I Love Typography

Web fonts — where are we?

With all the talk about web fonts, I think it’s time I tried to outline the present situation. I’ve not attempted to do so before, owing to the complexity of some of the material, and the speed at which things are moving.

Web designers are generally not interested in technical specifications, TrueType Hinting instructions, and extended OpenType permissions tables. They have one pressing question: when can I use font x in my web pages? Today, in Atlanta, Georgia, at TypeCon 2009, the faithful met to talk about Web Font Embedding: The New State of the Debate. At the foot of this article, I’ve included highlights from the twitter feeds of @typographica (Stephen Coles) and @splorp (Grant Hutchinson). Many thanks to them for the great job they did in reporting.

What web designers want

Web designers want more options, they want more fonts. sIFR, Cufón, and numerous other replacement techniques permit web designers to go beyond the so-called web-safe palette of fonts. However, all those techniques are, fundamentally, hacks. Moreover, their practical use is limited to headlines, or short bursts of text.

What type designers & foundries want

Foundries do not want their raw (.ttf and .otf) fonts uploaded to Web sites where they can easily be downloaded (stolen). @font-face permits linking directly to the raw font file. When I say raw, I mean an uncompressed, unprotected font file, just like you’d find in the font folder on your computer. [see also Stephen’s comment below.]

fonts

Downloading those font files would be as easy as downloading an image. For obvious reasons, foundries don’t want that. In fact, no-one wants that. Here, the music industry comparison doesn’t work. The type industry is in fact, not an industry; it’s not regulated by any kind of governing body, and the industry comprises thousands of small players — the vast majority of type foundries have a staff of one. Font piracy hurts them.

Solutions

Way back in 1997, Microsoft developed its proprietary EOT (Embedded OpenType Format — basically a compact version of OpenType, that permits sub-setting), that only supported Internet Explorer. Hoping for widespread adoption, Microsoft opened it up for all, and in 2007 submitted their EOT proposal to the W3C (for inclusion in CSS3). Later that year, the proposal was rejected, for, among other reasons, security. In 2008, the proposal was resubmitted:

The Embedded Font Format (EOT) was developed by Microsoft to enable OpenType fonts to be linked to web pages for download to render the web page with the font the author desired. This appendix specifies the format of the .EOT file so that User Agents can download, extract and temporarily install fonts of the .EOT file suffix that are included in the @font-face definition of a CSS style sheet. Example pages can be found at the Microsoft Typography site on Font Embedding for the Web.
Downloaded fonts are only temporarily installed on the user’s machines for use by the particular web page while the page is actively being used.

I once heard EOT described as DRM icing on an OpenType cake. Once EOT was associated with DRM (and whether it’s strictly DRM is debatable), then EOT was doomed. For all the technical features of EOT, see the W3C’s Embedded OpenType (EOT) File Format. So what happened to EOT? To cut a very long and complicated story short: it didn’t gain the necessary support from foundries. [I was wrong; see Richard Fink’s comment, & Thomas Phinney’s comment.] Remember, the W3C is not mandated to push these formats through, to run around drumming up support. The consensus must come from the foundries, and from distributors.

.webfont

Recently, two highly respected type designers, Tal Leming & Erik van Blokland (they are programmers too) proposed an alternative to EOT. It’s not proprietary, and its implementation is relatively uncomplicated. Via twitter, H&FJ described the .webfont proposal as:

Smart, compact, open, elegant, forward-thinking, realistic. — source.

Basically the .webfont font is a compressed file (perhaps .zip), comprising two files (the actual font data, plus info.xml). The embedded permissions or meta data are then read by supporting browsers, that could determine whether the font should be downloaded and displayed.

With such huge support from type foundries and many in the type community (even TypeKit supports it), the dot webfont proposal could well be a winner. So, we’ll all be using .webfont by this time next year, right? No. First, the W3C needs to be convinced that the majority of type vendors support the .webfont format. Then, and only then, will its slow wheels begin to turn. Then the browser vendors need to come aboard the .webfont ship, and build support for this new format into their respective browsers. Though the .webfont format is, in my opinion, the best proposed solution, don’t hold your breath. It will be years before we can start to link to .webfont files in our CSS.

If you’re not already confused, then let me introduce you to David Berlow’s (The Font Bureau) Permissions Table for OpenType proposal. (Technical specification here). Without getting too technical, I think Berlow’s proposal can be summed up thus: embed ‘meta data’ in the OpenType font file. These data will be information about the permissions for which the font is licensed. For example, the permissions table (not separate from the font file, but embedded) would include information about permitted use; e.g. whether it can be used on a web site — previewable for web.

The proposal does not require any change in font format; it only requires that more data (about permissions) be stored in the font file. Some have pointed out that its greatest strength — XML to describe the permissions — is also its greatest weakness. What’s to stop users from opening font files and editing the permissions? Another of its obvious strengths is that it does not require any kind of wrapper, and can be used with @font-face, which will soon be supported by most, if not all browsers.

In the meantime

While we’re waiting on .webfont et al., there’s Typekit, a simple solution that involves web-only font linking licenses. Basically, a font file, or a subset of the font is stored on a third-party server.

typekit-customise

You pay a subscription to Typekit to rent (not buy) the font. The rest is simple enough. Include a call to a JavaScript file (that handles delivery of the font, I guess), and simply include your ‘subscription font’ in your fontstack, like:

#introduction .one p {
font-family:"skolar-1","skolar-2","Palatino","Georgia","Times","serif";
}

Great to see David Březina’s Skolar on screen. Go to for a beautiful web to see Typekit in action. Typekit is still in beta, but it looks very promising.

beautiful-web-detail

One of the most exciting aspects of the Typekit solution is best described by Thomas Phinney:

…the most interesting thing about Typekit & Kernest is they provide a service, a subscription, a brand new model for font licensing.

Multiple jars of jelly

We need consensus. They only way a consensus can be reached is through compromise. There exists no governing body of type, so there can be no democratic vote. The closest thing we have to consensus is the list of foundries that support the present .webfont proposals.

Despite concerns about the security of the .webfont format, most of the larger and important foundries have come out in favour of the .webfont proposal; and that’s what really matters. See @typegirl’s Most of the important foundries are supporting #webfont list.

If no consensus is reached then .webfont will forever remain a proposal. If there is consensus, then perhaps at the very soonest we’re looking at .webfont in our browsers by 2011-2012 at the earliest. @splorp sums it nicely in <140:

We just need to have one #webfont initiative to start solidifying. That’ll help. Right now, we’re tip-toeing around multiple jars of jelly.

Regardless of which format or proposal actually wins the fight, type designers are going to be very busy indeed. Most fonts are not optimised for on-screen viewing, so, if they are to compete with those that already are (e.g. Verdana), then they have lots of work ahead of them. (Type Designers have the joyous prospect of mastering TrueType hinting instructions).

Final thoughts

In my opinion, EOT is as good as dead. [Cf. Tiffany’s comment below; and Thomas Phinney’s.]

EOT may be dead, but Ascender Corporation is proposing EOT Lite — think of it as a less restrictive implementation of the original EOT. In what way is it less restrictive? Well, the new EOT Lite does away with URL binding (limiting use to a specific domain or URL), and proprietary compression technology (MTX compression) — the two principal objections to the original EOT specification. Ascender hopes to have it up and running within months. [added July 21, 2009].

Will .webfont ever come to our browsers? Who knows. But with the backing of the majority of influential type foundries, it could. In the meantime, TypeKit appears to be a viable, workable solution. And Typekit is now. I know I’ve omitted mention of other proposals like EOT Lite or Kernest from Ascender Corp., etc., but this article is intended as a non-technical, brief [laughs] overview. If you have questions or comments, then please leave them below.

[Update (July 21): fontdeck joins the fray.]

_______________
Highlights from TypeCon 2009’s Web Font Embedding panel discussion, courtsy of @splorp and @typographica


Eleven font nerds and a microphone. @typographica is live tweeting from the huge #webfont panel discussion this morning.
Audience unrest already. Only one web designer on a panel of 11. Via @nicewebtype — @splorp
Using a service like @typekit or @kernest is similar to buying/selling faces through distributors like @myfonts.… except that you’re not renting the typeface via a modified license. It’s not yours to use perpetually. — @splorp
@opentype They don’t want fonts that users currently have to be used as web fonts. Should be a separate license and product. — @typographica
“… the thinness of the wrapper is disturbing …” — John Hudson on the .webfont format. — @splorp
“Any new font format we come up with … takes years to be implemented.” — Bill Davis, Ascender — @splorp
Dewitt: Foundries, if you don’t have a license that addresses the web, do it. Have a position that allows for some of these things. — @typographica
Bill Davis — “I think we’re on the cusp of something happening very quickly here …” That’s because web designers are frustrated. — @splorp
There’s a big ostentatious EOT Lite petition at the door. Presumably placed there by Bill Davis. Waiting to hear his invitation to sign it. — @typographica
John Hudson of Tiro Typeworks is explaining @typesupply’s .webfont format. “… super easy to implement … ” — @splorp
Mason: The W3C doesn’t lead, it follows. It takes steps once groups find consensus. — @typographica
Hudson explains the problem with a new format: IE is slooooow. IE users are slooooow to update. (Could take years for .webfont to be real.) — @typographica
Bill Davis — “I think we’re on the cusp of something happening very quickly here …” That’s because web designers are frustrated. — @typographica
@_Baylink I don’t think that’s the case. The foundries aren’t terrified, they’re cautious. None of these solutions will ‘eliminate’ misuse. — @splorp
Thomas Phinney: URL binding was a non-starter because vendors don’t want to enforce that. Worse, users might be open to DMCA liability. — [see Thomas Phinney’s comment for clarification.]
Basic Q. What happens when any of these schemes don’t work? A. You’ll get the fallback font. Whatever specified in CSS, what you see today. — @typographica
Verdana is still a pretty good typeface. — John Hudson — @splorp
Gabrowitsch (F[ont]F[ont]): We support .webfont and “EOT Lite, Medium, or whatever”. As long as there’s no chance to use existing fonts on the web. — @typographica
Relevant commentary from Tim Brown @nicewebtype: Type sellers, web fonts, and Typekit. link — @typographica
Typekit is willing and excited to incorporate .webfont proposal in their product. — @typographica
Final question from moderator: how urgent is it?
A. Garrick Van Buren (Kernest): We’re 10 years behind.— @typographica
Hudson: You can say yes to .webfont or EOT. It’s not an either or situation. But foundry weight behind one format will influence browsers.
End. Applause. Matthew Carter first to stand. Maybe wants out of here. — @typographica
Looks like everyone is walking out without batting an eye at the giant petition. Not sure the pen has been touched.
— @typographica

_______________

Further reading:
@font-face in IE: Making Web Fonts Work
Web fonts now (how we’re doing with that)
Web font licensing: the basic idea
Type sellers, web fonts, and Typekit
List of foundries supporting .webfont
Typophile
@typegirl @splorp @typographica @nicewebtype
@font-face in action
Jeffrey Zeldman Questions The ‘EOT Lite’ Web Font Format
Audio from the Web Fonts Panel at TypeCon2009


Tags:                                                       

  1. Thanks for this reasonably exhaustive round-up.

    I’m hoping, though, that it will be obsolete in 6 months.

  2. Thanks for the brief overview!!!!

  3. Great summary, cleared up some questions I had from reading all the random tweets.

  4. Nice summary. It is funny that this is a *brief* overview, but the effort is still applauded! Should be quite understandable, even for those new to the issue.

    I have to say, I am not sure about EOT being dead already… yes, hopefully it will disappear/be firmly replaced by a superior alternative, but (for the time being) it is a very viable option to bring more fonts to the web. Unfortunately, I imagine we will be serving up several formats and delivery methods for the foreseeable future – until the webfont paths narrow a bit more.

    Cheers,
    Rob

  5. I think you’ve managed to sum up the state of affairs with typography on the web very well. I’ve been reading tweets and pages about this subject all afternoon and this is by far the most comprehensive article I have come across! Thank you!

  6. zork

    I’m sorry but DRM just isn’t going to work any more than it did for the music industry and that does not have anything to do with how large the foundries are compared to the recording industry players.

    DRM is just easily defeated snake that won’t really accomplish anything at all and it isn’t like the internet is totally free of unauthorised copies of fonts right now, is it?

    I seriously doubt the web is going to adopt any form of DRM in a serious way for something as fundamental as this. What you’re going to have though, is @font-face and foundries are free to let people take advantage of it at a cost or perish.

    Read this for an alternate take on the issue:
    http://diveintomark.org/archives/2009/04/21/fuck-the-foundries

  7. Great article, thanks for the overview. With both the webfont discussion and the XHTML2/HTML5 discussion taking place simultaneously, it’s challenging for us web designers to stay on top of things right now.

    However, there’s one thing I’m still confused about: how would the .webfont file format deter font piracy? From what I understand the embedded permissions meta-data would prevent someone from downloading a .webfont from my site and reusing it on their own. Is there something about the .webfont file format that makes it more difficult to hack these permissions?

  8. One of the most exciting and interesting aspects of the TypeKit solution is also the one that makes it unreliable. I don’t see myself using or supporting anything involving a third party. Microsoft had the basic idea right ages ago, which one should expect from a company full of developers.

    “What’s to stop users from opening font files and editing the permissions?” A large percentage of users will edit the permissions regardless of the implementation. The software industry has a poor track record of producing unhackable files.

  9. “You pay a subscription to TypeKit to rent (not buy) the font.”

    What happens to the website when the rent payments stop?

    Whatever solution is adopted it needs to be something that will have longevity and preserve properly in archives.

  10. TypeKit may be a short-term solution for nervous foundries, but long-term I’m nervous about consolidating too much power with a single provider.

    BTW, I always find the piracy issue a little specious; the major difference between the media and font industries is one of demand - media is massively attractive to a large audience, fonts only to a very niche one.

  11. taris

    “uploaded to Web sites where they can easily be downloaded (stolen)”

    You mean where they can be copied and the copy than used in a copyright infringing way.

    Not one byte on my server ever got stolen. Bytes, as opposed to matter, are unstealable. If you describe things simpler than they are, you make them wrong.

    And besides: DRM will work someday. I think arround 1984 the technology needed for DRM will be used comprehensively enough. Than we will all be happy - Or at least nobody will be able to use a computer font to write something else.

  12. You should reconsider the comparison with the music industry.

    Just because the industry isn’t lead by multi-national conglomerates doesn’t mean it isn’t an industry. The big foundries are very much like record labels, not in that they have thousands of employees, or earn billions every year, but in that they have economic & cultural influence in the market.

    But there are little guys in the music industry, too, that are just as hurt as the big guys by piracy. At least, those who don’t adapt to changing conditions.

    DRM didn’t work for the music industry, because people understand a model of ownership that’s continually reinforced – unless stated otherwise up front, when i’m given something, it’s mine. And people understand, at this point, that digital duplication of a file (whether that’s .mp3 or .otf) is zero cost. We’ll all acknowledge that it took work to make the original, and most designers, I expect will still assign some of the value that they would to the original work, to the no-cost duplicate, but to consider the duplicate to be equal in value to the original doesn’t work.

    So our second solution: renting typefaces. Typekit seems like a good solution for renting, if that’s what you’re looking for. It’s a good way to retain control . If you can convince people to rent, you’re all set. I bet to do that, cost per rental is gonna have to be pretty low, but maybe that’s the plan.

    Really, I think the foundries should do some tests with just allowing @font-face. Release a font or two, that costs what it would cost anyway, but the EULA allows for embedding. The audience they’re selling fonts to is an honest bunch of people who appreciate what it takes to make a good font. I think they’ll be impressed by how respectful we are, and how supportive we’ll be.

  13. Dave Crossland

    Kernest isn’t from Ascender, its from a single individual, and it currently carrys only freeware and free-software fonts.

  14. Stephen Coles

    Nice summary, John. Though I do think it’s important to include EOT Lite. .webfont is clearly the foundries’ favorite long term option, but if they want a non-raw format within the next couple of years they really need to come together and back some variation of EOT. There’s no other option out there beyond the web services like Typekit.

    Correction in your Twitter summary: Ivo is with FontFont.

  15. Stephen Coles

    Foundries do not want their raw (.ttf and .otf) fonts uploaded to Web sites where they can easily be downloaded

    They also don’t want customers who have licensed fonts for desktop use to consciously or unconsciously break their license by using them on the web. The idea is that a unique web font format to make licensing more straightforward.

  16. johno

    Rob
    I think if it hadn’t been for TypeKit, then EOT (in the form that Ascender Copr. is using it), would still have some life in it. Personally, I think that TypeKit has killed off any chance of an EOT Second Coming. Having said that, I think it’s commendable that Ascender Corp. is doing something, and doing it now.

    zork
    I think you could be right. I think that any proposal that comes to be associated with DRM will, in the long run, fail. What do you think of TypeKit?

    Kyle

    How would the .webfont file format deter font piracy?

    I don’t think it will. It’s another layer of protection, so in that way it’s a deterrent, but not much one.

    Peter Gasston

    I’m nervous about consolidating too much power with a single provider.

    That’s understandable, but I do think we will see many more following in TypeKit’s footsteps. The competition will drive down prices; though I’m not sure what else it will do. If TypeKit gets it right (and the foundries feel they are ‘fairly’ compensated, then I think that perhaps it has the potential to be more than an interim solution.

    taris
    Perhaps ‘stolen’ is not the best choice of words.

    Not one byte on my server ever got stolen.

    But bytes could be take without your permission. If I use your copyrighted content without permission, then what am I doing? Technically, I may not be stealing, but I’m using it without your consent. Perhaps it’s simply ‘copyright infringement’ — if it is not stealing, then it is still wrong.

    I’m sorry, but I don’t understand anything you wrote in your final paragraph.

    Micah
    You make some sound points. I shall rethink. Thanks for your contribution.

  17. Stephen
    You’re right about EOT Lite. After I get a little sleep, I’ll append an overview.
    I was tempted to make the EULA connection (it’s very important), but ran out steam and space.

    Thanks for the correction. Ivo might be surprised to learn that he’s now working for FontForge :) Off to give him his old job back.

    By the way, how many names on that ‘ostentatious petition’?

  18. Well if these delivery method catches on I have a feeling web developers will have EOT format for IE8+ and everything else will use .webfont because it would be too easy if every browser rendered web pages in the same way.

  19. When I left there were 6 names on the EOT Lite petition. But I don’t think it’s that important. We already know from the panel that all the major foundries are open to any format that isn’t raw.

    > Personally, I think that TypeKit has killed off any chance of an EOT Second Coming.

    Typekit is really impressive and I’m confident it will be successful but it won’t be the only way to deliver type on the web. There will be fonts that are unavailable on the service and sites that won’t use a third-party service.

    (Written while sitting just a few rows behind Typekit’s Bryan Mason and Ryan Corver on a flight from Atlanta to SF. You really notice turbulence when you’re typing on an iPhone …)

  20. dave

    Sorry, requiring some wacky new drm system is the opposite direction from where everybody else is going.

    And for your music industry comparison, it’s just a matter of scale. There are a few big foundries/music labels, and then smaller ones and then orders of magnitudes more individuals (both musicians and font creators).

    It’s just the font industry is so small and specialized that nobody outside the industry can name anybody in it.

    You could argue for exactly the same wacky drm for other web content from photographers, graphic artists, the news industry.

  21. TypeKit is, as far as I can see, just a very weak layer of DRM. The only protection seems to be a check for the Referer header in the request, which is trivial to forge. I would imagine it just takes a couple of lines of JavaScript and voila, a Firefox plugin (or something similar) to download any custom fonts on the page.

    .webfont is just goofy. Instead of putting metadata inside the font file, where it belongs, they want to put it in a separate file and then zip that with the font file. Architecturally that’s really awkward, and how does it do anything at all to protect the font? I can just look up the URL on the page, download the webfont, unzip it and take out the font file.

    Everyone hates to say the word “DRM”, but that’s what all of these are attempts at, and it just will not fly. DRM is so technically, legally and ethicall nasty that the only groups with the weight to push it into existence are 800lb gorillas like the MPAA and RIAA. Compared to them, the font industry is nothing.

    I love typography and fonts, and I once worked at a type company on font digitization software. The only way forward that I can see is to live with the fact that people are going to make unauthorized downloads of commercial fonts, just as they do with MP3s.

  22. Tiffany

    I don’t think EOT is dead it has just become EOT Lite. And even though there is a publicly available list of foundries who back .webfonts there could also be a list of foundries who support EOT Lite. (It just hasn’t been made yet.) If, for instance, we can get Monotype and Adobe to state their preference for EOT Lite then those two foundries alone (especially Monotype) have a lot of weight to throw around. And nevermind Microsoft’s weight.

    But this aside I think it is going to come down to different things being available and each foundry’s EULA allowing or not allowing each. Sadly I don’t think there will be just one format with all foundries lined up behind it. Why would they? They don’t all stand behind the same EULA at present. They don’t even all agree on allowing PDF embedding.

    What I do hope is that there can be enough consensus made within the foundries to present their hopes to the W3C and get the browsers on board. Without the browsers’ support their worse nightmare could become a reality.

    While it is applaudable that some foundries support @font-face use, maybe at the end of the day they are shooting themselves (and other foundries) in the face. (No pun intended.)

  23. I take this was intended to illustrate to those not in the know. No point at all was given as to why EOT (once open) was bad, how .webfont differs from it, and why it is a better solution.

  24. All DRM systems are doomed to failure by their very nature. The only way DRM can work without third-party involvement is if you give away the keys with the lock, otherwise the recipient has no way to open it! How secure is that?

    Sure, you can rely on keeping the key with an external source (a la TypeKit) but if that source disappears, then you’ve lost access entirely. Just look at the DRM fiascoes that Yahoo, Google and others have dealt with when lack of interest causes the services to become too expensive to maintain. Everyone in the ecosystem suffers.

    If demand just isn’t there and TypeKit disappears, what happens when your clients start calling up because their websites suddenly look different? Or worse yet, differences in widths break the designs…

    DRM just doesn’t solve a damn thing for anyone. Even the RIAA is finally admitting it: http://torrentfreak.com/drm-is-dead-riaa-says-090719/

  25. Tiffany

    Either the foundries get some protection or web designers/developers lose access to the fonts.

    I guess I just don’t understand what is wrong with a little data protection. *throws up hands*

    The foundries* aren’t going to give away any keys. They want to continue holding the keys. I don’t blame them. The keys are the foundries’ to keep. Users may say, “well then we just won’t use there fonts.” Fair enough. But, just like some of us agreeing to some of the more strict EULAs, we’ll continue picking and choosing our fonts based on what is allowed.

    This is good for users — Some foundries will find ways to supply us with fonts for use on the web.

    This is bad for users — Some foundries, maybe the foundries we really want to be able to use, won’t supply us with fonts for use on the web.

    This is the case with foundries already. Some foundries do allow PDF embedding and some don’t. Now there will be another part of the EULA to consider when licensing fonts.

    *They being the foundries who want some protection.

  26. Oliver

    I cannot see a situation where @font-face and TTF doesn’t win. The font foundaries do not have as much control as they think they do, any copy protection is ill-thought-out and inherently insecure, and certain foundaries are going to see a commercial advantage in licensing for @font-face and TTF which is going to put financial pressure on the rest to follow.

    Ultimately, what wins isn’t down to the font foundaries or the W3C (it’s well known that the W3C are without teeth), but rather the browser vendors. And I can’t really see the browser vendors implementing anything other than @font-face and TTF within a reasonable time-frame, leaving it as the only choice and de facto standard.

  27. Sure, any of the foundries can decide to keep the keys, that is entirely their right. More power to them, I wish them luck in surviving past another decade. Meanwhile, the foundries that choose licenses which allow for @font-face embedding will prosper thanks to providing a product that is usable.

    This is precisely the same problem that every other industry is having to deal with, from music to movies to just good old fashioned news. The problem comes down to pure computer science. One’s and zero’s are infallible. They can be reproduced an infinite number of times without any degradation in quality. Furthermore, there is no appreciable associated cost of reproduction.

    Metal type has a clearly definable cost of reproduction. Milling equipment, casting tools, wages, cost of the metal you’re casting the forms out of, etc. But with digital type (and digital anything, for that matter), there is the cost of the electricity to keep the computer up and running. Which is negligible. And it is a number that is constantly decreasing thanks to technology advancing at a breakneck pace. Run the function out to infinity and the limit reaches, you guessed it—zero.

    (Just wait until physical products can be duplicated information technology style…)

    Of course, this does not even touch on the associated cost of -creation- and THAT is the center of the problem. And in reality, I think it is something that needs dealt with as a separate issue. There are two economic factors at play—cost of creation and cost of duplication. However, it is certainly the touchier of the two subjects…

  28. “So what happened to EOT? To cut a very long and complicated story short: it didn’t gain the necessary support from foundries. Remember, the W3C is not mandated to push these formats through, to run around drumming up support. The consensus must come from the foundries, and from distributors.”
    Unfortunately, this statement is totally divorced from the facts. EOT had broad support from font-makers. In addition to Microsoft, both Adobe and Monotype were on board. Just those three have a greater investment in fonts than probably the whole rest of the font making community put together.
    And the idea that the developers at Mozilla and Opera were in any way significantly influenced in their decision by “support” among font makers is simply untrue. As someone who’s followed the issue carefully, it’s almost funny. Do you have any evidence to support this notion?
    Plus, if a permissions table is included in an OTF file, almost all applications and especially browsers will simply ignore it. In fact, as a legal matter, it is wisest for them to do so. So how much value do you think it will provide?
    Incidentally, you did not mention the non-DRM version of EOT that is now, today available. EOT “Lite” as it’s been called is far from dead.
    A good explanation of EOT Lite written in response to a question from Jeffery Zeldman as posted on his blog is here: Jeffrey Zeldman Questions The “EOT Lite” Web Font Format
    I was at the TypeCon Web Fonts conference this morning. It will probably end up on YouTube soon and you can see if the reality matches up with the twitter feed that you’ve relied on.

  29. I think one of the features that are overlooked from the tool to generate EOT files, that somewhat deals with avoiding re-use of downloaded font files, and most importantly, that is somewhat of a must for asian fonts with thousands of glyphs, is subsetting.

    I do not particularly think that subsetting is a thing to be addressed by the format specifications (since a font file containing a subset of the real font is still a normal font). But I guess that, whenever things get down to a somewhat established standard, a need for a tool that allows us to create those font subsets in the standard format (be it .otf, .webfont or whatever) will arise.

  30. johno

    Richard
    Thank you. I’ve now added a link to your EOT Lite article.

    I’d like to hear your thoughts on why EOT, in its original form, failed, and who you think was responsible for its failure.

    Plus, if a permissions table is included in an OTF file, almost all applications and especially browsers will simply ignore it. In fact, as a legal matter, it is wisest for them to do so. So how much value do you think it will provide?

    Very little.

    Am looking forward to seeing that video.

  31. You forgot to mention Typotheque’s web font service system, the only one being developed by a type foundry: Typotheque Web Font Service preview

  32. dave

    Gustavo,

    Typotheque’s web font service system is an example of how stupid foundry’s can be. To even get a quote for how much they will charge you, you need to submit “number of fonts licensed, the terms and length of any agreement, future development requirements”. What, I’ll take the web page down in 3 months? Or redesign it to no longer use the font? And I need to describe how my web site may change in the future?

    What web designer is going to propose this to their client? This amounts to “It costs a variable amount of money to rent part of my design. As your company grows, it’ll cost more, even if nothing on the website changes”.

    Open mouth, insert gun in mouth, pull trigger. Rinse barrel and repeat.

  33. TypeKit does look promising indeed, even if only as a short-term solution until something better is worked out. But one thing to keep in mind that doesn’t get as much mention as maybe it should is that it will only serve fonts from foundries which have an agreement with TypeKit. In other words: if the font you want to use on your website isn’t in the TypeKit collection — which I was told during the NYC demo last month would be more on the “boutique” size/scope — you’re out of luck.

    Another thing Mark Simonson pointed out to me regarding the much-cited .webfont support list — which his name is on — is that a foundry’s support of .webfont doesn’t necessarily mean they think it’s a perfect solution or that they oppose other competing models; only that they generally support the ideas behind it.

  34. @dave: Time wil tell.

    Good luck with OFLB & Cantarell.

  35. taris

    Thanks for the Answer.

    “But bytes could be take without your permission.”

    Again, you mean copied. :-)

    “If it is not stealing, then it is still wrong.”

    If wrong means against your law. Yes.

    “I’m sorry, but I don’t understand anything you wrote in your final paragraph.”

    I meant that there is only one way to make DRM work - The radical way of killing computers as a machine you can you use like you want to.

    Full govermental/industrial controll about anything that happens on the internet and every machine connected to it. Only certificated software will work.

    As you may have seen by Iran, that may not be a good choise.

    Sadly, I am sure it will come anyway, because so many people just want to make there living, and don’t see what comes packed with the total IT-security wich they need for enforcing their copyrights. Music industry first, than font designers, book authors and so on. Nobody of them want anything bad, they don’t even understand what they have to do with the political/IT changes we are facing.

    Like you.

    http://www.fourmilab.ch/documents/digital-imprimatur/

  36. Bert Vanderveen

    I like the approach taken by Typotheque: they generate a specific typefile that stays on their server.
    http://www.typotheque.com/news/web_font_service_preview
    IMO: foundry happy, designer happy, end-user happy.

  37. (First, I would say I’m sorry for the poor english you’ll read… but I’m french…)

    I just want to say that TypeKit is really easy to hack. It only takes 2 minutes to get the OTF file (and I’m not an expert).

    I think Typotheque Web Font Service is a better solution ( see this news: http://www.typotheque.com/news/web_font_service_preview ). It uses a secure connexion to load the font.
    Even if it’s possible to get the TTF file, it’s more complicated and it’s not really accessible for most people. I have to ask a friend of mine ( who is a computer programmer ) to hack this security.
    He tells me that if you see a font on your browser, so you can get it.

    I think there’s no miraculous solution ( even a font in a pdf can be retrieved ) but we have to choose the least worst option.

  38. Kerim Friedman
    Peter Gasston
    “What happens to the website when the rent payments stop?”

    Precisely. And what happens when the bandwidth requirements surpass the capacity? Even Google occasionally has difficulties delivering content on time.

    micah rich
    “Really, I think the foundries should do some tests with just allowing @font-face.”

    When all of the browsers support it, @font-face will happen. With or without some of the foundries. Security through obscurity won’t prevent that.

    Jens Alfke
    “The only way forward that I can see is to live with the fact that people are going to make unauthorized downloads of commercial fonts.”

    The sooner everyone comes to terms with this the better. This discussion is beginning to read like problems are being solved with little understanding of the implications those solutions impose on functionality of the web in general.

  39. Philip Taylor

    elmimmo: I have a tool at http://fonts.philip.html5.org/ that does subsetting of OpenType files, and the code is open source. It tries to do it in a high-quality way, e.g. automatically preserving ligatures. I’d be interested if people want to use or extend or adapt the code.

  40. Regarding EOT Lite vs .webfonts they are very similar options. The advantage of EOT is that it is currently supported, whereas .webfonts is not yet implemented. In the end though they are just slightly different, unprotected, formats that are specifically made to be used online. They will still be illegally copied (or stolen, no need to split hairs) but they should make usage rights more clear for the users. I am supporting .webfonts, but I would also support EOT Lite (providing Ascender will make a nice compiler).

    All current web font rendering options just vary how far people need to go to get the fonts – and no paths are ever very long. Very unfortunately, these font serving systems (typekit & typotheque at least) have already been hacked – and they aren’t even officially open yet. This does not bode well for the confidence of foundries.

    At the end of the day we need to do what is best for our customers and try not to worry about the piracy too much. We have to accept the fact that there will always be those who do not respect EULAs or IP and there is really nothing we can do about it. My hope is that being a customer-orientated foundry will at least be appreciated by the real customers.

  41. and some more questions……..

    who buy fonts?
    why “ordinary people don’t know nothing about typography?
    whyordinary people don’t buy fonts
    why designers use fonts without permission?
    why font designers are like a “golf club?
    the use of fonts in design can help font designers to become more popular and sale more if “ordinary people get the posibility to see those beautiful fonts?
    what do you think if you design a site “visitors don’t steal graphic elements, css…………?
    why font designers afraid to jump in the deep water?

  42. @dave That is not accurate.

    You are referring to our OEM licensing, which is different from the upcoming Typotheque Web Font Service. We hope to make Web Font Service as easy as it is possible in terms of licensing. There will be two licenses: Full license that covers print (current EULA) and web use, and Web only license. Web only license will cost a fraction of our existing font license, but users will pay for bandwidth. You can monitor the traffic directly online, so there won’t be any surprises

  43. Rob

    Precisely. Great example is Jos Buivenga / exljbris (http://exljbris.nl). Not only is he creating some of the most well respected new free fonts, using the free weights and styles to to drive sales of others—he’s openly encouraging the use of @font-face!

    I know as soon as his Calluna and Calluna Sans are released, I intend to be first in line at the checkout stand!

  44. First of all, John thank you for the summary! It was much needed. I find it actually very useful that it did spark some disagreement. It helped me to revise some questions/answers and understand situation I have not been following.

    [1] I think we (type designers) maintain an old illusion about how our market works. People do not buy fonts because they could not get them for free anywhere else. They buy them out of sympathy, understanding the value of our work and/or legal reasons. They could get them for free, easier, and faster (!). It is not that we would be shooting in our faces. It is more like we have been shot already. We already accepted the piracy as a burden of our business.

    If I am right in this view, we do not need any kind of DRM. The expected “new” web piracy won’t change a thing. I would very much like to see some study or educated estimate re this view. Or at least an authoritative opinion. It is crucial information for designers in order to evaluate the formats properly. Otherwise, they are just left aiming for the most security.

    [2] What we, however, want is a tool to limit webfont licences exclusively for web. We want to make a profit out of this #webrisk and keep distinction between web and print fonts. Why? If I am not sure whether opening my fonts for web use is going to make me money I would rather keep the new market separated from the old working one. That is the motivation behind the web-specific format. Acceptance of non-security, but limited to web.

    Personally, I think that opening to web market is surely going to make a profit. An objectively, we are not going to have strictly (that is: not-convertible) webspecific format ever. Not with current technologies where the fonts are described with curves. The only option I see is bitmap fonts &c.

    .webfonts is just bundled metadata with print font (we can have them in OT table as Berlow suggest, why another format simple-to-hack when most are not going to care?), EOT Lite is a very thin wrapper as far as I understood, but at least not so trivial. It will become easy convertible (assumption), but at least something. Typekit and similar tools offer only limited security by obfuscation. So far too easy to circumvent. These techniques are not imo worth complicating life of paying customers. Even though the interface is sexy, it is still another interface.

    Therefore: prepare the fonts for web (have them subsetted, add web exclusive license, permission tables, …) and go naked! Or if you are shy, have EOT Lite.

    [Please note that this is still an opinion under development, was and will be revised, and it is not an opinion of TypeTogether.]

  45. Sidenote about obfuscation methods. Some of them impair the fonts’ kerning and OT features. Not a good idea. My concern is purely of egoistic designer. With my limited abilities I tried to produce as good font as possible and I don’t want crippled copies of my font ripped off from various webservices floating around. I want constant quality of my work world-wide, nothing smaller, if possible.

    The very first person who paid a compliment to my typeface on web was the one who posted it on a Russian download server (being it ripped from my MATD PDF specimen). Just thought it illustrates the situation pretty well. :-)

    [one more note, what I am posting here is a result of a weekend emailing with friends and colleagues who helped me (in)form my views]

  46. (Re: @dave “Good luck with OFLB & Cantarell” – I thought ‘dave’ was Dave Crossland. Sorry for the confusion.)

  47. I think this post is very interesting but it has problem with its solutions..!!
    The first point, if you want to deal with browsers you must have free fonts for many reasons.. some of them :
    1- not anybody can pay to buy fonts.
    2- The people like open source in everything ( Google like and support that)

    The other point I think this is not right when you say “the W3C is not mandated to push these formats through, to run around drumming up support. The consensus must come from the foundries, and from distributors.”

    The W3C is the responsible and it is put rules and standers ..!! can you tell me how W3C is not mandated to push these formats through, to run around drumming up support.??!!

    thanks

  48. Thanks for the comprehensive article (I’m doing my own overview, this is part of the research…), it helps in shedding light on the broader spectrum of web fonts.

    Although I haven’t tried Typekit yet and for that reason shouldn’t voice an opinion, I can’t really see myself renting fonts for use on web sites. At least my own sites. Partly because I’d prefer to use the same fonts, if possible, on more items than “just” the site, e.g. print stuff, PDFs, excetera. Renting a license for the web site alone isn’t really a feasible option for me.

    In the end, although Typekit, Kernest .webfont and @font-face all offer interesting possibilities, I still think that well constructed font rules in the normal style sheet sense is the best way to go. Even with these new (and “new”) techniques and services, I doubt that we can expect *every* browser to handle esoteric fonts in identical fashion anyway (see Malarkey’s browser comparison on forabeautifulweb.com) … (and I didn’t even mention IE…)

    But it’s an interesting topic, no doubt. ;)

  49. Torbjørn Vik Lunde

    I’m still in favor of pure OpenType (maybe with the addition of permission tables). I think .webfont seems like a better solution than EOT though.

    I still don’t see why type designers deserver or need protection(DRM) more than musicans, photographers, writers, etc. In addition: when has DRM ever really worked? When has it ever done anything more than hurt the loyal honest consumers?

    Even if the DRM *did* work I don’t think it would change anything. Just do a search for “fonts”, or a famous font name on The Pirate Bay and you see what I mean.

    Typography on the web is a untapped market and I think the first foundry to embrace it will do well.

    I know I’ll pay for a non-stress way to get more choices in type.

    Also, when are we going to get fonts optimized for the web? Fonts that look good on Windows, and not only OS X. Fonts that compete with Georgia and Verdana.

  50. Matt

    Just a wild guess, but it wouldn’t surprise me to learn that EOT got shot down for no other reason than it was a Microsoft initiative. Standards organizations have a very nasty habit of automatically writing off anything that comes out of Redmond, no matter how good it is. This lets them continue to feel morally superior because they refuse to allow the 800 pound gorilla to dictate the direction of their industry (even if that direction is a lot better than the one they wind up taking), but it generally isn’t very good for anyone else.

  51. Sidene Damayún

    I think the whole royaltie-driven business is doomed. No more “I make stuff once and then I expect to be paid every time others use/listen/see my work”.

    That’s because a copy is not the same value of the original. The first one took expertise and time to produce. The digital copy, being exactly the same, not. And we can obtain the copy easily, for free.

    You can try to force DRM, laws, whatever: your business model is doomed. Change it or disappear. That’s how capitalism works. Face it, evolve or die.

    All other considerations are bullshit that time will prove wrong: I will not pay for something that I can obtain for free, be it a font, movie, or song. Because right now, if I pay for it is for pity, for “doing it right” or because “I like to support the artist”. No serious business can be sustained by the good will of the customers.

  52. Bert Vanderveen

    Hee Sidene, how about giving us your home address? So that we can all come by and help ourselves to your booze and the rest of the stuff in your fridge? Or do you brew your own beer?

  53. David Crossland - thank you for the clarification. The fonts currently within Kernest.com have licenses supporting @font-face use - the Kernest.com is built to support for-fee fonts as well.

    Fred K - Kernest and Typekit are services built atop CSS’s @font-face declaration. Also, not every browser supports @font-face today, just as not every browser supports sIFR, cufon, or even images. There will always be a need to specify a back-up font in the CSS, even if it’s simply ‘monospaced’.

  54. Torbjørn Vik Lunde: I still don’t see why type designers deserver or need protection(DRM) more than musicans, photographers, writers, etc.

    I think that this is a particularly salient point. Copyrighted data is sent over the wire unprotected in many formats. While I certainly understand the problems with the model, there’s going to be some kind of change. Whether this involves proprietary foundries being shut out of the web or a system like EOT where it’s easy to circumvent, but the user has to do something by hand, is yet to be seen.

    As for me, if I have a strong need to use nonstandard fonts, I’m just going to stick with free ones, and leave all of this craziness to the rest of you folks.

  55. Oliver

    Otba Mushaweh: “The W3C is the responsible and it is put rules and standers ..!! can you tell me how W3C is not mandated to push these formats through, to run around drumming up support.??!!”

    That’s not the W3C’s job. It makes recommendations in the hope that they reflect community consensus and that they will be adopted. Ultimately, however, it is in no position to mandate or even pressure browser-makers what to implement, not least because it has no way of enforcing compliance.

    In the real world, what tends to happen is that browser vendors tend to pick and choose what to implement pretty much arbitrarily based on some combination of usefulness, ease, what competitors are doing, whether somebody is interested in implementing it, whether it agrees with their principles, and importance. This approach means that standards can be implemented in months, years, or even never.

  56. Oli

    I got half-way through writing a comment this morning, but @David Březina and @Torbjørn Vik Lunde have said it already. However that’s not gonna stop me! :) Background: I’ve implemented @font-face with subset fonts including .eot (thanks Philip!), bought fonts, and have been waiting a month so far for a foundry to reply to post-purchase questions on whether their license allows subset @font-face use.

    New format proposals are a nice idea, but are five years too late. Even if .webfont was universally adopted and *implemented* in all major browsers tomorrow (HAR!) we’d still need to wait until 2014 for widespread adoption.

    @font-face works now, but the hoops required to generate .eot files make cross-browser support painful, and the lack of licensed professional fonts doesn’t help at all. Also the chance of all browsers supporting one method in the near future are nil (and anyway, ref: 2014).

    Font foundries seem to fear a new world where people download fonts and use them *without paying!1!* except, um, that’s already happening yo. What font creators don’t seem to be thinking about are their customers. The people who do buy fonts would like to:

    * not have to wade through pages of legalese to work out what is permitted
    * not deal with DRM, Byzantine licensing and other bs that makes paying to use fonts far more onerous than if they were stolen
    * pay for and use fonts on the web already!
    * be treated with a little empathy

    Font creators who look at this from their customers’ perspective and make it easy for people to buy and use fonts (whether via Typekit, another system or just intelligent licensing) are going to make money. Anyone who doesn’t is going to be leaving cash on the table, pissing off customers, or both. Of course, that’s a valid choice. The current situation denies paying customers the opportunity to give font creators money.

    For anyone who thinks DRM is a necessity (rather than a fundamentally flawed idea that makes paying customers suffer), subsetting alone will make the fonts ‘broken’ and more trouble than they’re worth (ie easier to just torrent for anyone who is too cheap to pay). What we’d really like are CMS plugins for autogenerating subsets that are hosted on the customer’s server, and font creators such as Jos Buivenga empathetic enough to license fonts allowing this, which would work fine with a subsetting requirement.

    Phew. Thank you, I feel better now :)

  57. Outside of designers and type admirers, I’m fairly sure that type usage, thought and piracy do not exist. I mean is anyone’s mom going to take the time to look at source code she doesn’t know how to read and download a copy of Helvetica Neue 45 Light if it’s using plain old @font-face. The answer is no.

    It’s not to say I don’t have sympathy or consider the possible losses type designers and providers endure. Probably like every other student on the planet back in the day, I had a completely bootlegged set of essential typefaces that I used on projects. If you were to put the retail value of all this together, it would probably at least equal a years tuition at any school. We had to do this otherwise we’d walk into a job interview with Geneva and Courier all over our portfolios.

    The issue of piracy is a bit overblown. Very few people are going to spend any amount of time trying to steal typefaces. Even designers. Type costs can be built into a project proposal, just as anything else can. So if you really need Knockout, you just expense it and afterwards it’s yours to keep.

    As far as web usage goes, I don’t think the foundries should consider this a retirement fund, as focusing on the business is not what they do best. Many are successful already. Monetarily, if they were adamantly monitoring usage, they could probably all be driving Aston Martin’s and living in tropical locations, but it’s just not the case, so why is it now?

    In terms of service and profit to owners, i think what Kernest and TypeKit are doing are really the best options. As for security I cannot speak to any of that (especially @webfont - don’t understand it at all). Why don’t the large foundries post their bread and butter faces through these folks and see how licensing and usage goes. I mean….is it really that big of a risk?

    Seems a whole lot of fuss over nothing at all.

  58. After reading the article and all the comments it’s apparent that unfortunately we still have no viable solutions to using non web-safe fonts, each one has its CONs & PROs.

    At least, web designers today can freshen up the Web with @font-face using fonts designed by Jos Buivenga.

  59. @all
    Having met at length with Garrick Van Buren of Kernest, Bryan Mason and Ryan Carver of TypeKit, I can tell you these guys are web professionals who have a very clear idea of what the market is going to accept and what it won’t.
    I too, have reservations about having a third party in the middle between my sites and the fonts.
    They are perfectly aware that this is of concern and they certainly don’t want to lose users of their services because of it.
    We’ll see how this concern is addressed by each of these new services.
    And hey, there’s no reason why you can’t tell a font-maker that you’re not comfortable doing business this way and is there an alternative?
    I don’t think Kernest or TypeKit is insisting on exclusivity. It’s a good question and I’ll be asking and writing about it.
    We’re in the very early days of this. These guys have so much work to do, I’m getting tired just thinking about it.

  60. Tiffany

    Note that Jos (exljbris) only allows @font-face embedding in his free fonts.

    What David (Březina) said deserves repeating. One of the larger issues for foundries, from what I gather, is they don’t want the fonts they provide for web use to be the same those provided for desktop fonts. Most foundries want division.

    Also, when are we going to get fonts optimized for the web? Fonts that look good on Windows, and not only OS X. Fonts that compete with Georgia and Verdana.

    And it is kinda laughable to think that you can expect high quality fonts without some compensation to the foundries. These web fonts were the result of hugely funded projects. I’d imagine if you wanted to create another one of those projects you *could* get more fonts just like that.

  61. Tiffany

    Note that the paragraph starting “Also, when…” is me quoting someone from above.

  62. Tiffany is absolutely right. Only my free fonts can be used for @font-face at this moment. I would like that to change in the not too distant future. Be sure that I’ll keep track of things.

  63. All
    Thank you for your contributions. Still digesting.

    Richard Fink
    You’re welcome to publish that here, if you wish.

  64. Great round up, good to have a concise post on the topic.

    I’m hopeful for .webfont, it may be early days but it looks as though it already has a lot of support from both the design/web community & foundries.

    I haven’t had a chance to read into Typekit massively but I was thinking it would be nice to see a hierarchical pricing structure commercial/non-commercial/student usage (or traffic based), thoughts?

  65. Just saw this today! I’m stoked that someone out their is working on a solution to make both type foundries and web designers happy. Until .webfont, this may be the way to go… Any more beta invites left?

  66. it would be nice to see a hierarchical pricing structure commercial/non-commercial/student usage (or traffic based)

    Yes, Typekit will have pricing tiers.

  67. Xiao

    You people are all missing the big picture!

    Do you even understand what the majority of users will be for this technology?

    It is the average people, who will NOT pay for professional fonts no matter how professional they are.

    They will use free, “crappy” fonts.

    That’ll be 99% of the users.

    Heck, even if there are “pirate” sites out there offering ripped off professional fonts, and even when there are GOOD free fonts, they would still use ugly free fonts because they “look better”.

    Most people don’t need professional fonts, they just need DIFFERENT fonts!

    Also, what about people making their own fonts from their own, or their child’s handwriting? When the people who make the font are the same ones that implements them, or when the fonts they used are free fonts why should the foundries have the ability to get in-between and have a say in those matters?

    You can’t let the 1% dictate what hoops the other 99% has to jump through in order to make sure their (outdated) business model succeed.

    Just because your fonts cost money, that doesn’t mean you will hinder the development of the web to implement some complicated, frustrating scheme to ensure the profits of your 1% marketshare.

    It’ll be difficult to track down and enforce your right on the font files, but so it is for music, graphics, and text. You don’t have any right to have it easy. None of us do.

    You’re gonna have to work for your own meals, not forcing the rest of us to work for you.

    It has always falls on the copyright holder, or in this case, since the design themselves aren’t copyrightable, the font file author to enforce their own works.

    This has been for books, music, images, and video.

    Just like fonts, most of the professional written works are copyrighted, but there’s not DRM on text, you can’t force the web to accept a standard that enforce meta-data on a paragraph of text, and block all text unless they are proven copyright free.

    That’s because even though most of the high quality literary works are copyrighted, what we read most are from someone’s blog, forums, public discussions, that might not win any major prize but is what everyday life is about!

    You can’t DRM life.

  68. Xiao - I’m not sure what you’re going on about. Anyone can make a font that will work with @font-face right now. The foundries are looking to manage their own products, this has no impact on license-free fonts.

  69. Tiffany

    I don’t think anyone would suggest that the foundries do not deserve compensation. But, that is precisely the problem every single industry is having. As I state in a previous post, DRM is inherently flawed and, from a computer science standpoint, can not work. This is why every industry that tries it exhausts their resources and is forced to find a new business model.

    Developers have known this for a long time. The people writing the DRM systems know that it is just a matter of time before someone walks through their code, looks at how the lock works, and releases the keys to the kingdom. The only function DRM has is slowing people down, and thanks to the rabid hatred coders have for such systems, the time it takes for any DRM system to get cracked is also dwindling to zero. Major systems get cracked in days, if not hours. (As noted by Thomas and Rob, TypeKit has already been subverted, the fonts are easily obtainable.)

    If we are to move forward, we’ve gotta stop wasting time on the protected versus non-protected issue. True protection for digital goods does not exist at this point in time. Here’s an analogy for you: Houses in the computer world aren’t built with windows because glass doesn’t exist. With computer science, if you can see in, you can reach through the open hole in the wall and take the product out.

    For any industry dealing with a digital product, the focus must be this: “How do we profit off something which can be copied infinitely at no cost?”

    The prominent answer right now, especially among the pre-digital age stalwarts in any industry, is a walled garden. But then you’re only set until someone throws your content over the wall… (Which, is absolutely illegal, however that doesn’t change the facts of how digital systems function.)

    Better, more sustainable methods involve immediacy (“Gotta have it NOW!”) and providing a unique, custom experience. Now, I’m young in my typography geekdom and I’ve never attempted to digitize a glyph in my life, so unfortunately I can provide no type-specific input on how to achieve such things. Creative typographers will have to do that part.

    But I can note some great examples in other industries. Game publishers are creating special editions that include unique, limited run items. The new Call of Duty: Modern Warfare 2 “Prestige Edition” include functioning night vision goggles. Halo 3 and Gears of War 2 had special editions that included a “Spartan” helmet and “Lancer” assault rifle, respectively. Definitely take a look at Josh Freese (drummer for A Perfect Circle, Devo, formerly for Nine Inch Nails). When purchasing his new album, you’ve got several “tiers” to choose from that include all sorts of unique goods and experiences—signed albums…t-shirts…drums…simply hanging out…Freese’s Volvo 940. Starting at $7 and heading right on up to $20k and $75k packages. The $20k package has sold and included a drum lesson and two songs, written and produced, about the purchaser!

    Digital goods are a new ballgame. I’m sure the scribes weren’t too happy when typesetting started taking hold, this is just another step in the evolution of information delivery. Time to do what creatives do best and—be creative!

  70. Jens Alfke

    Tiffany wrote: “Also, when are we going to get fonts optimized for the web? Fonts that look good on Windows, and not only OS X. Fonts that compete with Georgia and Verdana.”

    Maybe when Microsoft improves its font renderer?
    Seriously, MS’s renderer is much more dependent on hinting than Apple’s. And hinting is probably THE most difficult part of digitizing a font. Since hinting’s also unnecessary for print, it’s understandable why most font foundries ignore it or don’t put much work into it.

    “These web fonts were the result of hugely funded projects.”

    Again, much of that expense was because they’re so extensively hinted. I don’t see any way that this kind of effort would happen again unless, as before, development is subsidized by a large company and the fonts made available for free. Because anyone wanting to recoup a large investment in a font project sure isn’t going to let it be posted online.

  71. John, Your comment “In my opinion, EOT is as good as dead” has me wondering if you fully understand where EOT is at, and Ascender’s proposal for EOT Lite.

    As you may know, Microsoft, Adobe and Monotype opened up the proprietary EOT format and published the specifications to the W3C. Members of the W3C (specifically representatives of Mozilla and Opera) said that 2 features of EOT would not make it fly: the patented MTX font compression and URL Binding.

    Subsequently Monotype has offered the MTX font compression under GPL if the browsers said they would implement EOT.

    We proposed EOT Lite as an alternative as well (with no font compression and no URL binding). Our proposal specifically addresses the two issues that the browsers found objectionable.

    The beauty of EOT Lite is that it works today in IE, and could be implemented in months by the other browsers. As others have noted, any alternative web font format will take years for widespread adoption.

    That’s why we believe EOT is not dead. Far from it. It is alive and well. And it could become the solution that brings font-specific, cross-browser web pages to the world very quickly AND provides a level of comfort to address the piracy issues that are of significant concern to type designers & foundries (with raw TTF/OTF fonts).

  72. Tiffany

    Jens,

    I was quoting someone else with that phrase. The html didn’t work.

    Randall,

    So you are suggesting that type foundries need to come up with even more products to make their fonts worthy licensing? I’m sorry, but aren’t the fonts themselves worth licensing?

    So what do you suggest? That a foundry license a font and let people do what they will with it? Shouldn’t a foundry be allowed to make choices on how they license their type? (Yes, this might limit the market, but limit can drive up desirability.)

    Some foundries, ShinnType comes to mind, already are or are working on separate licensing tiers dependent upon usage. Basically they offer TTF for the web with limited glyph palette, no bells and whistle, etc. and then the full glyph palette, all bells & whistle OT version for the print world. Is this acceptable? I think this is fair. Do you?

    Is it just me or is there a lot of “I want my cake and I want to eat it too” in these comments?

  73. Jens Alfke

    Bill Davis wrote “The beauty of EOT Lite is that it works today in IE, and could be implemented in months by the other browsers. As others have noted, any alternative web font format will take years for widespread adoption.”

    I don’t buy this at all. Why is it that EOT Lite, which sounds like it has some rather complex differences from regular OpenType, could be implemented in three entirely different browser engines in months? It sounds, rather, as though it would be significantly _more_ complex than adopting any of the simpler proposals.

    “Monotype has offered the MTX font compression under GPL”

    That would be no help at all to WebKit, Chromium or Mozilla, all of which use source licenses incompatible with the GPL, or of course to Opera which is not open-source at all.

    Frankly, the easiest route sounds like it would be for the one remaining holdout (Microsoft) to adopt the @font-face standard that everyone else has already implemented. If not, MSIE’s ever-shrinking market share may solve the problem in a few years…

  74. ron

    all these ideas are bad as several people covered above. but i’ll add the sooner everyone admits that only we care about fonts and the vast majority of our clients dont give a shit, the sooner you’ll all realize what a waste of time, energy, and money all the above webfont solutions are.

    the only thing that has a future is using the growing number of free fonts that let you use the @font-face tag simply and without worry. That’s what I’ll be doing instead of chasing the dream of using some hacked solution just so I can use a fancy pants font instead of a similar free to use one that gets the job done without headaches or extra expense.

  75. Tiffany
    I have now fixed my CSS blockquote; now works within comments, and is styled differently.

    I need to rethink comment styling — make it easier for visitors to style their their comments.

  76. Tiffany

    Foundries absolutely deserve to gain from their hard work. But, as I said in a previous comment, it is entirely up to the individual foundry to make the choice. They are within their rights to license their creative any way, shape or form they please. More power to them, that’s exactly the way it should be. But, the foundries that accept the limitations of the technology they’re working with and focus on innovating new strategies stand to gain the most.

    We’re in a radically different market reality thanks to digital technology. The game has inherently changed and new ideas must be brought forth in order to compete. My main point is that DRM does not work and trying to debate systems that involve it is a complete waste of time. Other solutions must be conceived.

  77. Tiffany

    Randall,

    Then I don’t think you and I sit too far from one another, basically. Although I do think foundries should be able to differentiate the print market from the web market. And in turn a new format, or some other way of keeping the fonts for the web on the web and the fonts for print (including PDF, etc.) on the “page”. (Although I realize that many don’t acknowledge a difference.)

    Do you (others?) think a limited version of a font is DRM?

  78. Tiffany

    Randall,

    Although, I should add, I’m still sitting just on this side of some form of protection for the foundries. But what I don’t know. If what the wind blows is correct that definition needs to happen quickly otherwise A) foundries will be stuck with raw fonts and B) this means very few foundries will license font linking and you (we) will all be stuck with sIFR and such for most foundries.

    I’ll add you are right. Those foundries that figure out a way to work with the web industry will benefit, definitely. If we can go back several years, when PDF embedding was still new, you’d find a lot of foundries not allowing it because of fear. Now a lot of those foundries do allow restricted* PDF embedding. But for my needs I’m covered. I didn’t fight against them, but I did create conversation and was able to see that change happen. So, it is good that this conversation is happening, but at the same time I don’t see a meeting in the middle.

    *Restricted in that it is usually internal and/or service bureau. No public sharing of these PDFs. And especially no resale of the PDFs, that is usually extended licensing.

  79. Jens -

    No, EOT (full or Lite) is not overly complex. You can read the specification here: http://www.w3.org/Submission/EOT/

    What we have proposed is to address the two stated concerns of the browser makers to EOT.

    Truly, this is not that difficult to solve. Unfortunately I think a lot of folks are getting sidetracked or not fully understanding the issues (hey, possibly including me with respect to the GPL comment I made about MTX).

    So let me try to recap the situation we have today: web designers who want to implement a downloadable font to enhance the display of their web pages need to use EOT for IE 6/7/8, and can use raw TTF/OTF fonts for the most recent versions of the other browsers that support font linking.

    But because almost all commercial type designers/foundries do not allow their fonts to be posted onto web servers in raw TTF/OTF format, web designers don’t have a solution to use the commercial fonts they desire.

    OK, so now we have new services such as TypeKit or Kernest that obfuscate the raw TTF/OTF font files. If type designers/foundries embrace these services, then this is one way to get around the lack of a single web font format across all browsers.

    But these services impose a layer of cost on type designers/foundries that will be passed along to web designers. Also, some web designers may not want to keep their font assets on their own servers along with the rest of their site assets. So TypeKit and Kernest may not address the needs of all of the web designers.

    So there is still a need for a single web font format to address commercial fonts. And yes, I believe that EOT (full or Lite) is a good short term solution. So do many others.

    Is the .webfont proposal better? Or what about the proposal for .zot?

    We think both are great ideas. But unfortunately they will take years to reach broad adoption.

    Note: you can read our proposal for EOT Lite here:
    http://blog.fontembedding.com/post/2009/06/29/Revised-Web-Fonts-Proposal.aspx

  80. Bert Vanderveen

    How strange that people here keep postulating that no one is going to pay for fonts on the web…

    I think that as soon as there is a standard that can be used by the majority of users (in other words as soon as the major browsers use that standard) — a lot of companies will use (custom) type to extend their corporate identity to the web in ways that are not possible now.
    They will commission type, they will want exclusivity, they will want to stay ahead of the competition… And that will lead to a lot of smaller companies to adopting less expensive, but still high quality type.
    And the bell curve model being what it is, after that everyone’s nephew and obscure church club will use the most awful freebies available to have their own ‘look’.

    And let me assure you: the day Amazon premiers their own webfont, legions of lawyers will protect that exclusivity. And so on for the rest of the big corporations.

    Webfonts WILL be a viable business, BECAUSE there is no such thing as free beer for all.

  81. To a developer, suggesting a new format when there’s one that works very well, isn’t patent encumbered and already functioning…makes us cringe.

    And realistically solves nothing. Software makes it trivial to convert from JPG to PNG to GIF to TIFF to whatever you want. And in the case of loss-less formats—converted without degradation. So with pixels, likewise with cubic and quadratic splines. If the bits are on your machine (which, they have to be in order to render the font) they can be copied, no matter what wrapper you put them in.

    Oh, and another thing I don’t see anyone point out in these discussions. What of the type previewers that allow you to demo arbitrary text strings? Set the previewer to the largest size, pop some screenshots, trace it in Illustrator and you’ve got a reproduction that is close enough to make anyone willing to pirate fonts very happy. These people not only can’t tell you what the difference between a textura and fraktur blackletter script is, but don’t even know such a difference exists! Those of us who know the value of having a complete typeface with small caps, text figures for numbers, ligatures, hinting, etc. are still going to be willing to pay for proper type.

  82. dave

    “Do you (others?) think a limited version of a font is DRM?”

    Well, it depends on what you mean by “limited”.

    If by limited, you mean a font having a subset of the characters/configuration (as provided from the foundry to the person designing the web site, not the technical process of the web site delivering the font to the client web browser, which may optimize the font to speed delivery). That amounts to font ‘lite’.

    If by limited, you mean licensing restrictions, then that just reduces the possibility of their fonts being used, either because of an explicit restriction of the license, or just because a designer just won’t be bothered to wade through the license trying to figure out if they can use the font or not.

    If by limited, you mean some kind of technical limitation preventing it’s use in some way, even using a different format, it’s just something new to make things more difficult and to get in the way of people to be able to do the right thing (if they are going to at all). With a different font format from existing ttf (or whatever), there either has to be tools to convert from ttf or whatever to this new format, retaining most/all of the fidelity of the original, or else this new font format won’t be used at all (as who will bother taking the time to reimplement their existing fonts for the web (assuming the foundry cares about how it’s font looks on the web vs print or in local documents on the computer).

  83. Oliver

    For those who think EOT is simple, I’m sure both Mozilla and Apple would be ecstatic to accept patches for Gecko and Webkit respectfully.

  84. Oliver

    Actually, disregard that - according to the Webkit bug, the EOT format is apparently patent-encumbered due to a patent on compression owned by Monotype, meaning Mozilla won’t be able to implement it in Firefox due to its open-source nature.

  85. dave

    “But because almost all commercial type designers/foundries do not allow their fonts to be posted onto web servers in raw TTF/OTF format, web designers don’t have a solution to use the commercial fonts they desire.”

    Replace references of font with “music”, “graphics”, “photographs” and you can see the problem is not with technology, but with the foundries. For a decade, the music industry refused to make their music readily available for a reasonable price on the Internet. Now, they are making money hand over fist on it (while screwing the actual musicians over while their at it). Some people are willing to dive in, some are just willing to stick their toe in, and some insist it’s just not safe unless they can make their own special pool with a 10 foot fence around it.

  86. House Industries is a great example of what’s possible when a foundry focuses not only on their core assets (the typefaces themselves) but the opportunities within halo products. Many of their faces are shipped in limited edition packaging that is a collector’s item in and of itself—added value. That’s on top of t-shirts, furniture, books…even pillows.

    For instance: http://www.houseind.com/fonts/tikitype

    “In addition to eight fonts and a Native Art collection, you get 12 surf tunes on a multi-session CD from Estrus Records including Man or Astroman, Impala, Volcanos and more!”

    So, yeah, “type foundries need to come up with even more products.” Not because filling out your product halo is a requirement, but because your competitors are doing it better than you and the market desires it. That’s just a basic business principle.

  87. k.l.

    One of your “Final thoughts”: Most fonts are not optimised for on-screen viewing, so, if they are to compete with those that already are (e.g. Verdana), then they have lots of work ahead of them.

    What did Microsoft pay you for saying so? ;-)
    Switch on your Mac and see that whichever typeface you use, TTF or CFF-OTF, hinted or unhinted, you’ll get good results. A bit too bold, yes, but you can recognize the original design even at text sizes.
    If, to achieve a halfway comparable result in Windows you need to serve (a) a particular font format and (b) extra instructions, then my conclusion is that something is wrong with font rasterization in Windows (regardless if “normal” greyscaling or “ClearType”). With @font-face and a large number of typefaces of various formats and quality to choose from, it should become hard to justify why only a few especially prepared fonts should look ok while the rest is barely legible or requires guessing at what the typeface was supposed to look like.
    It is not the fonts. Apple provided the proof of concept.

  88. Bert Vanderveen

    Randall, over here in Europe we buy and eat our cereals for their nutritional value, not because of the nice trinkets included in the box. That part is called marketing…

  89. Bert

    And that is the part that creates profit…which is what the foundries are after, so they can continue to create great type.

    In markets where basic needs are met, where there is plenty of “nutritional value”, going above and beyond is what makes the big money. Sure, you can get a generic MP3 player for $50 now that will play your music back just fine…but why does the iPod dominate? Providing a better experience—and marketing.

  90. I’m failing to see how TypeKit is preferable to something like sifr. It’s still going to bog down your page, only now you’re paying for it. Are designers really expected to design a site, and then ask their client that they pay to use a font on the website?

    The idea of forking over hundreds for a font then having to pay for it endlessly to use it on a site makes my stomach turn.

  91. johno

    In case you missed it: fontdeck, from Richard Rutter & co., joins the fray.
    Twitter: @fontdeck

    fontdeck

  92. dave

    johno: You do realize that it appears that ‘fontdeck’, at least so far, is just a single static web page with a nice icon, with no actual information about what it might actually be about other than “web service delivering real fonts to your website”.

  93. I was quoted on twitter as saying:

    “Thomas Phinney: URL binding was a non-starter because vendors don’t want to enforce that. Worse, users might be open to DMCA liability.”

    That’s not quite an accurate representation of what I said. It was because BROWSER vendors didn’t want to enforce it, and because browser vendors (not users) were worried about DMCA liability.

    Above it was also written: “So what happened to EOT? To cut a very long and complicated story short: it didn’t gain the necessary support from foundries.”

    That is really not at all the case. The reason EOT failed a decade ago is because it was a Microsoft-only technology so users didn’t want to use it and didn’t demand foundries support it. The reason EOT failed again eight months ago (with the proprietary issue removed) was that *browser vendors* refused to support it because of DRM concerns (with a bit of patent concern as well), even though foundries actually supported it at that point.

    I agree that EOT in its original form is dead. But EOT Lite is not. WFNT (.webfont) has a number of nice properties, and a number of drawbacks. So does EOT Lite. I agree that it is in foundries’ best interests to get behind one option and push that to the W3C. But writing nine paragraphs about WFNT and none about EOT Lite is not a fair way to deal with that situation.

    Cheers,

    T

  94. Dave
    Richard Rutter (of fontdeck) first mentioned he was working on something Typekit-esque in response to some article about the Typekit proposal. I shall try to track down the source for you.

    k.l

    What did Microsoft pay you for saying so? ;-)

    I wish they had paid me to say it ;)

    Thomas
    Many thanks for clarifying, especially in regards to the failure of EOT. I shall add links to your comment in the article. I had intended to qualify my EOT is dead with a paragraph about EOT Lite. Perhaps after several more cups of coffee, I’ll do just that. If I were really cheeky, I’d ask you to write that paragraph :)

  95. Matthew Buchanan

    Great, lively discussion. If Typekit et al are the stepping stone to an eventual cross-browser solution, I’m interested in how many of our clients might be willing to pay ongoing fees for the use of different/better type in their sites (especially when many of them specify Arial as their corporate face for internal documents. :) A one-off licence I can probably sell, ongoing payment on top of hosting fees will be a new challenge.

  96. Oliver, regarding your comment about the MTX compression, the patent IP issue is what EOT Lite solves.

    Again, EOT Lite removes MTX compression and URL binding (the 2 features that the browsers objected to).

    Only the existing version of EOT is dead. But the new version we proposed is very much alive and worthy of serious consideration and discussion.

    Sidenote: We apologize to everyone for coming up with “EOT Lite” in the first place. It sounded better than “Version 0x00020002a” as a way to easily discern the differences in the current version of EOT with our proposal for a revised version.

  97. Audio of TypeCon2009 Web Fonts panel discussion now online. Pass it on.

    http://typecon.com/talk.php?id=333

  98. You don’t need tipekit, webfont or any other solution.
    You can start using real fonts on the web right now.

    Just use font-face and and declare 1 rule using eot for IE and another rule using OT for firefox and safari.

    I’m doing it right now and you can read MuseoSans on mi site, on any of those browser. Have a look:
    http://www.pabloimpallari.com.ar/

    Creating the eot file is easy as 123 using ttf2oet:
    http://code.google.com/p/ttf2eot/

  99. …….”@Xiao:It is the average people, who will NOT pay for professional fonts no matter how professional they are”……..

    i guess you miss the point Xiao to “pay for fonts if you are a designer is not a payment but sharing resources to continue to enjoy our life with creativity, and create a happy way to get there. Look the example of Firefox Add-ons Contributions Pilot idea, we know………..lot of people will steal anyway but to support development, and the people who create is important, and naturally not everybody is ready to share, but that ones just garbage on the web with “strong “democratic explications (that ones pass good on design milk as printed toilet paper)

  100. Bert Vanderveen

    Yep, type designers’ feng shui… Rewards to be reaped in the afterlife.

  101. In case you missed it on Twitter, I’ve just now added a short paragraph on Ascender Corperation’s EOT Lite.

  102. Oliver

    Then I have no real objection to EOT Lite, assuming that the people who are trying to encourage it are willing to do the leg-work and the advocacy needed to get it into Firefox and Webkit. Universal support of EOT Lite could dramatically speed up font adoption since IE won’t be able to drag adoption back.

    However, I think in the long run there will be zero advantage to its obfusation. It’s trivially crackable, software will almost certainly ignore any licensing bits, and in the long-run all that will happen is that applications will support using EOT files directly. But if some false short-term protection makes font foundries feel better, it’s hard to be against it.

  103. Oliver

    Nevermind, it seems apparently Mozilla is actively against EOT Lite as well, so it’s utterly doomed as a format.

  104. @Jos Buivenga (exljbris) I bet you’re pretty happy right about now that you made some fonts available with @font-face permitting licenses, if only for the free publicity alone! :) If (when?) you make paid @font-face fonts available please let us know how it works out.

    @Jens Alfke Given how pants text looks in Internet Explorer (well, in Windows generally) I actually wonder if .eot will be ignored and @font-face will only be used as a progressive enhancement for modern browsers, at least in the short term.

    @Bill Davis I think you overestimate the speed of browser updates. While IE users would take about 5 years, even modern browsers would require a couple of years before widespread adoption of any new format. AND as @Jens Alfke mentioned you’re talking about something that’s complex to implement (ie won’t be ready any time soon). 2011 is as head-in-the-sand as 2014, and IMHO “within a year” for usable adoption is optimistic to the point of comical.

    Also, I was under the impression that recent versions of IE actually required a URL in the URL binding field (whereas it could be blank in earlier versions). Is this incorrect?

    @Tiffany Is it just me or is there a lot of “I want my cake and I want to eat it too” in these comments?—more like a crowd of hungry people outside the cake shop with money in their hands looking at the delicious window display being told “we don’t want your type around here” ;-)

    @johno A list of allowed elements/attributes would be a nice start ;-)

    @Tiffany by “limited” do you mean subset? No, as long as the subset can be changed (preferably automatically) when the content changes.

    @bethany Are designers really expected to design a site, and then ask their client that they pay to use a font on the website? Yes. This is called “commerce”, or as Homer Simpson’s brain defined it ‘the exchange of money for goods and services’. It’s quite a common idea :P Of course the difference between a one-off price and a how-much-can-we-screw-customers-over-per-month price may indeed be nausea-inducing.

    @johno Richard Rutter first mentioned Fontdeck in Andy Clarke’s Webkit article comments.

    Finally, here’s a summary of web fonts and .EOT by Bert Bos from 2008. Does anyone have a link to a summary of EOT Lite that covers browser positions?

  105. .webfont is clearly the foundries’ favorite long term option, but if they want a non-raw format within the next couple of years they really need to come together and back some variation of EOT.

    I second this.

    Personally, I think that TypeKit has killed off any chance of an EOT Second Coming.

    I don’t think so. It’s a nice solution for now, but don’t forget all the people who don’t want to have any third party involved in their work [corporate web sites alone]. Living with Typekit & Co. forever means to accept them as a kind of a standard. It’s an option, but shouldn’t be a standard.

    Ivo might be surprised to learn that he’s now working for FontForge :) Off to give him his old job back.

    Thanks a lot =)

    When all of the browsers support it, @font-face will happen. With or without some of the foundries. Security through obscurity won’t prevent that.

    The most important browser won’t support ‘raw fonts’ because Microsoft understands the foundries’ issues. Even if IE will support it, this needs a few years as well. So why not .webfont or EOT Lite?

  106. dave

    Ivo: “The most important browser won’t support ‘raw fonts’ because Microsoft understands the foundries’ issues. Even if IE will support it, this needs a few years as well. So why not .webfont or EOT Lite?”

    This kind of thinking really hurt the music industry. Microsoft’s DRM let all the music labels set their own limits. Everybody came up with their own, and they made the online music stores use those limits. Customers stayed away in droves both because the DRM itself sucked AND because for a given price, the customer didn’t know what they were getting (some songs would only work on one computer, some on a couple computers, some on portables). It took Apple to define a common set of limits for all the music on their store for people to start purchasing music in volume online (and got the industry to stop using DRM, for which it’s making more money online [and growing faster than ever]).

    And Microsoft bending over to get Blu-Ray support into Vista added non-trivial costs to all computers, even ones that don’t have a Blu-Ray drive. Hardware needs to be created so it can detect if somebody has modified it. Software drivers need to be written to test and make sure the hardware hasn’t been modified. It might not be executing all the time (or ever if the system never gets a Blu-Ray drive connected to it), but the support for all that needs to be in both the hardware and software.

    Simply put, industries don’t seem to take well to big changes such as these, and when the industry gets to set it’s own terms to make the change (we know best because it’s our IP), they pretty much blow it every single time.

  107. k.l.

    Jens Alfke:

    EOT Lite, which sounds like it has some rather complex differences from regular OpenType

    EOT Lite = EOT = wrapper around regular OpenType. Nothing “rather complex” at all.

    Pablo Impallari:
    The question is not if you already can use EOT for some browsers and TTF/OTF for others (that’s not news at all), but if the license allows allows you to link to TTF/OTF fonts.
    (Also see Tiffany’s clarification above: “Note that Jos (exljbris) only allows @font-face embedding in his free fonts.”)

  108. miha

    Pablo:

    Creating the eot file [of Museo Sans] is easy as 123 using ttf2oet

    It is actually more complicated, license says you may not modify the font.

    I also agree with k. l., there is something wrong with Windows font rendering — evenmore, it’s especially jaggy in headline sizes, and it looks like it’s not even anti-aliased. Don’t worry, you will notice it when @font-face comes.

  109. zork

    Firefox and Webkit are together probably important enough not to ignore in this and I am fairly certain they will not support any DRM or semi-DRM effort. Anything they implement will have the code out in the open so breaking the so-called protection will become less than trivial too.

    Professional photographers get along without obfuscation and DRM in the way browsers display images, video with HTML5 is going to be similar. I think type should too.

  110. Chris Fynn

    “most of the larger and important foundries have come out in favour of the .webfont proposal”

    Larger & Important? Where are Monotype, Linotype, Adobe, Bitstream…?

  111. Philip Taylor

    @Oli: “I was under the impression that recent versions of IE actually required a URL in the URL binding field (whereas it could be blank in earlier versions). Is this incorrect?”

    I’ve generated EOTs with this code (used e.g. here), which sets a blank root string, and it appears to work successfully in (at least) IE8 on Vista. I’d be interested to know if there’s any version in which it fails.

  112. Larger & Important? Where are Monotype, Linotype, Adobe, Bitstream…?

    FontFont is the largest foundry that already supports ‘.webfont’ as a long-term solution officially. But as far as I know the other “big” foundries would also be fine with it, but may have some concerns about the possible time frame. In my opinion this is not a reason not to support this proposal.

  113. First, let me state clearly that I’m not a fan of any DRM. But I could not ignore the flaw in this argument I read over and over the last few days:

    quote by zork: Professional photographers get along without obfuscation and DRM in the way browsers display images, video with HTML5 is going to be similar. I think type should too.

    The huge flaw in this argument is that 72dpi 500x350px photos are not usable in any professional publication and that the same videos, like youtube videos, cannot be used by another video producer to create a professional work (except if he wants the pixelated look). Unlike raw fontfiles, better said the perfectly sharp vector descriptions inside it, which can be used for any professional work afterwards.

    The only parallel is to authors of written words, but this has been the case forever, unlike music and video which are both newer mediums (in their recordable, copyable, digital form).

  114. Thanks for this roundup! It was really solid and informative. Typekit is the most reasonable workaround at the time and I cannot wait until .webfont is approved…Great article!

  115. The huge flaw in this argument is that 72dpi 500x350px photos are not usable in any professional publication and that the same videos, like youtube videos, cannot be used by another video producer to create a professional work (except if he wants the pixelated look). Unlike raw fontfiles, better said the perfectly sharp vector descriptions inside it, which can be used for any professional work afterwards.

    That’s the point. [Watermarked] 72dpi JPEGs are not 300dpi TIFFs or RAWs. Even an 128 kBit/s MP3 file is not CD quality. But raw fonts is still the original data, that is usuable like the data you will get purchasing it legally.

  116. @oli
    Also, I was under the impression that recent versions of IE actually required a URL in the URL binding field (whereas it could be blank in earlier versions). Is this incorrect?
    Yes, incorrect. As in wrong. Weirdly, it’s backward compatible. EOT “Lite” as it’s called works in IE6,7,8.
    Johno’s included a link somewhere on this now lengthy page to my explanation at:Jeffrey Zeldman Questions The EOT Lite Web Font Format
    Also, coverage of the Web Fonts Panel at 2009 is also posted on Readable Web: Web Fonts At TypeCon 2009
    There is one argument in favor of EOT Lite as an interim solution that is difficult to argue against: it is the quickest way to get Web Font support to the most people in the shortest amount of time.

  117. Tiffany

    Customers stayed away in droves

    Customers did not stay “away in droves” from iTunes. Even with DRM people spent money. Yes, some people did so begrudgingly, but they did not stay “away in droves”.

  118. Tiffany

    Oli,

    If that is the case then I hope for all the cake lovers that something like .webfonts and/or EOT Lite is approved quickly so people can start licensing to their hearts content.

  119. Chris Fynn

    raw fonts is still the original data, that is usuable like the data you will get purchasing it legally.

    Yes, but no company in their right mind is going to embed a font on their website which they don’t have a proper license for. Unlike in print, doesn’t the fact that a font contains the original data actually make it easier for the copyright holder to detect and prove their font is being used and take legal action if that use is unlicensed?

    Font embedding may hurt sales for use in printed publications, where it is far more difficult to detect and prove unlicensed use. However, as others have pointed out, in most cases any designer prepared to use a font in print without paying for a license can already easily obtain a copy of almost any desired font via a bit torrent site or similar source - The availability of fonts embedded on websites is not going to change things much in that regard.

    .webfonts will still contain the original data and extracting an original font from its wrapper will probably be little more difficult than opening a zip or rar file is today.

    IMO the best protectiion for copyright material, including fonts, is proper licensing and, where necessary, legal action - not DRM, obfuscation or new file formats.

    - CF

  120. @k.l. While EOT Lite may be easy to describe, implementing it is most probably not easy.

    re: Also see Tiffany’s clarification above: “Note that Jos (exljbris) only allows @font-face embedding in his free fonts.”; to tell the truth I don’t think this is relevant. Whether the license allows @font-face use (plus .eot generation and subsetting) is of primary importance. Whether a designer then has to bill the client or not for the use is secondary.

    @miha Yet the two free fonts are licensed for @font-face embedding. Does that mean .eot font embedding/subsetting too? I don’t know. Why isn’t this stated clearly in non-legalese? I don’t know that either.

    Dear font creators, how about a human-understandable summary of your license? Surely you dislike reading that stuff as much as us other non-lawyers, right?

    @Philip Taylor OK maybe that’s an urban legend or something I imagined—if it works in IE8 then I guess the URL binding field can be blank (I checked IE6/7). My mistake.

    @Richard Fink I stand … er sit corrected.

    (@Tiffany too) Also, we already have web font support, just a dearth of fonts. Even that might not be a problem if a few font creators try new licenses; only a few good fonts would go a long way. I think we’ll be seeing a lot of Fontin and co on the web soon.

    Pragmatically I don’t think designers are going to wait several years for magic bullets to gain widespread adoption—they’re going to use what’s available (and permitted) now, with the tools that work now. That currently means paid fonts and text as image, or one of a growing number of … free fonts with @font-face.

    If font creators want to wait several years in the hopes of a magic solution that’s their choice, but I’m betting some won’t, and they’ll be the ones web designers buy fonts from.

  121. I love to personalize everything i can on my comptuer. great info

  122. Walter Permican

    “Downloading those font files would be as easy as downloading an image. For obvious reasons, foundries don’t want that. In fact, no-one wants that.”

    What do you mean “no-one wants that”? Everybody (but the foundries) wants precisely that. And that it what will ultimately win.

    There are plenty of great free-as-in-beer fonts out there that this proposal will never get off the ground.

  123. Torbjørn Vik Lunde

    @Tiffany:

    Who said anything about no compensation? I’d pay for a good typeface optimized for the web, and I think a lot of others would as well.

    “Basically they offer TTF for the web with limited glyph palette, no bells and whistle, etc. and then the full glyph palette, all bells & whistle OT version for the print world. Is this acceptable? I think this is fair. Do you?”

    As much as I’m skeptical towards separate formats for print and screen I think this is a lot better than using EOT or the like.

    @Jens Alfke

    Well, MS‘s renderer might be bad, but it‘s the one that they are using. What are we going to do? Make Microsoft use another renderer?

    (I‘d love for MS to have a better font engine though!)

  124. You are indeed right. Web designers always look for more options. There is nothing that can satisfy them and that is what motivates them to create unique designs. Personally I don’t believe in web-safe palette of fonts and I have already worked with various typefaces. And I always crave for more.

  125. Mono

    Well… I think it’s time to face reality instead of trying to change and force things to happen in an ethical and correct way. We all have at least once downloaded a few songs via P2P, we all have more than once used cracked software, and most of us (graphic and web designers) have NEVER bought a type in our lives, even if we proudly show our good taste on choosing and using them. And whoever denies this fact is either a lair or a stupid.
    Having that in mind, I think that Type Foundries should understand that fighting this war is only preventing them from selling more fonts (potentially) and is leading some people towards using Arial instead of Helvetica and Comic Sans instead of anything else.
    And as for downloading the raw ttf file… there are at least 100 sites/blogs/forums from where you/me/we can download, for example, the FULL emigre collection without paying a penny.
    So, in conclusion, type foundries are as “dead” now as they would be if they let their TTF files be used via CSS.

  126. Mono

    Oh… and one more thing. Whenever we Graphic Designers send a project to the printer, we send our Quark or InDesign files along with the pictures and, guess what, the font’s TTF (or whatever format our font was).
    So, potentially, the printer will keep our font files and will share them with everybody else, not giving a damn about royalties, permissions, restrictions, etc etc.
    And again, if anyone says these things don’t happen…

    From my point of view, the only chance Type Foundries have to get a share of the web market (because at the end this is what the whole discussion is about) is to behave the same way they behaved when “software” came along: STAY COOL and get your fonts used as much as you can, both by paying and non-paying clients/designers. And eventually you will sell more (and never less) products.

  127. Torbjørn Vik Lunde

    “We all have at least once downloaded a few songs via P2P, we all have more than once used cracked software, and most of us (graphic and web designers) have NEVER bought a type in our lives, even if we proudly show our good taste on choosing and using them. And whoever denies this fact is either a lair or a stupid.”

    While I admit to have downloaded various material via P2P (usually to find out whether it is worth actually buying) I think the claim that most graphic designers never have bought type is a pretty bold claim.

    I’m barely a graphic designer at all (had a small internship, but plan on taking some more education), but I have already bought type. Every serious designer I’ve ever talked to have many fonts. When I was at a very short internship at a agency they had literally hundreds of fonts, and I’m pretty sure they paid for all of them.

    I do agree though that if foundries worry less about losing money, and not least: stop thinking in the way that every pirated font is one that people would otherwise buy. I don’t think that is necessarily the case. I think by focusing on the untapped market of web fonts they could probably make more money than they did before, not less. (This is just my thoughts though, I’m by no means an expert in economics. *laughs*)

  128. mouras

    Interesting discussion. I’m building sites since 1995 and I have a few points to add:
    1. @font-face is the SOLUTION, right here, right now. Web must have its own fonts. There should be created central repository by W3 with open fonts.
    2. The Font companies are just stupid. All these years there is an immense market open for them and what do they do?
    3. Independent font creators should be more open to the net and start licensing fonts to websites at low prices so everyone can afford them and not have to steal them.
    4. .webfont is a marketing solution. I just don’t like it.

  129. Mono

    @Torbjørn Vik Lunde
    You are right, it is a bold claim.
    But I think it is a fact that if at least a 20% of the designers would have ever bought a typeface, there would be a “Type Industry” by now!

    @ mouras
    I agree with you in almost everything.
    Almost because I believe that nobody HAS to steal anything… people just CHOOSE to do so… and we are not talking about a slice of bread not to starve, if you know what I mean ;)

  130. Interesting observations. I’ve done some work with embedding .swf fonts and it seems like a fairly practical implementation if having a unique font for the text of your site is a priority.

    But you’re right that finding fonts optimized for web viewing is the real trick, because it’s not a priority for most designers.

    That said, some years ago I designed a set of web fonts which were tested really extensively and which seem to work quite nicely. They’re free to download and try out and feedback is welcome. There at fontcraft.com.

    Dave

  131. 2. The Font companies are just stupid. All these years there is an immense market open for them and what do they do?

    It’s not stupidity, they just don’t know how to deal with the concept. For a web font to really work it has to be available for free so they can become really widely disseminated, and companies that want to make money have a hard time figuring out how to do that with a product which they have to give away free in order for it to be useful.

    Dave
    fontcraft.com

  132. Cliff’s Notes Version
    1. We’re f*cked.

    And it’s too bad. The expanse of fonts on the web would be a great help for designers and developers.

  133. Russell, no we are not!

    Let me try to explain how I see this.
    1) Web designers and many companies/brands want to use commercial fonts (or their custom/bespoke fonts) with @font-face to enhance the typography on their websites.
    2) Type designers/foundries want to license their fonts for use with @font-face, but NOT in the raw TTF/OTF font format.
    3) Web browsers today support two different web font formats: IE supports EOT, while FireFox/Safari/Chrome/Opera support raw fonts.
    4) New web font formats are being proposed to establish a single web font format that can work across all browsers (EOT Lite, .zot, .webfont).
    5) New services are emerging that can deliver both EOT and raw fonts (with the raw fonts being obfuscated to prevent font piracy).

    Over the past few months this topic has really heated up and all the stakeholders (web designers, type designers and browser makers) are talking about how to solve this.

    So that’s good news IMHO.

    We came up with the EOT Lite proposal because we saw a way to bring a single web font format to all the browsers in the shortest time frame. Yes, other proposals like .zot and .webfont might even be better, but these will take 3 to 5 years to reach a significant adoption rate (due to lengthy upgrade cycles and users not upgrading automatically to new versions).

    We are selling web fonts licenses now (in EOT Lite format) on our commercial font website: http://www.AscenderFonts.com. Monotype is also offering EOT web font licenses. And if other browsers add support for EOT then web designers can start deploying a single web font format for all their commercial fonts or custom fonts.

    In summary, I don’t think the situation is as bleak as you think.

  134. What am I missing here? Why isn’t the .swf format a viable way to do web fonts. From what I can tell it works pretty decently.

    Dave

  135. Why even bother? With CSS3-polygons you can make your own fonts on the fly! It does take a little longer, but you can always automate the process with PHP/Javascript and the end-result will be smooth. :3

  136. Chris

    The situation seems very black-and-white to me, as a some-time web designer and some-time programmer, I offer the following comments to the foundries.

    (1) I will pay (and have paid) a reasonable price for commercial fonts, because I like my personal work to look good, I like to do a professional job for my commercial clients, and as a matter of principle I will support those who help me do this.

    (2) I will not pay for or use any commercial font with a stupidly restrictive licence agreement, or any licence agreement I cannot quickly and confidently interpret as a non-lawyer. For example, I automatically rule out fonts that have any restrictions on PDF embedding today. Once web font embedding takes off, it will be added to my list of unacceptable restrictions. There will be plenty of places supplying fonts that are good enough where I don’t have to worry about these things or pay my lawyer to check the agreements, so I have no reason to jump through your hoops if you are too short-sighted to do offer your fonts on similar terms.

    (3) I will not make any of my own sites depend on an external service to provide resources, fonts or otherwise. This is a major flaw for both reliability and security, and is not and never will be negotiable.

    (4) Even if I personally were willing to depend on a third party, it is unrealistic to expect my professional clients to pay me for a job but then continue to pay an unknown “rent” in perpetuity to another party to keep it working.

    (5) I will not use any silly, DRM-encumbered formats. Give me the real deal and just let me use it as easily as an image file. Sure, someone else might download it and rip you off. But they were going to download it from somewhere else anyway if I didn’t have it, and you would have lost a sale to me as well.

    (6) If you want a basic degree of protection against being ripped off, just add a “licensee” field to OpenType, and allow licensees to redistribute your font specifically as part of their own web pages (in perpetuity, royalty-free, etc.) if they register the domain(s) where the font will be used with you. This causes minimal hassle for the user, and means you can search the web for unlicensed users automatically.

    (7) If you want a tiered pricing plan, how about charging just $5 per font (at least for the basic variations) for non-commercial use, but charging a premium for more specialist variations and/or commercial use? Look at Apple selling music: even everyday people will pay a small amount for something they value. A few deals with popular sites like MySpace to allow people to personalise their pages might build a somewhat lucrative market in casual, non-professional users that you don’t have at all right now.

    (8) Please give up on separating web from print use. It is debatable how enforceable any such agreement would be anyway depending on your jurisdiction, and in any case, see (2).

  137. This is an interesting academic discussion.

    However, in the real world, foundries better hurry, because designers can now upload fonts for cross-browser font embedding using the @font-face CSS selector. (Here’s How) Once designers discover this capability and begin using it in a big way, Pandora’s font box will have been opened.

    Personally, I don’t see TypeKit gaining traction. Designers don’t want to (a) pay a rental/subscription fee for fancy fonts; (b) rely on JavaScript to display those fancy fonts or (c) rely on a 3rd party server for their site design content.

    Cheers,
    -stk

  138. type.nasos

    “However, in the real world, foundries better hurry, because designers can now upload fonts for cross-browser font embedding using the @font-face CSS selector. (Here’s How) Once designers discover this capability and begin using it in a big way, Pandora’s font box will have been opened”

    Feel free to use @font-face with the free fonts, noone said
    dont do it, as come in business is not your concern, people
    who create the fonts can do what they want. Like it or dont
    is your choice.

    On the other hand i cheer up all graphic designer go sit work 1-2 years create a font, and upload it free, i hope u have a good support by ur family to feed you those years.

    The discussion is about the how the foundries will protect
    their products, or i didnt understand smth? How you people
    demand… (not advice) how they gonna run their work?
    There is the right away and there is a wrong way.

    For the friend that said that he send the fonts to the printer, well thats ur fault, and u cant realize the morallity that a designer must have. i never sended 1 font to the printer. When there is a problem im always there to solve it. simple ah?

    As for web designers, you all have more options now, with so many free fonts to use them on web, why you complain? Start using those, and until u get bored of them, the foundries will
    already have the solution rdy, no big deal, u were workin so many years on fewer options.

    What free? no decent hinting? well it will teach you a lesson.
    if u want quality staff, you must pay for them, coz behind quality is years sweatin of a designer.

  139. mono

    Dear type.nasos,
    Right now I don’t have the time to say everything I’d like to, but I couldn’t wait telling you this: YOU ARE MISSING THE POINT.
    If you want to make a living out of Type Design, I am afraid you WILL HAVE to adapt to this “new economy” as everyone else is doing. And if you want to know how to survive for 1-2 years while designing your font for free download, well… go ask the Google, WordPress, Indexhibit, Drupal, Firefox, and a long long etcetera guys…
    My point is that if you want a share of this market you will have to do as everybody else is doing.
    Or… you will have to design fonts on demand, for single clients hoping that your hard work won’t spread for free.
    If I were a type designer, I wouldn’t doubt it for a second: I would accept the “rules”, I would let my work spread, I might even ask for donations…

  140. type.nasos

    actually my point of my post was more about the morality
    of designers nowdays, I believe its the start of the black hole.
    Personally my friend, i choose the path of designin custom fonts for my clients only, and fonts that i use for my own design. But in my studio when the web department want to use a font with @font-face now, i have teached them to read and respect their licence, as they want people to respect your work, you have to respect theirs. So they dont whine in forums, they wait until the foundries come up a solution. How a studio dislike his artwork get stolen, is the same feelin if a studio copy-steal-use illegal the fonts. Getting off point, so ill stop here.

  141. I’m hoping it does come to our web browsers soon to be honest.. Would look really nice! :)

  142. What a great summary! I had no idea tangible progress was being made at all, so this is a welcome surprise.

  143. Web fonts are applicable + popular art, though not everyone is aware of it. Now the progress is getting faster. Everything is getting faster nowadays.

  144. Just thought I’d confirm publicly that Fontdeck is more than just a static page, it is a real product currently undergoing development. As soon as there’s something complete to play with, ILT will be at the top of the list for a preview.

  145. Noneya Business

    You can still ‘extract’ webfonts. You are only creating an extra hassle for your paying customers, alienating potential returning customers, and still getting ripped off as you say. Except now you are ripping yourself off more by also paying extra to have the fonts hosted by some 3rd party who has ignorant faith in their Flash, or meta data, failure of a protection scheme. Scheme is really all it is, a scheme to take advantage of you.

  146. When do I get DRM for my HTML/CSS/JS and graphics? Why do people bother publishing books when they can be so easily photocopied? Shouldn’t we hold out for better Physical Rights Management before publishing? And you’re telling me people can just cut and paste these words I’m typing? Why do I even bother?

    Are we serious? Cars don’t automatically stop at 55mph, but we still have speed limits. If someone takes my design, that’s a copyright infringement (in some cases - thanks CC). Then we deal with it from there. Why is it any different for font foundries? Why the extra protections?

    As designers and web publishers why should we feel any sympathy for the foundries trying to get something none of the rest of us have? We’re not wining. We know that DRM would hurt all of us.

    I’m happy to pay for music, but I’m not happy when you try to control what devices I listen to it on. I’ll pay for your fonts as well, if you’ll just let me use them.

    Dear foundries: we really are your biggest fans, and would be your main customers. Can’t we please buy your product?

  147. Thanks for this awesome overview about web fonts. :)

  1. TypeShit?—Jul 22, 2009

previous post: iQ font

next post: Calluna — a text typeface with flow

October Fonts September Fonts August Fonts July Fonts June Fonts May Fonts April Fonts March Fonts February Fonts January Fonts January Fonts November Fonts October Fonts September Fonts August Fonts July Fonts June Fonts May Fonts April Fonts March Fonts February Fonts January Fonts december Fonts October Fonts September Fonts August Fonts July Fonts June Fonts May Fonts April Fonts March Fonts February Fonts January Fonts December Fonts November Fonts October Fonts September Fonts August Fonts July Fonts May Fonts April Fonts March 2011 Fonts February 2011 Fonts January 2011 Fonts December 2010 Fonts November 2010 Fonts October 2010 Fonts September 2010 Fonts August Fonts July Fonts June Fonts May Fonts April Fonts March Fonts February 2010 featured fonts December Fonts November Fonts October Fonts September Fonts August Fonts July Fonts June Fonts May Fonts April Fonts March Fonts February Fonts January Fonts December Fonts November Fonts October Fonts September Fonts August Fonts July Fonts June Fonts May Fonts April Fonts March Fonts February Fonts January Fonts December Fonts