What are they suggesting should happen to the font once it's downloaded? It seems like current functioning solutions use Flash as a rasterizer in some way. Otherwise, I would think that from a security standpoint the rasterizer would have to be built into the browser. Alternatively, the browser would have to be able to turn a font on, which is a very platform specific activity, and not one that many users would want. If a TT rasterizer was built into the browser, perhaps there would be an opportunity to create a special font format designed only for web use. This is just speculation, but if such a thing were possible, maybe there's a way this could work that wouldn't be completely opposed by type vendors.

Maybe there could be a parking meter on the site. You could use the fonts as long as you kept feeding the meter:-)

You might also license fonts through the iTunes Music store and royalties would be automatically paid to designers and foundries?

ChrisL

Now you're thinking! Howabout the fonts being hosted by vendors? The web browser would have a plug-in that would render the font. The web-designer just enters some sort of key that lets the page load the font off the vendor-hosted server. Perhaps the system would only use sub-sets of the font, so it would be much more difficut for people to devise a way to steal the faces. Remember, iTunes doesn't make it impossible to steal music, it just makes buying it much more convenient.
The trick has always been that there has to be a way for everyone to make money.

For this crazy scheme to work, you'ld also have to get broad adoption of the plug-in. That could be tricky, because it would have to be something that IE/Firefox/Safari could include without licensing entanglements.

Web designers could pay by buying a key for a certain number of hits on a particular font. Of course, piracy would be rampant, but everyone would still be making more money, which is the important thing. Some users would buy fonts from legit sources for the same reasons they do now for print.

>What are they suggesting should happen to the font once it’s d
downloaded?

I think they're open to suggestions. But I think they'll likely leave this up to the browser maker - who may choose to use OS level services.

>It seems like current functioning solutions use Flash as a rasterizer in some way. Otherwise, I would think that from a security standpoint the rasterizer would have to be built into the browser.

I think they need to hear this feedback from the community.

So far the focus has been on free fonts...

-----Original Message-----
From: www-style-request@w3.org [mailto:www-style-request@w3.org] On Behalf Of Håkon Wium Lie
Sent: Tuesday, April 25, 2006 11:12 AM
To: Dave Raggett
Cc: www-style@w3.org
Subject: Re: Downloadable fonts and image replacement

Also sprach Dave Raggett:

> I am therefore looking for support for direct use of TTF files so
> that I don't have to use a special tool for embedded fonts like Weft
> [1]. There are plenty of fonts with open licenses that are perfectly
> good for most purposes, so a DRM-based solution isn't high on my
> wish list.

This is also my experience. There are thousands of TrueType font
families out there. Some examples:

http://www.fontfreak.com/fonts-g2.htm
http://www.1001freefonts.com/fonts/pfonts5.htm

Typically, four truetype files are zipped up into a file which --
roughly speaking -- has a size similar to a photograph. Some of the
fonts have legal restrictions that prevent their use, some only cover
the English alphabet, and some are poorly designed. However, there are
also good-looking, freely usable fonts out there. These fonts
represent an untapped resource for typography on the web.

- - - - - - - - -

So, the elaborate schemes involving micropayments to type deisgners don't seem to be on the table right now.

So, the elaborate schemes involving micropayments to type deisgners don’t seem to be on the table.

I guess I'm off on a lark. I could imagine the proposed ideas being a big disappointment to type developers large and small. A big problem for the small foundries is actually being able to go after pirates. According to this scheme, if a font without proper licensing ended up on the web, anyone could download and use it. One could use Google or a custom spidering tool to just cruise the web and download any fonts out there. Yikes.

Can you close your CITE tag?

I've made these arguments which is why I think they need to hear from those directly invovled in type design.

Cheers, Si

I don't think I ever close the CITE tags. I didn't realize that you had to, since it looks fine in my bowser. Thanks for mentioning that.

I realize that this has probably been discussed a lot over the years, by people much more knowlegable than I. It's really one of the great outstanding problems in typography.

A greater venue for free fonts just further cheapens the fonts for sale. If fonts fall out of the internet for free like rain, why would someone think paying a mere $30 was reasonable?
Software companies and OS companies bundle what appear to be free fonts like crazy already. Unplugging the internet drain plug with a torrent of free fonts would just wash type designers down the sewer with the flood.

ChrisL

Oh that explains it - I'm probably the only one seeing all the italic. Can you fix this too - http://typophile.com/node/19393 ;-)

>I realize that this has probably been discussed a lot over the years, by people much more knowlegable than I. It’s really one of the great outstanding problems in typography.

Actually I don't think it's been given much attention in recent years. It got a lot of coverage back in the early days of the Web (mid to late 90's) but fell off the radar. Now it's back on the W3C's agenda, and the impression I get is that things will move quite quickly, especially if no one from the typographic community raises any objections.

I think Web designers also need to let their voices be heard. Will they be happy with a situation that limits them to freeware fonts? How about corporate marketing departments that will have to decide if they're willing to post their expensive custom fonts on their Web site? At least they have a choice - those companies using off-the-shelf fonts in their identity work won't have the option of posting commercial fonts in this way.

What is the best address or venue to send our comments on this Si? I am sure W3C does not frequent these boards.

ChrisL

I think your best bet would be to read through the posts here http://lists.w3.org/Archives/Public/www-style/2006Apr/ - and hit "respond" - you may get a bounce with information on subscribing to the www-style email list - or it may just go through.

Alternately I've offered to collate feedback and forward it back to the decsion makers at the W3C.

Si,
I just sent my comment from my above post to the list. It just pops open my email account and I sent it in. I have not seen itpost yet. Itprobably takes a while to post.

THANKS!

ChrisL

Maybe this thread should be in the general forum to get more traffic? Maybe a kind birdie will put it in "Hand Picked"?

ChrisL

Sii, do you think there will be some way for Windows to activate fonts in a particular application context? Without it, I just don't see how this could ever work. I mean, no one's going to let their browser put a font in C:\Windows\Fonts\.

On the Mac, this could be implemented today. A browser could simply download the font to it's cache and activate it with the same API Adobe uses for the fonts in /Library/Application Support/Fonts/. If the browser downloaded a font with 'unexpected featues', the worst it could (hopefully) do is crash the browswer.

>Sii, do you think there will be some way for Windows to activate fonts in a particular application context?

For sure - that's easy... http://blogs.msdn.com/michkap/archive/2005/11/20/494829.aspx

Thanks for the link! Someone should implement this as a Firefox or IE plug-in. Is someone already working on this?

@w3c, ie, firefox, opera, safari...
Before embedding fonts and such, there is another important topic first -
hyphenation for different languages depending on fontsize/width of html-elements. Problems with readability and layout occur normally with so called fluid layouts, adjusted fontsize, etc.

regards

Am I the only one who sees a sinister side to this? It is like going to a baseball game where the hotdog vendors who come right to your seat give you the hotdog for free. If you get up and walk to the concession stand behind the grandstands though, you have to pay money for the hotdogs. How long are the lines at the concession stand?

ChrisL

IF this went through would it REALLY be that easy to get the fonts?

The level of discussion seems to range from people who have some clue about fonts and the IP issues involved (David Woolley) to those who seemingly don't (Anne van Kesteren).

One part that really got me (I can't find the exact thread) is where somebody says that, rather than being able to specify fonts to be downloaded, you could satisfy the creative types by adding support in CSS for 3-D dropshadows and extrusion effects. :-P Somebody else mentioned "word art"... hmm. I don't know whether to laugh or cry.

I do see the sinister side of this, but I also see some interesting opportunities. For example, it could be used for a website to display text in a language for which the user might not have text support. This a good thing for people who write with the worlds less common scripts. This might also be handy for corporate identity fonts, it could encourage more business to customize their type. Also, Trekkers and Trekkies could make Klingon available for people who still don't have an OS with Klingon installed by default.

>IF this went through would it REALLY be that easy to get the fonts?
>

If the browser downloaded the whole font, perhaps clever type designers might respond by adding 'data slugs' to the font file that would make them larger. For example, maybe you could add lots of text to the License field in the font to make it 2MB. That probably wouldn't cause the font to use any more resources in terms of memory, but a pirate would have to go to the trouble of modifying it so it would work as part of a website.

This would be like converting iTunes DRM files to MP3, you can do it, but most people don't.

Foundries would know immediately they see one of their fonts, which is not licensed for internet downloading, being used in the "linked" manner proposed. They can do a search, and prosecute. This would discourage unscrupulous web designers.

However, there would be market pull from legit web designers who want to pay to license "linkable" fonts, and once they use a font, pirate web designers would have easy access to it.

Perhaps if there was some way of putting a security code into a linkable font, so that it would only work from a specific, licensed URL...

The following is a quote from David Woolley:

"That's the effect of putting a TTF font on the web! Anyone can download it and use it in any web site. Also, unless you apply a same origin policy, which you haven't requested, other sites don't even need to host the font, they can simply deep link the font resource from another site."

ChrisL

Browsers could also be designed to cache fonts, so that a page would display a face correctly if you've ever viewed a page that used the font. I don't know if, say Firefox developers would promise not to do that, since it would be handy. Of course, that would be very bad from a piracy standpoint, unless the fonts were licensed for that type of use.

I think if, as the W3C intends, this is limited to freeware fonts then it may be of some limited use to a subset of web designers. But it does little for professional and corporate web designers who would like to use non-freeware fonts, and it provides no revenue opportunities for type designers.

But realistically will web designers limit themselves to freeware fonts? I don't think so, and I think the W3C will say that misuse of the system they promote is not their problem.

PS. Carl, you're CITE'ing again - all i see is italics. Don't make me install firefox. ;-)

sii, if you see these type of cite problems, you can always contact one of the admins.

I think if, as the W3C intends, this is limited to freeware fonts then it may be of some limited use to a subset of Web designers.

this is just silly. i think i'd rather see comic sans everywhere than see crappy free fonts everywhere.

Here's a survey I forwarded to the W3C which I felt might be useful in gathering the opinions of type designer's on font embedding.

Font embedding on the Web.

The W3C is again looking at technologies for delivering fonts with Web content. We’re trying to assess attitudes of those in the type design community.

1. What best describes your involvement in type design?

A. I’m an active type designer producing fonts for commercial licensing and distribution, deriving some income from type design.
B. I’m an active type designer producing fonts largely for my use in my own projects or for fun and derive no direct income from my type design activities.
C. I’m involved in the type design business (eg. foundry owner, reseller or tools maker) but am not actively designing type myself.
D. I’m a consumer of typefaces and am not involved in their creation.

2. If you’ve followed issues related to typography on the Web why do you think font embedding technologies such as Microsoft WEFT (EOT) or Bitstream TrueDoc (PFR) failed to excite the Web design community? Pick one or more…

A. Lack of a cross-platform and cross-browser solution.
B. System fonts are more readable and legible than most other fonts that could be embedded.
C. It’s too much hassle creating embedded fonts vs. making static bitmap text in a image editor like Photoshop.
D. Flash and PDF offer a better solution than HTML plus embedded fonts.
E. The fact that some type designers and foundries campaigned against Web font embedding or changed their licenses to prohibit embedding fonts using these technologies.
F. The extra bloat embedded fonts add to Web pages.
G. The fact that these technologies are proprietary and not open.
H. Other. Specify…

3. As a font maker do you currently support (through your standard font license, embedding permissions, or FAQ) embedding of your fonts in Web content?

A. Yes, we allow embedding and make no restriction regarding technology or Web use.
B. Yes, we allow embedding in Web content but only when the mechanism is considered secure.
C. No, we do not allow embedding of fonts within Web content.
D. We don’t have a policy on this.
E. Other. Specify…

4. As a font maker what are your biggest issues or concerns with embedding fonts in Web content? Pick one or more…

A. Extraction and reuse of embedded fonts.
B. Little likelihood of licensing fonts to users for use with this technology.
C. Customer support issues.
D. That people will just ignore license clauses that deal with Web font embedding.
E. Other. Specify…

5. If a new Web font embedding technology were proposed what would your priorities around it be? Pick one or more…

A. That it be an open standard allowing anyone to follow a spec to make tools that create embedded font objects and browsers and other software that enables the extraction of usable fonts from the embedded font objects.
B. That it be a closed, licensed system, that would help reduce (although not eliminate) the extraction of usable fonts from embedded font data.
C. That it require mandatory font subsetting, removal of hints, kerning and other data, to make extracted fonts less useful.
D. That it require a new opt-in embedding permissions setting that identifies fonts licensed for embedding within Web content.
E. That it only support digitally signed fonts making it less likely that an unauthorized person has changed the font’s embedding permissions.
F. Other. Specify…

6. If a new Web font embedding technology were proposed that allowed public domain, freeware or open source fonts to be zipped up and attached to Web pages how would you feel? Pick one or more…

A. That’s fine with me, I’d support that.
B. I’d be concerned that commercial fonts would be mistakenly used in this way.
C. I think the system would be deliberately abused and commercial fonts would be deliberately posted.
D. Don’t care.
E. Other. Specify…

7. If the system identified in 6, required that fonts be set to “installable embedding” how would you feel?

A. Fine with me, I’d support that.
B. I think the system would be deliberately abused by people flipping the bits in commercial fonts or by people renaming commercial fonts and setting them to installable embedding.
C. Don’t care.
D. Other. Specify…

8. Any other ideas, comments or suggestions? Specify…

Si, maybe we could create a poll... I mean, if you are interested.

Si,
I think this would be a valuable tool. Would W3C post it or would Typophile post it and send the results to w3c? The latter might get more traffic. What do you think?

ChrisL

Thanks, It did go through a few drafts with a few of the usual suspects helping me out. :-) So I think it's ready for prime time. However, I think interpreting responses will be tricky - even if it's not posted as a poll it seems to cover the main talking points.

I'm happy for this to be posted anywhere, but I think somewhere frequented by type designers would be the best place - so typophile would be a great.

Most of the posts in that thread on the W3C site are errant nonsense. The people posting understand neither fonts nor the needs of web designers.

Being able to use only any free or shareware font will NOT meet the needs or desires of web designers.

A proposal that allows the web to be used to seamlessly distribute fonts without any controls will result in a dramatic increase in font piracy to a level beyond our wildest nightmares.

Regards,

T

>this is just silly. i think i’d rather see comic sans everywhere than see crappy free fonts everywhere.

We actually paid an independent third party to analyze a large collection of freeware fonts on a bunch of measurable criteria including char set coverage, embedding permissions, (c) and TM strings, hinting errors etc., we didn't look at aesthetics. We found that few fonts passed all of the tests. We passed these results to the W3C.

I know the third party continued extended analysis on their own dime so I’ll see if they are willing to post these results.

I personally think that a web downloading system will be worthless unless all or most foundaries will allow their fonts to work with it. And I can't see that happening if it means the font that I buy and use on my web site winds up becoming installed on the computer of every person who visits my site.

The way I see it, the browser has to download the font, and render it, without ever feeding it into the operating system of the user. For one thing, if I visit a page with four fonts I don't have, and I am forced to download and install them, then I would balk at the delay.

The only solution I see is that the font is handled within the browser in a manner that keeps it off the user machine. No doubt this will not be a perfect solution to piracy of the fonts, since one could build a pseudo-browser that would do nothing other than harvest fonts off websites. But this wouldn't be much worse than the current piracy where fonts are stolen under the current "sharing" methods.

Of course, this means browsers would have to be greatly improved, but with current computing capabilities this should be possible, if things like ATM worked on 286 computers.

All foundries will never climb on board. We already have enough problem with all foundries allowing PDF embedding. (I want to cry.)

Expecting users to pay for fonts they're viewing (as someone suggested earlier in this thread) is a bit ludicrous. The font licensor is a totally different group from the "end user" who views the fonts on a site. You don't expect magazine subscribers to pay for the fonts they view. You don't expect viewers of a TV ad to pay for the music in the ad.

Simon, Thomas, I'm glad you both (via the companies you represent) are weighing in with the W3C.

When a type designer/foundry licenses a face to a publisher to produce a brochure, they do not give the font away with the brochure. The website would be giving the font away with the pageview. If you can get the font to view and print only that web page, fine, go for it. If the font then becomes available for the reader to apply to text of his own, forget it.

ChrisL

For some reason my comment to w3c list never made it to post. I got an email response saying it had been received and asked my permission to publish it (which I gave). I then got a message saying it was sent for review but that was 2 days ago. Who knows.

ChrisL

None of the solutions that involve the browser and OS will work for anyone. It would be illegal for people to use it.

Analogies to other industries, other media, are red herrings. We are having to reinvent things for the internet. Fonts are software. In print they are no longer software, easy. On the computer, they remain in their original state. We need to think of this as a new problem, not a re-staging of an old story.

PHP right now can render a (rasterized) image using any font you want. That's how a lot of type testers work. I don't see why a similar technology can't be used to make fonts work more generally on the web. It's secure since it's server side, making us fonties happy and it gives the web designer what he/she wants. I could see this just being a back-end function, something like settype (fontfilelocation, size, style); and it could even be used with CSS somehow. I know this could be done, it would just take work. They could resolve many issues, but it has to be a function of back-end programming, not user and OS and browser based.

Now how this would relate to a EULA is beyond me, perhaps a license for web use specifically.

Aaron.

Aaron is correct, if online publications were made up of static bitmap pages then you could draw a comparison, but as they’re not I think comparing the W3C proposals to print publications use and licensing of fonts is a red herring.

Onto the EULA issue - obviously most font vendors allow embedding of fonts. “Embedding” is reasonably well understood as a principal, and is explicitly defined in many EULA’s as well as in the TrueType and OpenType specs. It’s the encapsulation of a font (or derivative thereof) in a document file – PDF, .DOC, .PPT etc.,

What the W3C is talking about is not embedding – it’s the linking of loose or zipped up font files to Web pages which are downloaded and temporarily installed to render content. This is not embedding, so the embedding provisions of the specs and EULAs can not apply.

This is not “embedding”, this is “redistribution.” The EULAs attached to commercial fonts do not allow redistribution. Hakon and co. realize that this means that most fonts will not be legally usable under this scheme. However, they point to the large number of freeware fonts out there as being usable under the scheme, arguing that although not perfect they are better than nothing.

I think the bottom line is as follows...

• For font makers the fear has to be that if adopted and implemented web designers will not limit themselves to legitimate freeware fonts.

I am surprised at the small number of responses here on this important topic. I would have thought more people would be interested. :-(

ChrisL

I think Si is right in his assesment of the 2 pertinant issues.

ChrisL

"so the embedding provisions of the specs and EULAs can not apply."

Why is that? I mean, a font to be linked to AND downloaded, AND installed, something could look at the embedding bit somewhere along the line,,,, even if this is not embedding.. . . . . . . . no?

In the case of an imbedded PDF, as long as the security protection is in place at the time of creation, use of the font is limited to printing or reading that document. With the current web scenario, the font would be distributed to the reader of that page and be available for use in any way. That is not embedding, that is donation. Not something I would like to see.

ChrisL

>Why is that? I mean, a font to be linked to AND downloaded, AND installed, something could look at the embedding bit somewhere along the line,,,, even if this is not embedding.. . . . . … no?

...because it's not embedded within the document file or tied to it in any way.

Hakon has suggested* that if a font is set to 'installable' embedding then the font should be considered freeware. Not sure that's totally in line with the spec, and it's certainly open to abuse, but he's free to make that argument.

Embedding permissions ...

http://www.microsoft.com/typography/otspec/os2.htm#fst

Cheers, Si

*In a W3C-Style posting made yesterday by Håkon Wium Lie...

"I would argue that, while the embedding bits predate the web, the
permission for level 4 fonts to be "permanently installed on the
remote system" [1] is granted even without the font file being
"physically embedded" into a document. It's easy enough to create a
dummy document which "embeds" the font -- I could even argue that a
zip file provides a level of embedding. So, if a TrueType level 4 font
family is being served on the web, I think it's reasonable to use it
to display documents. (I wouldn't install it permanenty on the system,
though, but treat it as a cached resource.)"

> I am surprised at the small number of responses here on this important topic.

I am too. I deliberately didn't take an alarmist tone, but I think that there's a good chance something like this will go through unless the voices of type designers are heard. I think when it does become part of CSS it will be adopted by browser makers quite quickly considering their stated support of Web standards.

Most of the discussion I see on this topic is about the technology of how to do it, not should it be done. Not that there is by any means a parallel in effect but this makes me think how physicists must have talked about how thermonuclear devices could be built during the 1940s. It was exciting science and technology and therefore had to be talked about.
I sure would like to see more views on the "should we" rather than the "how could we" of this.

ChrisL

Maybe we should make a crosspost in the "What are you Listening to?" thread where over 118 people have responded?

Would that be too much "Playing the fiddle while Rome burns"? :-)

ChrisL

I am surprised at the small number of responses here on this important topic.

Why wouldn't the W3C do serious outreach, consulting with stakeholders, and seek input and approval from an organization representing the interests of font developers?

There is no "AIGA" for type designers, but surely something as far reaching as this merits, for instance, a special committee of ATypI?!

This could also be a conference topic at TypeCon in August.

I'm not familiar with the workings, but if Microsoft, Adobe et al can get together with the Unicode consortium to develop the OpenType format, there must be some kind of informal protocol to build on.

It would be reckless of the W3C to proceed unilaterally with something that could devastate an entire industry, and then say "nobody objected to our plans", based on response in online forums.

Si, tell these guys not to rush.

Do we have until August?

ChrisL