5 and Opera 10 (you can read a quick intro to @font-face at the CSS3.info site). Both EOT and @font-face let you use your own fonts on your Web pages regardless of what fonts the end-user has installed. Unfortunately, these two solutions aren’t compatible, and you have to go through some weird contortions to use your own fonts in a way that works in all of these browsers.
Further complicating matters is the fact, that in most cases, it’s probably not legal for you to use any old font you have installed on your computer. Most font vendors sell fonts with specific licenses that detail the allowed uses for their fonts. While you may be able to use a font you buy from Adobe to create a banner graphic, embedding it into a Web page is probably not allowed. Font vendors are understandably worried about the @font-face rule, since you have to place the actual font-file (truetype or opentype file) on your Web server. Anyone could look at your CSS, see the URL of the font and download it — essentially stealing the font. (The EOT file format includes protections that make downloading and reusing the font file impossible.)
Now this brings us to TypeKit, which promises to solve these problems. Basically, TypeKit is a hosted service that lets you use fonts from the TypeKit server. While the details haven’t been spelled out yet, it sounds like you access the fonts you want by adding a bit of JavaScript to each page in your site. The script contacts the TypeKit server and does whatever it does to get the font to the end user. (There are some details of how the service works over on the TypeKit blog.)
While this sounds like an ingenious solution to the “web font problem” (and it will be if it works), I’m not so sure it will succeed. Now you can call me an old internet curmudgeon (just not to my face, please), but I’ve seen many a great sounding Web service come and go during my many years on the Web. For me, there are a few things that make me doubt this will succeed:
- Cost. They haven’t yet set out pricing for this product, but if it works like other Web services there will probably be a monthly fee, perhaps also a scale depending upon how much traffic your site receives. I, for one, am not sure that I’d want to pay a monthly fee for the life of my Web site just to make sure my type is the way I want it. If I decide to stop paying, then I’ll lose the font and my pages will fallback to the old standards — Times, Arial, Verdana, Georgia, etc. Personally, I’d be happy to pay a one-time, per-site fee to a type vendor to use a font on my site. I buy fonts all the time, and find it fair that there should be a second “web licensing” fee for a font.
- Over-estimating the market. Hopefully the folks at TypeKit have done a lot of market research, and there is certainly a lot of interest in Web typography at the moment, but I have my doubts that the majority of people who build Web sites are obsessing about typography. Don’t get my wrong — I love typography. Ok, not as much as this fellow, but I do love typography. I spent years in print publishing obsessing over fonts, but I think the handful of excellent Web designers who are writing about Web type are in the minority. This isn’t really a criticism of the TypeKit service, just a feeling that perhaps it doesn’t have a sustainable business model
- Dependence. I don’t want the look of my site dependent upon another server’s up-time. If the TypeKit servers are down, you’ll get nothing but your back-up fonts (I also wonder how a page will respond while the browser is waiting for the javascript request it sends to the TypeKit server). In addition, what happens if the server is up, but is slowed by a barrage of requests from servers around the world? I assume the page simply won’t render until the JavaScript request is processed, so does the visitor have a blank page? We won’t know until we can see the TypeKit service in action, but I’m very wary of relying so heavily on a hosted service like this. (Sure you could say we do this kind of stuff all the time by embedding Twitter, Flickr and RSS feeds on our sites, but I like to keep external dependencies down to a minimum.)
- Most fonts just don’t look that good on screen. OK, I may catch a lot of flack for this, but there really are a handful of fonts that look good in small print on the screen. Microsoft got it right with their Core Fonts For the Web — handcrafted fonts designed to be “highly legible” on screen. Most fonts are designed for the incredible resolution of print, and look great when printed. Most fonts also look pretty good when they’re anti-aliased in graphics applications (that’s why you can make some highly legible logos and banners for your Web pages using Photoshop.) If you have Safari, check out a few examples of @font-face in action. To me, they just don’t look that good. The serifs, especially at smaller sizes, are fuzzy, and the letters generally lack definition. Hopefully, font vendors will eventually start designing fonts (I’m sure some already do) that are just intended for screen viewing, and which take into account the limited resolution of a computer monitor. I’m not that interested in making my text less legible, and certainly wouldn’t want to pay to do that.
I’m not sure what the solution to the font problem is, but @font-face is simple and elegant, and in the world of the Web, simple and elegant usually win. Once Firefox 3.5 and Opera 10 are out, there will be a large percentage of the Web surfing world that can use @font-face, and there will be lots of designers starting to take advanted of it.
