Si was a Reading graduate, Tom Phinney says.

DANIELS: How many people consider themselves type designers (One-third.) Web designers? (Few.) Kind of a half-and-half split. I’m lead PM at Microsoft for fonts. I work with all the product groups at Microsoft and with outside designers for anything that ships with fonts. I work with Peter Constable, who owns the font technology aspect – shaping engines, Fonts folder, things like that.

The font-embedding story covers the fonts and our relationships with font vendors and the technology side. Technology that supports embedded fonts in all our applications is something that Peter wrote. I’ve been involved in font embedding for 10 or 12 years at Microsoft, and even as an intern in 1995.

(Shows PowerPoint-like slide of history of type on the Web.) Anybody can shout out objections if they want. I’m thinking of Joe in particular.

  • 1994: Mosaic lets users define fonts.
  • 1995: IE introduces font face
  • 1996: EOTs, PFRs, and CSS [whatever they are]
  • 1998–2006: “Nothing much happens.” PFR dropped, Bitstream deletes the PFR Web-font maker. Mac IE drops EOT. Microsoft deletes Web fonts. Other attempts: Fairy, Flash, SVG, sIFR. Subpixels rendering makes unhinted fonts passable.
  • “Q2 2006: The ‘issue’ resurfaces.”

Web-development community wasn’t really interested in competing standards. Microsoft had introduced some fonts that anyone could download. It wasn’t particularly worthwhile for people to embed other fonts, because ours were kind of the best ones and they worked well. That sort of shocked the font-embedding side in the butt.

Font embedding is including a font with a document file so that someone viewing that document can see the font as the designer intended. The font travels with the document.

During 1998–2006, ClearType and similar technologies meant that regular fonts rendered better onscreen without the expensive hinting that went into Verdana. In 2006, the issue resurfaced. This was really a thread on the W3C font working group list initiated by Håkon Lie, the father of CSS: What can we do about the lack of an embedded font format for the Web? We had discussions. We gave each other’s ideas.

Lie decided that his ideas were better, and in 2006 published this article (“Microsoft’s forgotten monopoly”) and talked about how Microsoft owned the whole font world. The only way to move beyond that was to go with a plan that essentially brings raw standard font files to the Web.

Lie proposed that browsers look for TrueType files outside the local machine. CSS3 can already refer to such fonts, so we don’t need a new standard. He and Bert Bos invented CSS, which brings rich layout and typography to the Web. He’s not associated with the W3C anymore; he’s chief technology officer at Opera. But in the Web community’s mind, he’s associated with the W3C.

Hopefully people can see the obvious problem with just sticking TrueType fonts on the Web. Generally that isn’t something that font vendors allow, especially not an unprotected form. His response was that there are thousands of free fonts available. (Plus TrueType fonts carry information about permissible uses.) “Font designers will find an outlet for their creativity; users will get great content.”

Then last month, another article (at A List Apart), with the rhetoric about Microsoft font monopoly gone. Maybe a bit prematurely, he announced that Web fonts are back. We don’t se any browser support, but through an HTML-to-PDF converter that he’s associated with, these pages can be rendered. He points to Ray Larabie fonts as great fonts for the Web – several hundred in the public domain. These are the perfect fonts for Web pages. (Shows CSS code.)

Fonts are temporarily installed. What could possibly go wrong? But he does see there are issues, mostly æsthetics (“third-rate fonts” by designers who once used blink and animated GIFs). Must convince browser makers to add support for Web fonts. He doesn’t have to convince Opera. But he needs to mobilize Web developers to call on browser makers to add support for this.

I think we can identify some pretty serious issues in this plan.

  • Limits Web design to freeware fonts. Is that better than what you have now? Is that useful? Will commercial fonts be posted illegally?
  • TrueType-centric. (Because they’ll work with old operating systems.)
  • File size is a problem for Asian languages (8–12 MB in some cases). (Perhaps subsetted version, Twardoch interrupts.)
  • Fonts as viruses? (Fonts have been used to attack mobile phones and “even on Windows.”)
  • Browser support: Why no support in Opera after a year and a half? If he hasn’t done it… who knows?
  • The Web-design community: There have been these two articles, the W3C list, but I do not see Web designers lining up behind him with banners saying this is a good idea. On Typophile, I described this as a one-man crusade. It may be more than that. And it might happen, because he does have the clout.

In Boston, Berlow said it would be a great moneymaking opportunity: A client posts a font, and then gets a bill for it. This would open up a whole industry of font-policing. There are tools that make it very easy to rename a font. After a few high-profile cases, would people stop, or just weasel their way around the restrictions?

Microsoft’s response: It was clever of him to initially characterize this as the world against Microsoft.

  • Doesn’t address needs of Web designers.
  • (Through Ascender, they documented the limitations of freeware fonts.) They aren’t wonderful. They aren’t everything you say they are. Small character sets, incorrectly set embedding permissions, hinting errors.
  • Explained that the plan may lead to the misuse of commercial fonts. (That didn’t play well with the W3C. You can post commercial pictures or music; let’s not cripple the Web just because some people misuse it.)
  • Educate the type community, as at TypeCon, Typophile.
  • Then take the Embedded OpenType (EOT) format a W3C standard.

W3C sent that back (scribe did not understand this point), and the results are published. About 4,500 fonts tested. 95% failed in some way. Even beyond English it drops off. I don’t want anyone to take this with anything other than a grain of salt. It was a study we commissioned to show the W3C this wasn’t a plan that would hold water.

Advantage of EOT: Stable. In IE for 10 years. Many font licences do not prohibit EOT embedding. Compression and subsetting are well defined. Supports TrueType, OpenType (CFF only, though not implemented in IE yet). We had Ascender – and they may have worked with someone in the room – and looked at 30 licences.

They’re trying to document the proprietary EOT format (completed). Submit to W3C. Work with Monotype on standardizing the Microtype Express compression used in EOT. Replace the WEFT tool with something form the command line. Make a proof-of-concept, possibly to be released [possibly?].

I’m not here to say it’s all rosy and EOT is the way to go. Not all font vendors allow EOT embedding. Nowhere near immune to somebody hacking out a usable font from it. It’ll be more of a target the wider it’s used. It’s more work for browser makers. It’s an alternative to Lie’s idea, but doesn’t make his ideas go away [as they apparently desire].

(Various Q&A, untranscribed. I did say that CSS 2.1 is not fully supported yet, and all the good designers who are also standardistas would fear that this sort of thing would be used by tacky people who wanted to typeset their entire Web page in Cooper Black. Apparently Chris Wilson of Microsoft considers this an important upgrade for IE7. I’ll say here that if we get font embedding before we get real CSS 2.1 or even CSS 3 support, people are gonna go ballistic. It’s something of a frill, don’t you think?)

The foregoing posting appeared on Joe Clark’s personal Weblog on 2007.09.13 13:21. This presentation was designed for printing and omits components that make sense only onscreen. (If you are seeing this on a screen, then the page stylesheet was not loaded or not loaded properly.) The permanent link is:

(Values you enter are stored and may be published)



First of all, as of March 2019 this personal Weblog is on hiatus for some months.

Other reading

Topics of interest

Typography ⁓ graphic designTTCCanadian EnglishInversion

Archives by date

Just add /year/month/day/ to the end of site’s URL, blog.fawny.org. You can add just /year/month/, or just /year/, if you wish. Years are four-digit, month and day two-digit (with padding zero below 10). For example:

Archives by category

Copyright © 2004–2019

You enjoy fawny.blog