Font or Font Software

Yves Michel

IfSixWasNine said:

I am more of an alfa person, and this is a question for all the beta's out there.

Sorry for my ignorance, Mr Alfa! What is a beta member of this forum?

My ignorance must make me a gamma.

IfSixWasNine

Yves Michel said:

IfSixWasNine said:

I am more of an alfa person, and this is a question for all the beta's out there.

Sorry for my ignorance, Mr Alfa! What is a beta member of this forum?

My ignorance must make me a gamma.

Hi Yves. Maybe this is something Dutch that got lost in translation. In the Netherlands "alpha's" are people that studied one of the  alpha sciences. The alpha scienes being the study the products of human action. Think of languages, law, art, for example.

The beta sciences deal only with non-human nature. The emphasis is on exact calculations and experiments. Such as physics, information technology, etc.

Christine Bateup

IfSixWasNine said:

Hi Christine, I work as an IP attorney.  It is certainly not that I want to give legal advice on this forum. On the contrary, I am trying to understand on what ground Font software would be protected. This seems to me a rather principled question that should certainly be answered. I had hoped that people on this forum would be able to clarify this. I think it's particular helpful to the type design community to opine on these matters in public forums such as this. If the protection of Font software is so clear, surely there is no problem explaining it. If that protection is not so clear then it seems to me that it is also in everyone's interest to know, Especially now with the rise of AI that may make such font software obsolete in the future. 

I think you may misunderstand me. I'm not suggesting that you are seeking to give legal advice here, I am suggesting that you are seeking legal opinions from non-lawyer type designers, without disclosing your position as an attorney and for unclear motives. As a lawyer who works in this field, I can tell you that my personal view is that the copyrightability of font software (in the United States and elsewhere) is clear if one is focusing on the correct factual issues and examining the correct legal questions. As a lawyer, I'm quite sure you have access to a great deal of legal material, including on this subject, that the type designers here do not. I'd suggest going there to seek the answers to your questions.

Yves Michel

@IfSixWasNine
Hi,
Thanks for this explanation of a curious 'class' naming system.
But I'm quite proud, as a retired physical education teacher, to be alpha. Until now, I was alpha only at home (don't tell my wife!)

IfSixWasNine

Christine Bateup said:

IfSixWasNine said:

Hi Christine, I work as an IP attorney.  It is certainly not that I want to give legal advice on this forum. On the contrary, I am trying to understand on what ground Font software would be protected. This seems to me a rather principled question that should certainly be answered. I had hoped that people on this forum would be able to clarify this. I think it's particular helpful to the type design community to opine on these matters in public forums such as this. If the protection of Font software is so clear, surely there is no problem explaining it. If that protection is not so clear then it seems to me that it is also in everyone's interest to know, Especially now with the rise of AI that may make such font software obsolete in the future. 

I think you may misunderstand me. I'm not suggesting that you are seeking to give legal advice here, I am suggesting that you are seeking legal opinions from non-lawyer type designers, without disclosing your position as an attorney and for unclear motives. As a lawyer who works in this field, I can tell you that my personal view is that the copyrightability of font software (in the United States and elsewhere) is clear if one is focusing on the correct factual issues and examining the correct legal questions. As a lawyer, I'm quite sure you have access to a great deal of legal material, including on this subject, that the type designers here do not. I'd suggest going there to seek the answers to your questions.

Hi Christine, don't really understand what you are trying to blame me for, making it so personal and questioning my motives. I was drawn to this forum because of posts about font software. The whole discussion in this thread is about copyright, font software, whether fonts are software, etc. Many people posting in this forum are licencing specialists and lawyers. I am happy to participate in that discussion. I wanted to contribute to that discussion by simply asking what exactly the IP protection of fonts looks like, where font software is concerned. That question is relevant because it was also raised earlier in this forum that a Polish court ruled in 1997 that font software is not actually software. If you look in my profile (as you did) you can see who I am just fine, and if you want to google me that's fine too. I see that on this forum some people introduce themselves at length, and others don't. I haven't done that, but if that is the mores here then I certainly would have. Just because I work as a lawyer doesn't mean I'm not allowed to ask questions about the background of IP protection of font software, does it? On the contrary, part of my job is precisely to understand that sort of thing. I was hoping to learn something about that on this forum and -if you allow me- have a discussion about it. I think the purpose of this thread is to clarify things and learn from each other. What you are doing is dismissing me from this forum? For apparently asking questions that are uncomfortable for you? Frankly, I find that very curious and it doesn't advance the discussion on IP over font software either. 

John Hudson

Some people may be wary because over the years they have encountered a lot of bad actors seeking to sow confusion or doubt about the protectability of typeface design or font copyright, including representatives of people who have made a business ripping off other peoples’ work, and often in the context of US law where the protection of fonts continues to periodically shift in one direction or another. Personally, I think you are asking the wrong questions, and what should be interrogated is not whether fonts are software within the currrent legal framework of digital code as a form of literature, but whether that framework makes any sense, has ever made any sense, or is anything more than laziness and unwillingness to actually understand the value and nature of digital media on the part of lawyers, judges, civil servants, and legislators. This is not limited to fonts.

John Butler

A font on a computer is a file containing a set of instructions that the CPU follows to draw and position a set of letters (and/or other glyphs and marks) on a screen, printout or other rendering surface, physical or conceptual. By those criteria it constitutes software, no matter what some decrepit half-awake judge thinks who can’t even program his own VCR.

Simon Cozens

Counterpoint: A font on a computer is a database; it is explicitly structured like a database, containing multiple data tables of different structures, with each data table containing multiple records. Fonts are therefore data, not code.

James Puckett

That’s true, but a font generally contains code. For some writing system a font without thousands of lines of OpenType code is useless.

Mark Simonson

The code in fonts isn't Turing complete, though, like a programming language. It's more like macros, where you can automate certain kinds of transformations on the data. Type 3 PostScript fonts were Turing complete since PostScript was and you could include any PostScript commands, etc. 

Thomas Phinney

@Simon Cozens
A database can store pretty much anything. Including software source code. Git is a database, which you and I both use all the time. Its primary purpose, for which it was designed, is to store software source code.

So, even if one were to agree that a font is essentially a particular kind of database, that would not particularly determine whether or not it is software.

John Hudson

I remember this discussion taking place, almost word-for-word, 30 years ago on the comp.fonts Usenet group.

The definition of ‘computer program’ employed by the US Copyright Office* is very broad—encompassing statements and instructions that both directly and indirectly bring about a certain result—while also explicitly noting that not everything in a given program may be subject to copyright. It is a basic principle of copyright that only those parts of a work that can be copyrighted might be copyrighted, and hence a claim of copyright on a work is only ever a claim on those parts that can be copyrighted. So if a font contains copyrightable statements and instructuions that directly or indirectly etc. those may be subject to copyright, and if it contains those functional aspects of a program that are not copyrightable—‘algorithms, formatting, functions, logic, or system design’—then those parts may not be subject to copyright. To ask if a font can be copyrighted as a computer program—the term software is not used by the USCO—is another wrong question. The better question is which parts of a font might be copyrightable and which not?

My understanding is that, at least in the US, there is no definitive case law that answers that question. There is the encouraging summary judgement in the Adobe vs SSi case, which acknowledged that the positioning of outline points and control handles was not arbitrary or algorithmic and interacted with hinting code to produce certain results on screen and in printers, and that this meant the outlines in the Adobe fonts from which SSi made their derivative product may very well be subject to copyright. At which point SSi settled rather than risking seeing that summary judgement confirmed at trial. I think Judge White’s understanding of the technology in that summary judgement was admirable—and would hope other judges would apply the same level of investigative thoroughness and clarity of analysis—, but he is one judge in a relatively low level court, and the case did not go to trial. So, there remains no definitive answer to the question which parts of a font are copyrightable.

In recent years, there seems to have been a resurgence of—or even a doubling-down on—a kind of literalism in the idea of ‘writing’ code, as if the act of sitting down and typing consitutes creation of a literary work whose copyrightability is not at issue. If I am being generous in my sympathy, I can understand the desire to frame the creation of digital works within a familiar old model that the registrars of copyright understand, or think they understand, rather than trying to wrap their heads around new kinds of creativity using new tools and methods. But it is a dead-end for both creativity and law.

_____

*—To give just one jurisdictional example, and let us all remember that there is no universal jurisdiction, law, or treaty regarding copyright that employs a single set of criteria for determining whether a given work is protected or for how long.

John Hudson

@Mark Simonson

The code in fonts isn't Turing complete, though, like a programming language.

According to @Behdad Esfahbod, OpenType Layout GSUB is Turing complete. I have to admit that I did not understand the explanation, even though I was in the room when he explained it for the first time at one of the Unicode conferences, but that’s because my knowledge of programming is limited to what it can do and what the outcome is, not how it does it. I suppose that might also colour my understanding that the creative value in digital works is in what they do, not how they were ‘written’.

jeremy tribby

IfSixWasNine said:

Many people posting in this forum are ... lawyers

that's some accusation 🙂

Peter Constable

James Puckett said:

That’s true, but a font generally contains code. For some writing system a font without thousands of lines of OpenType code is useless.

I assume you're referring specifically to OpenType Layout "code"—the information included in GDEF, GSUB and GPOS tables. None of that information is "code" in the usual sense: none of the individual values in those tables gets interpreted as an operator / operation. It wouldn't be unreasonable to say that it's all declarative data. E.g., for single substitution (format 1) lookup table, the data consists of a list of glyph IDs (a coverage table), and a delta value; there is nothing expressing an operation to be performed. So, @Simon Cozens assertion that this is data and not code is reasonable.

At the same time, while lookup tables contain only such data, each lookup subtable type corresponds to a specific type of action to be performed. So it might be argued that the lookup type and format fields in the data, in a sense, together comprise operators. I still think it's a bit of a stretch to call it "code", though, as the type/format combination still expresses an abstract concept of the action to be take, not the precise operations that get performed on a glyph sequence.

John Hudson

With regard to ‘code’, US copyright registration guidelines for computer programs include a strong preference for source code deposits (compiled object code deposits are permitted, but are subject to the Rule of Doubt, meaning that they are accepted only on the applicant’s assertion that the code contains copyrightable authorship). Note also that this is specific to the process of registering a copyright: it does not imply that the existence of a particular form of source or object code is the basis for the existence of copyright. As the summary judgement in the Adobe vs SSi case indicates, what may consitutute copyrightable authorship in a font is not limited to code in the sense of statements and operators etc. in a programming language. 

Dave Crossland

https://en.wikipedia.org/wiki/Adobe_Systems,_Inc._v._Southern_Software,_Inc.#Functional_concerns

John Hudson

Adobe contended that the outlines only determined some of the control points, and that there was creativity involved in picking the rest. One glyph may be expressed identically in a variety of ways with different numbers of control points.

That’s not a very clear or accurate representation. Better, perhaps, to say that an apparently identical glyph shape can be expressed in a variety of ways.

As I recall, Judge White’s summary also acknowledged that the placing of points affected the rendering, so although the glyph might appear to have the same vector outline shape with different quantities and locations of points, those points are interepreted to indirectly affect a certain result in rendering. This was a point that I and another comp.fonts contributer had made in debates with King.

John Butler

I understand that on the one hand, the opinions of judges and juries are ultimately the only opinions that matter.
I also understand that on the other hand, they’re still wrong. Carve that on my gravestone.

John Hudson

I remember reading a paper by a legal scholar some years ago making the argument that the decision in the Eltra Corp vs Ringer case had been wrong. I like that kind of analytical work because it is outside of legal proceedings while still with the potential to influence future case arguments. Arguments that arise only within adversarial suits tend to be hastily constructed—most cases are on a schedule decided by the judge but determined by external factors—and are often trying to target a particular result—e.g. persuading the other party to settle without trial—rather than establishing a solid, well-reasoned standard that will survive the test of time.

Of course, such analysis only becomes functionally useful if introduced as arguments in subsequent cases, where the weight of precedence needs to be overturned by demonstrating the error of the previous judgement on legal grounds.
_____

Back in the 1990s, I had long conversations with Mike Parker about the Eltra vs Ringer case. Mike had been instrumental in bringing the case, at the recommendation of Deborah Ringer! — a good reminder that once a case goes before a court, the outcome can’t be guaranteed even if both named parties actually agree on what that outcome should be. Mike described the suit as an attempt to make an end-run around Congress, which had blocked the USCO under Ringer from granting copyright to typeface design. [According to Arthur Baker, one of the congressmen involved was the father-in-law of one of the principals of Castcraft, which was notorious at the time for pirating phototype display fonts; indeed, their business had been largely based on copying other peoples’ designs all the way back to the hot metal days in Chicago.]

Dave Crossland

Any chance you can remember the link to the paper? Seems contra the current US Copyright office stance lol

Thomas Phinney

Are you looking for a copy of the decision in Adobe v SSI?

John Hudson

I tried to find it the other day, Dave, but not very intensively. Will think about it some more and see if I can come up with better search terms.

John Hudson

This is not the paper I remember—it is much too recent—, but is along similar lines: ‘An approach to why typography [sic] should by copyrightable’ [direct PDF download link]

JoyceKetterer

I've been looking into this at length because the Zazzle case makes me nervous.  I think there's a chance that software that is clearly marketed globally and released on the internet qualifies for copyright everywhere it is licensed (because it has no one "country of origin").  In that case we'd qualify in France and the Netherlands, which seem to be the two best localities for protections of fonts (I think they might not even make the distinction between fonts and typefaces). 

That said, my understanding as a non-lawyer is that if one pursues a copyright violation in court you have to do it in the location where the violation happened.  AND localities are only obligated by treaty to grant the same level of copyright to you as you would have gotten there (meaning that if there's no copyright in the US for fonts there still wouldn't be copyright for a Dutch or French font inside the US).  But, that might be enough to backstop a claim of violating the EULA if it named French or Dutch governing law.  I'm not sure yet, still researching.