Safari release version supporting downloadable web fonts...

[aluminum]

"Sadly our confidentially promise with those folks won’t let me counter your theory"

oooh! Time to start rumours! ;o)

[aluminum]

"Sii, adding a winking emoticon at the end of that statement does nothing to diminish the insulting, discriminatory and divisive nature of it"

Sure it does. ;o)

Ray was likely mentioned as he is often mentioned as the example by the Hakon Lie (one of the 'inventors' of CSS/employed by Opera)

[Si_Daniels]

>Ray was likely mentioned as he is often mentioned as the example by the Hakon Lie

Correct.

>Moderators—-will one of you please have a word with Simon offline about this. Thankyou.

If the moderators think I'm out of line they can call me on it right here. If Ray or any of the people I've personally attacked on typophile can contact me off-list if they like.

[Si_Daniels]

>Don’t assume that if Arbo isn’t posting much he isn’t reading typophile.

"Arbo" ;-) You know, you can't pick your own "nickname", the community has to pick one for you. ;-)

[Florian Hardwig]

You know, you can’t pick your own “nickname”

Why, would that be arbotrary?

[billy.wq]

I'll preface this by saying that I would hardly consider myself a type designer and much less a lawyer;
But I was under the impression, at least in the U.S, that a typeface is not something that can be held under copyright - that is the letter form itself. That what is copyrightable is the "software package" of it's distribution - which could also include metrics and kerning.
Now if a web page is just displaying the letters of a typeface on it's page - without metrics or a packaged means of distribution, then how can this be considered infringement?

[aluminum]

billy...for the browser to display a typeface embedded, then it, indeed, has to load the 'software package' to render it.

This would be opposed to uploading an image made with the type...like a JPG or GIF, which would not include the typeface 'software'. Nor would it be type at that point.

[Si_Daniels]

Billy if you follow the link you'll see a snippet of CSS code... the font file, in this particular example one of Ray Larabie's excellent fonts, is linked and will be downloaded by the browser...

[Si_Daniels]

Ralf>If anyone thinks the same I would be very interested in sharing ideas.

I think that should probably be a focus of discussion.

Ralf>How would web fonts be licensed? Per user? Per domain? Per domain per year? On demand (per download from the foundry’s server)? At which prices?

I really don't have any ideas, all current licenses that allow for traditional embedding, with broad redistribution, assume some level of misuse, but what do you do if the fonts are fully exposed? You had a technique to hide the fonts, which introduces a barrier to misuse. Does this work with Safari’s implementation?

Ralf> How can commercial fonts be protected (which means: how can we make it a litter harder to get them)?

As you probably know we’ve submitted the EOT format to the W3C. That introduces a further barrier, but is certainly not impenetrable. I wonder if there are tables that can be removed from the fonts to make them less usable if extracted?

[Miss Tiffany]

Larabie fonts are also mentioned in the original article to which Si linked. Si is one of the least mean people I know and I say this as an associate of his as well as a moderator.

The simple truth is that if the web becomes a sea of people using the fonts currently available for use it will be fugly. True there are appropriate uses for many free fonts. There are some very good quality free fonts as well. But, if we are speaking in general terms I fear for what the web will become.

[dezcom]

"I fear for what the web will become."

Just think of what happened when Apple and Microsoft unleashed software that could include clip art as well as the clipart and bunches of fonts with limited uses a few years back. Soon, we were getting god-awful emails and wordprocessing files from wannabe "artists" showing us every font and piece of clip art at ther disposal! Not a pretty sight.

ChrisL

[Si_Daniels]

Thanks Tiffany,

The comment I made about what this would do to the reputation of freeware type designers associated with this effort (either by choice or not) was an honest one. I can see things going in a number of directions for them, they could become heroes of the new font-enabled Web, or they could find themselves unjustly blamed for the fugly – I’ll coin the phrase “Vinnification” :-( - Comic Sans is still a very popular font with the masses, even though through no fault of his own, Vinnie gets flack for it. Equally until Firefox and IE jump on the bandwagon nothing may happen. Likewise if mainstream commercial font vendors can find a way to be involved then it won’t be an issue.

The timing, with respect to the Business of Type event couldn’t be better, rather than focus purely on the issues of enforcement and risk we can explore ways for font makers to benefit from this technology.

In addition the "PS" comment about Ray’s fonts having their place, was an honest effort to counter any perception that I was anti-Ray in my earlier post. I know various Microsoft games groups use his fonts, and I was lucky enough to be able to license one of his fonts (through an agent) for redistribution in Halo 3.

[typodermic]

Thanks very much for defending me, James. I really appreciate it.

Freeware fonts are more likely to be used in lowbrow design hence I'm used to being a typographical punching bag. The more widely fonts are distributed, the more likely it is that they'll be misused. Young font designers, take heed: don't release freeware fonts unless you don't mind a future career as whipping boy/girl. Even if you make great fonts, you'll be remembered for fonts misused. I guess Vincent Connare knows how that goes. Praise for Magpie will always be overshadowed by anti-Comic Sans sentiment.

About the ugly fonts:

The first font in the example at the beginning of this thread is Primer Apples which was created for a German teacher at no charge based on drawings she had sent me. The dashed lines are supposed to look casual and have square ends. I don't know how such a font could be made much more beautiful but in the right context I think it looks fine. If Primer Apples was the only headline font in the above example and the theme was school, sewing or scrapbooking it would look alright.

Prime Minister of Canada isn't so bad. I looks a little "autotraced" but I think it looks inky and appropriate at about 12 points in print. In the context you've shown, it's been tracked wide and seems to have a fake bold applied to it. In proper use, it comes off as 1960s style casual pen & ink lettering.

Sexsmith is just cute. The early version was pretty weak but I've seen it used in beautiful ways over the years. In a context where cuteness is required, Sexsmith does its job well. Unfortunately, many freeware font sites still carry the old version.

Vahika was created as a by-product of a FontLab tutorial and yeah, it pretty much blows goats. You'd never guess that someone would use it for paragraph text but there it is.

Many creators of web font technology have contacted me over the years. True Font Family is the best one I've seen. It works right now with current browser technology and anyone capable of basic HTML/CSS coding can handle it. It works right now with no plugins. I don't know if it would be suitable for text but it seems to work well in headlines. There's an iGoogle app available as well.

[Si_Daniels]

Thanks for chipping in Ray, and please accept my apologies for any unintended offense.

One question, does the W3C/Safari mechanism meet with your approval? If so what's the feeling about readme.txt files being separated from raw fonts posted on the Web? I would have thought that would be a major problem for many freeware and open-source font designers?

Cheers, Si

[Ralf H.]

But, if we are speaking in general terms I fear for what the web will become.

I don't. The web already has some nasty features: Animated GIFs, the Marquee tag, the blink tag, ...
But just because they are there, doesn't mean they will be used all the time. At least not on professional websites. They are used on a Harry Potter fan site of a teenager. And this kid will be very happy to set the text in the Harry Potter free font. But that doesn't mean it will be used on Typophile. ;-)

You had a technique to hide the fonts, which introduces a barrier to misuse. Does this work with Safari’s implementation?

It does. In my example there is no public URL for the font file itself. The font is only delivered (from the foundry website) after an authenticated request from an external (i.e. the client's) website. So no one can just look into the soure code and download the font. But in the end, the full font file is delivered to every browser. So of course it IS possible to get the font. And this will always be the case. Whether the font will be extracted from a browser cache, from the RAM or from a EOT-file. There is no bullet-proof protection. We all know this sort of discussion from the music industry, so we can learn from their mistakes. I think we shouldn't embrace the possibilities in this new and huge market as soon as possible. There are millions of websites out there and I am pretty sure that web designers will not hesitate to buy web licenses. We just need to figure out the right models and conditions.

[jasonc]

>>I wonder if there are tables that can be removed from the fonts to make them less usable if extracted?
<<

This is brilliant.
I could imagine browsers making use of fonts without using the OS/2 table (just use the brute force head.Ymax and head.min values to do layout), yet removing that table would make the font all but useless when installed. Depending on how the glyphs are rendered, you might also be able to do a similar trick with the maxp table

[Si_Daniels]

>I think we shouldn’t embrace the possibilities in this new and huge market as soon as possible.

I think you're right, clock is ticking. I bought this to the community's attention at TypeCon Boston, by TypeCon Buffalo we may be too late?

>We just need to figure out the right models and conditions.

Any chance you can make it to Seattle on Apr 4th? ;-)

[Dunwich Type]

We just need to figure out the right models and conditions.

One could always host the fonts on his own server, allow clients to link to those as opposed to hosting them, and charge a per-connection rate. It’s not an easy option for indie foundries who can’t deal with the server stuff, but the big vendors could probably pull it off. And when really high-traffic sites scoff at royalties, they can just be pushed to commission new designs outright…

[Miss Tiffany]

@Typodermic: Yes. I agree. Context is very important when using ANY (freeware or otherwise) FONT.

@Ralf: True. True. Just because bells and whistles are available doesn't mean we need to use and/or see them in our daily surf n' turf.

[jasonc]

>>
There are millions of websites out there and I am pretty sure that web designers will not hesitate to buy web licenses. We just need to figure out the right models and conditions.
<<

Hear hear!
[despite my previous comment, where I was tempted by the technical issue to be solved] I think Ralf's probably right, that the type industry could benefit from the failures of the music industry in DRM. This is the ideal time to look forward at different models of revenue streams (I know, easy for me to say.)
In the end, though, it'd be the Web designer's customers paying for the font license, right? Whoever owns the server would pay. I could see this working, with a few kinks worked out. Perhaps the license could be based on a yearly subscription, or maybe even a per X number of unique visitors model.

Jason C

[Miss Tiffany]

@James: I wondered about that myself. With some foundries re-considering their EULAs to include web use why not just allow usage on the web via linking through the foundries themselves? Or would that be problematic too?

[Si_Daniels]

>Or would that be problematic too?

The foundries already post their fonts on Web servers - you just need to hand over your credit card details to get them. So security on the server is within the realm of possibility. But as Ralf said...

"Whether the font will be extracted from a browser cache, from the RAM or from a EOT-file. There is no bullet-proof protection."

The fonts end up on the unlicensed Web surfing end-users machine, where they can get them with no effort (or trivial effort) and that's where the foundries concerns kick in, and its a this point the concerns need to be overcome with the promise of significant revenue.

[jasonc]

>>
why not just allow usage on the web via linking through the foundries themselves? Or would that be problematic too?
<<

well, there could be complications, certainly. For instance, the web designer and their clients would then be dependent on the foundry's server, so if the foundry's server went down, the client's site wouldn't show the font. Also, the web client's traffic would be "inherited" by the foundry, who's server would use some of it's own bandwidth to provide the font (although I think that's in the noise, presumably the foundry would use some of it's profit to increase their own bandwidth.)

jason C

[Miss Tiffany]

Right. I misunderstood that. :^/

[Ralf H.]

I could imagine browsers making use of fonts without using the OS/2 table (just use the brute force head.Ymax and head.min values to do layout), yet removing that table would make the font all but useless when installed.

I wouln't be suprised if something like this would already work with Apple's implementation.