Web fonts and language equality
A while back I posted about an interesting experiment on sending fonts over the web as Javascript. Here’s a look at a more robust solution to the same problem: web fonts.
Check out this list of Wikipedias. Do you see a lot of question marks or little meaningless boxes? The reason you get that junk is that you don’t happen to have the right fonts installed locally on your computer.
That sucks. It’s not fair. Why should English or French or Russian or Japanese speakers be treated as first-class citizens of the web, but Tibetan or Khmer or Inuktitut speakers be treated as weirdos?
When “web fonts” catch on, this regrettable state of affairs could start to fade away. Content providers will be able to host fonts on the server, right next to the content requiring those fonts, thus ensuring that readers will be able to see content in any language.
John Resig has an article that explains how all this works: An introduction to W3C Web Fonts. Happily, as John points out, it seems that browser support for web fonts is on the upswing.
I’m convinced that a lack of fonts is a major barrier to an increase in the amount of content in certain languages on the web. Web fonts would go a long way toward fulfilling the “world wide” promise of the World Wide Web.
web fonts and the ability to add yet another source of untrusted code to the browser are totally unnecessary (and besides they only work for the browser, not the gazillon of other apps users use to see/manipulate text).
What’s needed is:
– a repository of trusted fonts with complete unicode coverage http://fedoraproject.org/wiki/Category:Packaged_fonts
— a way for apps to install fonts from this source as needed http://fedoraproject.org/wiki/Features/AutomaticFontInstallation
Hi Nicholas,
Thanks for your thoughts. I agree that a public repository of fonts is very important, and I think that your work on DejaVu fonts and other similar font-related stuff is a real boon to the community.
The reason that I think that web fonts are worth promoting is primarily a social one: web fonts represent a partial solution to users who have very little access at all to their own languages on the web.
Imagine, for instance, a Tamil-speaking immigrant to Vietnam (there are some). There are plenty of internet cafés in Vietnam, but how many of them will have Tamil fonts installed? How many café proprietors will allow their users to install fonts? Not so many, I’d wager. They probably run a recent Windows (unfortunately ;-) ) for games and such, but they will jump through various hoops to see to it that their users can’t change anything.
A Tamil web font, at least, allows that user to write an email home in his own language.
Yes, the real solution is free fonts for all scripts, distributed with every OS. But we’re not there yet. In fact, for proprietary OS’s, we may never get there, because it’s not in those companies’ interest, necessarily, to underwrite that work.
My point is simply that it’s not fair to speakers of languages with unusual scripts to make them wait for that universal solution, when web fonts are available now, and at least let them get in the game.
I may be overly pessimistic, but I feel that in the web fonts affair the i18n/translation community is taken massively advantage of, and its members are so blinded by small short term gains they do not realise they serve as an alibi to everyone else.
In the web fonts, proprietary foundry version, everything will be locked down so your Tamil immigrant won’t be able to write a mail home unless he uses a webmail account from some company that paid dearly for the use of a Tamil font (and his Internet will be limited to the sites of such companies).
In the web fonts, @font-face version, everyone will be able to use any font file he wants, which will only help copying of well-known proprietary fonts and spreading all the shady-licensed badly-encoded fonts out there. This will help a little short-term, and drain resources from free font projects mid-term.
A proper solution to fonts for everyone needs to include the legal aspects . That is, if you want normal people to use fonts without fearing trials. This won’t happen in the first case (just EULA madness all again), and this won’t happen in the second case (it used, for example, to be totally impossible to find a properly localised Linux system in Russia or China, so prevalent where the English 1$ copies of windows sold in every other street).
But, again, maybe I’m overly pessimistic.
I may be overly pessimistic, but I feel that in the web fonts affair the i18n/translation community is taken massively advantage of, and its members are so blinded by small short term gains they do not realise they serve as an alibi to everyone else.
In the web fonts, proprietary foundry version, everything will be locked down so your Tamil immigrant won’t be able to write a mail home unless he uses a webmail account from some company that paid dearly for the use of a Tamil font (and his Internet will be limited to the sites of such companies).
In the web fonts, @font-face version, everyone will be able to use any font file he wants, which will only help copying of well-known proprietary fonts and spreading all the shady-licensed badly-encoded fonts out there. This will help a little short-term, and drain resources from free font projects mid-term.
A proper solution to fonts for everyone needs to include the legal aspects. That is, if you want normal people to use fonts without fearing trials (that most people will get by without being sued is not a good argument — PTBs love arbitrary trials). This won’t happen in the first case (just EULA madness all again), and this won’t happen in the second case (it used, for example, to be totally impossible to find a properly localised Linux system in Russia or China, so prevalent where the English 1$ copies of windows sold in every other street).
But, again, maybe I’m overly pessimistic.
Web fonts, are a partial solution. We currently have two competing models; the WebKit method and the Internet Explorer method. And WEFT is old, archaic and buggy tool that will not even install on Vista, and may not work correctly in all instances.
Its a great idea, but until there is a single approach, and appropriate tools, it is still an immature technology.
Web fonts will do very little, unless the OS being used already has the capacity to render text correctly for the language in question. So Tibetan, Khmer and many other languages will remain restricted, even with web fonts.
You also need an input system to allow them to type in the desired language.
I suppose I’m pessimistic, but the effective use of these technologies will really depend on individual web developers building in the appropriate support into their web sites and web services.
I’m looking at using web fonts in one of our projects, and incorporating on-screen keyboards into the site, but there isn’t any way to get around the rendering issues. You’re still restricted to operating systems that can render your language correctly.
And here is doesn’t matter whether its a commercial or Open source Operating System.