Font copyright law changes coming to U.S.

124

Comments

  • 1) What André said. My reading of the copyright office rejection notices is that they are saying they will NOT accept TTX output any more, that they do not consider it programming code. If so, they probably wouldn't accept our code either.

    2) Stuart, we've already integrated a source code editor into FontLab VI. You can create glyphs through text entry and/or drawing tools. Same effect.

    Our format is not TTX output, but if the Copyright Office were to still accept TTX (XML), I expect they would accept our VFJ (JSON) code as well.
  • I'm curious what percent of the code needs to be hand coded in order for the affidavit to be true, and how is all this measured, recorded, and proven regardless?
  • @Thomas Phinney as per my recent conversation with Frank Martinez, TTX Output continues to be fully accepted when an affidavit is signed. The new requirement is the affidavit to accompany the submission.

    @André G. Isaak you're correct on many levels and indeed I'm not sure my workflow would change dramatically if such a window were added to my font editor. That said, if the UI environment included both a graphical and a text editor that worked harmoniously (e.g. draw a circle and the X,Y code shows up in the text window or type code in the text window and a circle appears in the UI) then it solves the issue in the eyes of the Copyright office as to if the font was hand coded.

    To be clear, today, hand coding TTX using a text based code editor isn't believable to any party and I would agree 100% of font designers don't work this way. That said, it becomes more believable if a true option presented itself to demonstrate a hybrid text/UI based font editor existed if only for the purpose of being able to continue to be able to submit fonts for copyright under this new requirement.

    And, it's entirely likely some folks may work with the code window in a hybrid fashion as they develop their typeface as a side effect.

    This I feel is the ONLY viable path forward as opposed to lobbying the US Congress to remedy this issue and nobody has the time or money to fund a lobby for the Font Designers Local 402 to get fonts copyright law fixed for our purposes. Of course I'm being a little cheeky with my reference, but it's hard to agree that the efforts to build such a tool that nobody is asking for or may use frequently has merit other than for the purposes of being able to copyright fonts in the US.
  • That said, if the UI environment included both a graphical and a text editor that worked harmoniously (e.g. draw a circle and the X,Y code shows up in the text window or type code in the text window and a circle appears in the UI) then it solves the issue in the eyes of the Copyright office as to if the font was hand coded.

    ...
    if a true option presented itself to demonstrate a hybrid text/UI based font editor existed if only for the purpose of being able to continue to be able to submit fonts for copyright under this new requirement.

    And, it's entirely likely some folks may work with the code window in a hybrid fashion as they develop their typeface as a side effect.
    That's pretty much exactly what we did in FontLab 6.1. We intend to make a preview version available soon (as in, this month).

    It has uses beyond resolving copyright issues.
  • Cory MaylettCory Maylett Posts: 75
    edited September 20
    My understanding has always been that the copyright office considers letterforms to be utilitarian objects, which, to them, means they're not eligible for copyright protection. I'm assuming this does not apply to dingbat fonts since they're not letterforms and are purely small pieces of original artwork compiled into a font. Does anyone have any insight into whether or not this is the case?
    If so, could the inclusion of a few original dingbat glyphs into a font make at least those glyphs in the font subject to copyright protection as original artwork? Someone with a font editor could, of course, remove those copyrighted glyphs and claim that the remainder of the font had no copyright protection. Even so, the copyright office might still approve a font for copyrighted status if it includes original, non-utilitarian, non-letterform artwork.
  • @Cory Maylett you are correct that the alphabet and forms cannot be protected or patented in the US as art or similar. Outside the US, in most countries, they actually can be protected as such.

    The only thing that can be copyright protected as a font is the source code (aka software) which is treated the same way a novel is treated in the eye of the copyright office regardless if it contains dingbats only, partial dingbats or no dingbats. The end-run you propose is a non-starter in the eyes of the copyright office.

    Otherwise, the font name can be trademarked but that only protects the name of the font.
  • Cory MaylettCory Maylett Posts: 75
    edited September 21
    Just to clarify (not meant to be argumentative), the design of a utilitarian object (like a spoon, a nail, a bottle or a letter, numeral or punctuation mark) can't, according to the copyright office, be copyrighted — only patented.
    Drawings of plants, horses, insects, etc., can, of course, be copyrighted since the drawings are original art objects and not, by definition, utilitarian.
    It doesn't seem as though the copyright office objections are about fonts themselves, but rather about (1) the utilitarian nature of the letterforms and numerals fonts contain and (2) the code making up the font not being an original copyrightable program — just a digital description of utilitarian letterforms.
    I hesitate to attempt a logical argument here, since the copyright office arguments seem selectively illogical to begin with. But since dingbat drawings are decorative art objects, wouldn't dingbat glyphs fall outside the definition of them being utilitarian?
    In other words, draw a horse or a duck or a wombat in, say, Adobe Illustrator and those drawing are copyrightable as original artwork. Take those same exact vector-drawn illustrations and insert them into a font and logic would seem to dictate that those drawings are still the same copyrightable, non-utilitarian pieces of art that they were before — they're just being stored inside a software container called a font instead of in another software container called, for example, an EPS, .ai, PDF or .psd file.
    Again, I'm not suggesting that logical arguments are necessarily meaningful to the copyright office. I'm not even suggesting that it's a viable approach, let alone any kind of solution. I'm mostly just tossing out a somewhat harebrained thought in response to what I consider half-baked reasoning for denying copyright protection to original typeface designs. The real solution, I think, is getting the laws changed, but I don't see that happening anytime soon.
  • FWIW, I've never attempted to copyright an illustrator file with images on it so I can't speak to if it'd be protected as source code or patented as 'art' but I get your point that we can agree the copyright office is consistently inconsistent :dizzy:
  • Would a command line that could do things like ‘draw circle x y r’ make then font ‘hand coded’? 
  • Would a command line that could do things like ‘draw circle x y r’ make then font ‘hand coded’? 
    Yes, I believe so - as long as the commands were person authored and not say generated by an autotrace method.
  • So, MetaFont files are more likely to be accepted since the are written as commands, instructions, equations, etc., yes?
  • It seems to be a question of how you create the file. Do you code it, or use visual tools?
  • George ThomasGeorge Thomas Posts: 440
    edited September 23
    If the USPTO is going to require this, then in all fairness they must also require that all software be hand-coded in Assembler. If they are going to be ridiculous, then by all means carry it to extremes.
  • Thomas PhinneyThomas Phinney Posts: 1,109
    edited September 23
    I am not disagreeing with the “daftness” of this distinction. But it is exactly the distinction the USPTO is making.

    The natural way to do this at the software level, at least to us as we have implemented it in FontLab 6.1, makes no distinction between visual input and text input in results, and represents those results in both forms, as well.

    One would hope this could be informative to the PTO.
  • It has come to my attention that a claim of copyright infringement needs to be enforced where the violation happened.  After learning that it hit me like a lightning bolt that all the high profile cases that were brought as copyright were not just by american foundries (which we'd expect) but also against American companies.

    That should have stood out as weird.  65% of our customers are outside the US and I'd guess that's common among us foundries.  In case you're thinking American companies are just worse at license compliance, a quick look at our past infringement settlements tells me otherwise (disproportionation not american in comparison to over all sales).  

    In practical terms, given the international nature of our customers, that means that copyright really is a very is a very weak weapon.  Legal recourse is cost prohibitive enough without having to go to another country AND I can tell you for certain that several of our customers were incentivised to settle specifically because they would have had to defend themselves in America.  

    Additionally, I would point out that trademark protections help in a large percentage of cases since infringers rarely change the name of the font. 

    This post may be more cold comfort from the wet blanket but it made me feel better - that we are not completely screwed.
  • Stuart SandlerStuart Sandler Posts: 229
    edited September 24
    @Georg Seifert I don't believe the command line tool would solve this since it is effectively a 100% text based editor and nobody could sincerely create fonts in code only without some visual feedback.

    Let me suggest another approach if incorporating a code window into an existing font editor is too complex.

    Perhaps a visual TTX code viewer/editor be created that loads and saves only TTX files but can import existing OTF and TTF files and allows the user to select snippets of code from the TTX table names via a dropdown menu and allows the user to 'edit' this code.

    I've attached a crude 'minimum viable product' example of how I'd imagine this tool may appear.

    Basically the visual output is on the left side, the TTX table selector in the upper right and the code window in the lower right.

    It would of course also need the ability to create a new TTX file and save/save as the TTX file.

    Lastly, the only tricky part is that it'd need some way to visually show 'kerns' or 'features' in the left window as appropriate to the code.

    This would likely satisfy the requirements of the US Copyright Office although it should be noted the user must be able to create a 'New' TTX file and effectively build a font from nothing to prove this can be a legitimate font creation method.
  • Mark SimonsonMark Simonson Posts: 955
    edited September 25
    We can already "create a new TTX file and effectively build a font" with a text editor. The only reason nobody does it is it's too tedious. Adding a live preview would only make it slightly less tedious. Since nobody would use this hypothetical tool anymore than they would use a text editor to create a font from scratch, I don't see how the existence of such a tool would make any difference.

    On the other hand, I actually think I could honestly say that a TTX file, generated from a font I created, was written by me, albeit indirectly. But even writing "directly" in a text editor has an element of indirectness compared to, say, writing by hand. You type on a keyboard, but what you see on the screen is necessarily mediated by application and operating system software, first to put in into memory, and secondly to display it on a screen. You are not directly turning bits of memory on and off individually, although that is the indirect result of your key presses. That this all happens instantly hides the indirectness.

    If I make a change to a font and TTX it, the changes I made will be plain to see in the TTX file, just as if I had made the changes directly in a text editor. There is a one-to-one correspondence between the font I have created and a TTX version of it. Therefore, I think I could honestly claim to be the author of the TTX file. Without my actions, it would not exist.

    So, maybe this is not as big a problem as we think it is. It all comes down to what you think "each line of software code for type face software works must have been written by the author" means. Maybe we are thinking of it too narrowly.

    Why not sign that affidavit?
  • Khaled HosnyKhaled Hosny Posts: 219
    edited September 25
    I think it is worth noting that a TTX file is an XML interpretation of a binary font file, it isn’t a 100% identical representation of it and many parts of the binary file are omitted or recalculated during compilation. De-compiling a font with FontTools then compiling it again will not always produce a bit for bit identical result, and bug in FontTools can even corrupt the compiled font (which is not unheard of).

    So people shouldn’t be to fixated on TTX files; if they can be accepted, then any other non-binary format can be and it is even more plausible for someone to manually write or edit a, say, UFO font or parts of it than a TTX file, since the former provides a much higher level representation of the font. UFO vs TTX is more like C source vs disassembled binary.
  • Mark SimonsonMark Simonson Posts: 955
    edited September 25
    I don't think plausibility of authorship is the issue. The idea that it hinges on whether I literally typed in the TTX file (or UFO, which isn't a single file anyway, or some other text representation) is beside the point. Saying I could have plausibly written this text representation of the font I made seems like a weak argument. That's not how we author fonts and I can't imagine we ever will. I think it's focussing on the wrong end of the issue.

    The Copyright Office says they would accept a TTX file if the font author signs an affidavit saying that the TTX file "was written by the author".

    A printed copy book is not a 100% identical representation of an author's manuscript, yet they are both protected under the same copyright claim. The author of a manuscript is just as much the author of the printed book. The author did not create the printed copy, but it was derived from the manuscript.

    A TTX file is nothing more than a different representation of the font I authored. It is derived from my original work. Therefore, I believe it is not dishonest to attest that I am the author of the TTX file.

    A lot of the discussion here seems to revolve around the assumption that the Copyright Office is requiring this affidavit as a means of blocking registration of fonts. ("You didn't type all this in—gotcha!") On the other hand, maybe they just want an affidavit.
  • @Mark Simonson as per my conversation with Frank, until the US Congress directs the US Copyright Office and since they have 'punted' on how expressly to address fonts, the US Copyright Office has basically said, until further notice, this is what we need.

    So technically, yes, they will accept TTX files accompanied by an affidavit that it was hand coded and it's left to the font designer to determine their comfort in signing such a document.

    My primary suggestion was to more deeply integrate a text editor within existing UI based font creation tools so that technically the probability of such a statement could be believable whereas today it is at best suspect.
  • Mark SimonsonMark Simonson Posts: 955
    edited September 25
    I don't know. Adding a TTX view to a font editor just so you can claim that you might have made the font by typing it in seems to be a sort of smoke-and-mirrors approach to satisfying the affidavit. It seems more defensible to me to assert that I did author the TTX file, regardless of the indirect, automated steps needed to produce it.

    I mean, if an author dictated a novel to a typist, would the author not be able to claim copyright? But he would if he had a typewriter on his desk, because then he might have typed it himself?
  • And we've come full circle . . .

    Again, reading from page one of this thread, your point has already been addressed as we cumulatively assessed the nature of what they will accept and why.

    At present, zero font editors are hybrid (UI/Code Window) so it stands to reason that generated output from a UI font editor isn't the same thing as a font that is developed in such an editor, your point wouldn't satisfy such an affidavit if I'm solely focused on what the Copyright Office wrote in their letter to Frank.
  • Kent LewKent Lew Posts: 793
    edited October 8
    I’d be curious to find the proceedings that lead to the Vessel Hull Design Protection Act of 1998, which amended Title 17 (the Copyright Code) by adding Chapter 13 to explicitly provide protection for the design of vessel hulls.
    Here’s the first section:
    §1301.
    (a) Designs Protected.—
          (1) In general.—The designer or other owner of an original design of a useful article which makes the article attractive or distinctive in appearance to the purchasing or using public may secure the protection provided by this chapter upon complying with and subject to this chapter.
          (2) Vessel hulls.—The design of a vessel hull, including a plug or mold, is subject to protection under this chapter, notwithstanding section 1302(4).

    (b) Definitions.—For the purpose of this chapter, the following terms have the following meanings:
          (1) A design is “original” if it is the result of the designer’s creative endeavor that provides a distinguishable variation over prior work pertaining to similar articles which is more than merely trivial and has not been copied from another source.
          (2) A “useful article” is a vessel hull, including a plug or mold, which in normal use has an intrinsic utilitarian function that is not merely to portray the appearance of the article or to convey information. An article which normally is part of a useful article shall be deemed to be a useful article.
          (3) A “vessel” is a craft—
                (A) that is designed and capable of independently steering a course on or through water through its own means of propulsion; and
                (B) that is designed and capable of carrying and transporting one or more passengers.
          (4) A “hull” is the frame or body of a vessel, including the deck of a vessel, exclusive of masts, sails, yards, and rigging.
          (5) A “plug” means a device or model used to make a mold for the purpose of exact duplication, regardless of whether the device or model has an intrinsic utilitarian function that is not only to portray the appearance of the product or to convey information.
          (6) A “mold” means a matrix or form in which a substance for material is used, regardless of whether the matrix or form has an intrinsic utilitarian function that is not only to portray the appearance of the product or to convey information.
    This always seemed to me the most potentially useful precedent for an “industrial” art receiving copyright protection in the U.S.
  • Thomas PhinneyThomas Phinney Posts: 1,109
    I don't know. Adding a TTX view to a font editor just so you can claim that you might have made the font by typing it in seems to be a sort of smoke-and-mirrors approach to satisfying the affidavit. It seems more defensible to me to assert that I did author the TTX file [...]
    I agree entirely! That is why the feature we added to FontLab VI was done in a way that actually promotes editing and is useful. Used it just yesterday to debug and delete something I had done wrong with an imported vector graphic.
    [...] regardless of the indirect, automated steps needed to produce it.
    I am not at all convinced the copyright office considers that the same thing. One of the reasons (though far from the main one) for having the feature in FontLab is to give people a tool to help show why the distinction that the copyright office seems to want to make here, between visual authoring of code and typing of code, is absurd. If done properly, there is no difference between the results, no way to distinguish, these different approaches to authoring font code.
  • John SavardJohn Savard Posts: 192
    Kent Lew said:
    I’d be curious to find the proceedings that lead to the Vessel Hull Design Protection Act of 1998, which amended Title 17 (the Copyright Code) by adding Chapter 13 to explicitly provide protection for the design of vessel hulls.
    The thing is, typefaces aren't vital for national security. So it's unlikely to be useful as a precedent.
  • Kent Lew said:
    I’d be curious to find the proceedings that lead to the Vessel Hull Design Protection Act of 1998, which amended Title 17 (the Copyright Code) by adding Chapter 13 to explicitly provide protection for the design of vessel hulls.
    The thing is, typefaces aren't vital for national security. So it's unlikely to be useful as a precedent.
    Or are they? http://www.cs.columbia.edu/cg/fontcode/
  • Cory MaylettCory Maylett Posts: 75
    edited October 11
    Kent Lew said:
    I’d be curious to find the proceedings that lead to the Vessel Hull Design Protection Act of 1998, which amended Title 17 (the Copyright Code) by adding Chapter 13 to explicitly provide protection for the design of vessel hulls.
    One might need to look no further than the $39 billion dollar per year U.S. boating industry and the lobbying group supporting it.
    https://www.nmma.org/press/article/21978

Sign In or Register to comment.