Font copyright law changes coming to U.S.



  • In U.S. law, a registered copyright is considered "prima facie” evidence, which means that it is sufficient to establish fact, unless disproved or rebutted. In cases of font piracy, it would likely be enough to force a settlement prior to a trial, as the defendant would be liable for the kinds of cost coverage and multiple damages that Thomas Phinney describes.

    I am of the opinion that enough has changed forever about fonts and the type business since the Copyright Act of 1976 so that full copyright protection of type design and data should be subject to full copyright protection. But to get this into law will take a lot of doing: from building a clear-cut case (based on a clear explanation of the creative work involved and a demonstration of equivalent work that is so protected) that can be understood by politicians, lawyers, and judges, to building political support. 

    The question I have for all of you is: Who, in your opinion, will line up against such an effort? Who are the corporate interests who would believe themselves to be disadvantaged by the advent of such protections? Are there any credible actors within the type community who would testify against copyright protection?
  • Nick ShinnNick Shinn Posts: 1,404
    How would it be possible to police such a thing? It would just end up with those with the deepest pockets winning law suits. And judges with no type background deciding whether or not Myriad is Frutiger.
  • ^  What Cory said, 100%
  • Kent LewKent Lew Posts: 874
    edited August 2018
    Enforcement may be one thing. But the problem will actually begin in the copyright examiner’s office.
    If earnest and informed type folks can argue about whether Myriad or Segoe are too similar to Frutiger, for instance, then what hope do we have leaving these determinations to a bureaucratic examiner to approve?
    Witness how the Trademark Office refused to grant trademark protection to Matthew Carter’s Postoni (commissioned by the Washington Post) because it was deemed too similar to Adobe’s trademark for Postino.
    (Yes, this was about a trademark; but the issues & challenges will be analogous.)
  • How do you think this will affect DMCA takedowns for copyrighted fonts that wind up on free font/pirate font sites? Will DMCA requests for fonts made in the US be rejected because font copyrights are no longer recognized?
  • If something isn't copyrighted, then a DMCA takedown, which is a copyright-based claim (Digital Millennium Copyright Act), would not apply. So yes, it could have a huge effect.

    Or to put it another way, if copyright isn’t in play, the only way to chase after a web site distributing your fonts is if you can prove the web site violated a license agreement they had with you. But (1) that is an entirely civil matter, and (2) the web site would have to actually be in a contract with you.

    The phrase “good luck with that!” comes to mind.

    (And hence Cory’s concerns of just a few messages back.)
  • But fonts do not do things in the way programs do. Yes?
    But a python script is not doing anything, either. It is read, parsed and then some functions are executed that do things.
    The python script is compiled at run time or "just in time" (interpreted) and executed, not precompiled and directly executed, but howveho many layers of virtualization are involved, it is descriptions of logical processes. 

    OpenType fonts are data, that do not describe logical processes, but are the input to such processes.
  • But the lack of registration doesn't mean there is no copyright.
    No. But registration allows (in the USA) the copyright holder to collect:
    - statutory damages, where infringement can get compensation regardless of actual loss
    - treble damages, where you get triple the actual loss

    Without registration, the copyright holder is limited to actual damages.
    But since previous registration was privisiopal and no one ever took this to court to clarify it, does it matter?

    And, how is loss calculated?

  • @Dave Crossland
    That calculation of loss is part of the legal process. The plaintiff claims some amount, the defendant might counter with a lower number (possibly even zero), and either they will settle for some number in the range from zero to the amount claimed, or if the case goes through a complete trial, the court will make a final determination.

    Speaking as somebody who has been asked where I would be willing make such a calculation for a party in such a case, I can say that my approach would depend on the particulars of the case. The hard cases to come up with a valuation for the infringement, would be those in which the infringement involves a usage, for which the plaintiff does not normally have a set rate.
  • I would love to get some more data here. So...

    So, to summarize, font lawyer (and former US design patent examiner) Frank Martinez has been suggesting that some recent rejections of digital font copyright applications are part of a general trend. He believes that the copyright office has had a policy change and has filed a FOIA (Freedom of Information Act) request to get more info. The Copyright office has responded saying that they are not going to give him the info for free. One way or another he hopes to learn.

    Offline, I have had a couple of our colleagues chatting with me, and a bit skeptical of whether there is a general pattern or trend, suggesting that there could be some more mundane explanation. This despite the fact that one of them has themselves received a couple of recent rejections.

    Personally, I simply have a vested interest in our industry, and want to clarify what the heck is happening. I honestly think both explanations are plausible. The two rejection letters Frank shared had interestingly different rationales, but it is hard to know what that means without more data. (Evolution in copyright office thinking? Related to the content of the applications? Or just obfuscation?)

    So... I would like to simply gather some data. I would appreciate hearing from anyone who has had a response to a font copyright filing from the US copyright office in 2017 or 2018. (Frank thinks the change started in fall of 2017).

    I am willing to keep responses completely anonymous, but also happy to share aggregate data—and even details if you are OK with sharing such details. So feel free to reply off-list, if you like.

    My survey basically asks:
    • what font copyright applications you have had accepted or rejected, including how many and when, since the beginning of 2017?
    • for each, are you also willing to share the rejection letter contents, if there was a rejection?
    • are you willing to have me share your survey details with others? or only as aggregate data?
  • Thanks for this effort Thomas. Will be interesting indeed! 
  • Well, we will see how many responses I actually get. I am not counting my chickens before they hatch, as the saying goes. But even a half dozen responses could be interesting and enlightening.
  • John SavardJohn Savard Posts: 412
    edited September 2018
    A TrueType font, an OpenType font, and so on, encodes characters as a series of curved lines that go from one point to another.

    The fact that one is using a visual tool for specifying those lines, instead of typing the numbers in by hand, should not conflict with the eligibility of the font for protection in the same manner as a computer program. Therefore, this change is a misguided one.

    As well, it is a very real and very serious threat to font designers. Copyright is indeed the first line of defense, not a last resort. It is because the fonts are copyrighted in the first place that the users of those fonts are compelled to enter into a license agreement with their vendor.

    Oh, naturally, you can require people to sign the license agreement before they can download the font, and you can require licensees to maintain security on their computers with your font. If you're going that route, though, you had better be prepared to hire a lot of lawyers, and digitally watermark each individual copy of your font you deliver to a client. Have fun.

    And even that won't protect you from someone autotracing their own vectors from the shapes... so you really don't want a change in the law that gives you no choice but to go there.

    Given how much effort Congress has put into providing more and better copyright protection to the movie and music industries, I can only think that companies like Monotype haven't been making enough campaign donations.

    Oh, and one other point. It is correct that copyrights don't have to be registered to exist and be enforceable.

    However, that applies to works that could have had their copyrights registered, because they were works of a type that were eligible for copyright protection.

    Here, the stated rationale for not accepting registration of fonts not typed in as hand-coded XML is that they do not meet the required standard of originality for a copyrighted work. Not because they are short of staff or don't feel like registering them, but because they lack an attribute essential to having any kind of valid copyright, registered or not.

    Now, of course, that's their interpretation of the law - and they're not the Supreme Court. So courts aren't obligated to find one's unregistered copyright invalid for the same reason that it couldn't be registered. I'm not a lawyer, so I can't say how strong a presumption that this would create.

    But I do think that it is a mistake to assume that this change in policy on registering copyrights will have no effect on the effective copyrightability of fonts in the United States in general.
  • To restate my proposal, @Thomas Phinney and @Georg Seifert, would need you to add a toggleable live text editing window to your respective font development applications that would display the live XML code (same as TTX generates) and gives the user the ability to edit such XML and apply changes in realtime to the font itself.
    This is similar to the model used in VTT, in which hinting can be viewed or edited using the graphical UI or as code (at a couple of different levels in the case of VTT).

    Of course, once you have a visual editor and a text editor that produce identical outcomes, the notion that only handwritten code constitutes a computer program for copyright purposes is revealed as a nonsense. There's no special creative merit in typing.
  • Can one get copyright for a movie? Nobody asks a director to hand type the mp4 data. If you make a animated movie, you drag some handles, apply some properties and the hit export. 
  • I could add a mode where you can hand edit the glyphs data structure. But editing the final ttx data? To find a way to do that that even makes remotely sense??
  • Can one get copyright for a movie? Nobody asks a director to hand type the mp4 data. If you make a animated movie, you drag some handles, apply some properties and the hit export. 

    The problem with this analogy is that Movies, unlike typefaces, have actual copyright protection so no one has to convince the copyright office that a movie is a computer program.

    I’m rather ignorant of US copyright laws, but one thing I wonder is whether webpage design is subject to copyright. HTML much better meets the definition of 'markup language' which the copyright office weirdly applied to XML. And, I suspect, few web-designers code HTML exclusively by hand. If HTML is subject to copyright, then the argument that TTX-generated XML should not be seems rather weak to me.
  • Why not just print the .glyphs text contents?
  • @John Hudson and @Georg Seifert to be perfectly clear, I'm not suggesting adding such a window because it's a great new feature, it simply creates for US font designers a meaningful path to demonstrate that fonts CAN be handcoded even if the hand coding is primarily the data output of manipulating graphical shapes that we can access.

    Ultimately, the desired result is that the ENTIRE TTX code output must be displayed in a single scrollable text editing window that can be edited and any such edits applied to the font.

    That said, could there be a dropdown menu that allows users to edit 'chunks' of TTX data in the text editor window by type? Sure! But again, there must be some sort of view mode that shows ALL the code together.

    Georg, I'm quite sure with your knowledge of your application there are likely slicker and more efficient ways to skin this cat, I'm simply proposing a method that will be acceptable to the US Copyright Office of which an application must be able to allow.

    @Dave Crossland the .glyphs text output isn't useful because it's simply a text output of data, we already can do that with TTX, what we require is a font editor with a text entry window that proves the font can be created with hand code (even via graphical assist)
  • @Thomas Phinney that sounds great but Frank would likely know if the JSON format can be an acceptable font format to file a copyright with. Otherwise you're spot on regarding the source window interactivity.
  • All the formats that have been discussed for representing fonts could be criticized as being “data” formats rather than “application source code” formats. (But of course, things that nobody disagree about being programs can be represented in such formats, so this particular objection by the copyright office seems bizarre to me.)

    One also wonders what the copyright office will say about AFDKO-style GSUB and GPOS code, which seems beyond dispute as to it being software code.

  • One also wonders what the copyright office will say about AFDKO-style GSUB and GPOS code, which seems beyond dispute as to it being software code.

    not so fast ;) 

    Behdad said he thought gpos is Turing Complete, although I didn't see a demo. 

    But if it's not Turing Complete, it's not really software. 

    It may just be markup. 

    Which is what xml almost always is, and what USPTO said won't be able to make it as software.
  • John HudsonJohn Hudson Posts: 1,627
    edited September 2018
    But if it's not Turing Complete, it's not really software. 

    ...what USPTO said won't be able to make it as software.
    Can we be careful with terminology, please? My understanding is that USPTO does not have a definition of 'software'; it has a definition of 'computer program', and as far as I know Turing completeness of the programming language used is not part of that definition.

    Turing completeness means that a language can be turned to any arbitrary task, but non-Turing complete languages exist that have deliberately restricted functionality (e.g. Coq). The USPTO definition of a computer program seems to me easily broad enough to encompass non-Turing complete programming languages.
  • To reiterate, TTX output is accepted code and has been accepted code by the US Copyright office therefore it shall continue to be accepted code and no new code format is necessary nor will it solve the issue at hand.

    What is no longer accepted is TTX output where the designer has not also submitted an affidavit claiming they have 'hand coded' the font software they're submitting.

    Therefore, to satisfy the US Copyright office, we must integrate a text editor into an existing font editor which would allow the entry and editing of said code (TTX output) so that such an affidavit can be true.
Sign In or Register to comment.