Web fonts, Free fonts, and modification requirements

Thomas Phinney's picture

There's something I'm trying to wrap my head around, and not quite succeeding.

Many "free fonts" either:

1) Do not permit modifications, or
2) Require one to change the name if the font is modified

How does this interact with web font services like TypeKit and Kernest, which offer such fonts? Do we choose to believe that an EOT wrapper is not a modification? What about subsetting?

I'll note that fonts embedded in PDF are usually subsetted, but most fonts actually used in the world do not permit modifications in their license terms.

Comments welcome. I'm especially interested in the opinions of people who are licensing "free fonts" under such conditions.

T

abattis's picture

My understanding from conversations on the OFLB mailing list with Nicolas Spalinger, co-author of the SIL OFL, is that EOT conversion is seen a 'modified version of the file' and therefore requires renaming; a subsetted font file is also seen as a modified version.

Thomas Phinney's picture

Funny thing: the viewer never sees the name of the font in this situation, no? Even the web author doesn't really see/care much, except for the name they use in their CSS/HTML, which may be outside the control of the folks doing the font serving anyway.

Thanks,

T

Richard Fink's picture

@thomasphinney and abattis:

TP, when I first read your post my feeling was, "More talk about how many angels can dance on the head of a pin."
Licenses. Legal bull. Sick of it.

But now I'm trying to wrap my head around the same things.
No choice. Last night I was reading the SIL OFL.
Thanks for the info, abattis.

With Typekit, I have no way of knowing if their EULA agreement is the same as is being offered to me when I buy the font. Right?
But anyway, I think that the bottom line for Typekit, Kernest - or anybody for that matter - is the "Am I In The Clear?" question. Does it seem like the intent of the creators is that I can do the particular thing I'm looking to do without a take-down notice arriving at my ISP. I've got enough problems, thank you. And I want to do the right thing by the font's creators and anybody who downloads it and plays around with it. If it's a sub-set, it should say so.

Is it unwise to say that, as a rule of thumb, if you are modifying the original file in any way you should, at least, modify the name and/or make a notation in the version string?

Don't know. Anybody?
(And then of course, making a notation that you've modified the file is, in itself, a modification. I would speculate that at least 400 angels can fit on the head of a pin. Maybe as many as 500. :-)

rich

dberlow's picture

>Funny thing: the viewer never sees the name of the font in this situation, no?

Correct.

>Even the web author doesn’t really see/care much, except for the name they use in their CSS/HTML,

Which is a file name, not a font name.

>...which may be outside the control of the folks doing the font serving anyway.

Not sure about that...;)

Cheers!

Richard Fink's picture

@TP

BTW - This intersects with the turn taken on the thread Different Fonts Named The Same: Does This Happen Frequently?
Reserved names and the requirement to change names for derivative works ensures that, going forward, we're going to see a lot of renaming.

Regards, rich

Richard Fink's picture

>...which may be outside the control of the folks doing the font serving anyway.

TP is referring, I think, to the way @font-face works in CSS. The font file is given a font-family name, an alias of sorts, which doesn't need to reference anything within the font file.
The web author can choose any font-family alias for the file the service company is providing. Kernest, for example, suggests a default font-family alias, but you're not bound to it.

rich

Richard Fink's picture

Oops, I forgot the usual disclaimer: I am not a theologian.

Thomas Phinney's picture

I agree that it's a pretty specific and detailed question, but for folks like TypeKit and Kernest, who are running these services and trying to operate within the bounds of the font licenses in question, it might actually matter. Even if it does seem to go into angel/pinhead territory.

Cheers,

T

Richard Fink's picture

@TP

>it might actually matter
Yes, of course it does. If and when you try to enforce the provisions. If Typekit or Kernest are violating EULA's the licensors do have remedies they can pursue. A EULA is a private contract and enforcement is, and rightfully should be, a private matter.
You can try until the end of time to explain to average folks that the software product they are buying is not the product itself, it's only the license to use it, subject to a list of restrictions, and they will still not understand that because it is so counter-intuitive. I, being quite average, have trouble understanding it on an emotional level, myself. If I can't use the software in a manner that makes sense to me, as defined by me, why are you offering to sell it to me?
In patent law there is the concept of patent exhaustion - once you've paid money for a patented thing, there is an implied license and you can do what you want with it. Perfectly intuitive. Once I pay money for something, I own it. Done.

I'm being flip because EULA's can be quite comical. The iTunes EULA says, in capital letters, too, that the software:
IS NOT INTENDED FOR USE IN THE OPERATION OF NUCLEAR FACILITIES
Try wrapping your head around that one. iTunes, this is.
There's so damned much convolution I don't think there's any possibility of unscrambling it without Federal legislation. A EULA is a contract and a contract can say anything you want. And it's a one-way negotiation.

Regards, Rich

Christopher Slye's picture

I'm not really following this. Typekit or whatever similar service still has to sign an agreement with the owner of the "free font," don't they? Well, let me put it this way: If the free font license doesn't allow something that a font subscription service does, then the font subscription service has to make an agreement with the font owner before serving the font, otherwise they are very simply violating the license.

I don't know what Typekit or Kernest or anyone else is doing, legal-wise, but I think it would be wise to get explicit permission from a font owner to serve their font, regardless of how permissive their standard license seems to be.

Syndicate content Syndicate content