…the phrase “derivative work” is coming from copyright law,
I wasn’t intending to talk about intellectual property in a legal sense, when I meant it quite literally as “an idea that belongs to a designer” and how easily such things may now be subject to mimickry or unintentional simulation—not what courts may decide is permitted if and when an alleged derivation is tried.
Two thoughts occur immediately in response to this:
1) Are all ideas ownable? If not, who determines whether a particular idea can be owned? Insofar as that question demands an answer, I wonder whether there is any discussion of intellectual property possible that is not also a discussion of the legal aparatus that exists precisely in response to that question.
2) Are uninstantiated ideas ownable? If not, is there any sense in which intellectual property exists other than as instantiated in made things? That being the case, doesn’t it also make sense that what is owned must be reckoned in the nature of the thing made, in which case both the USCO analogue that a piece of software is intellectual property in the same way that a novel is and your notion that a static typeface design is more robustly intellectual property than a dynamic typeface design seem to me equally incoherent, as neither takes proper account of the nature of the things made.
I wonder whether there is any discussion of intellectual property possible that is not also a discussion of the legal aparatus that exists precisely in response to that question.
A discussion of ease of copying need not address whether copying or exclusive ownership is legal.
Are uninstantiated ideas ownable?
I’m not sure I understand your line of reasoning here. But I would say yes. For instance, if I were to produce a Univers font of “1” weight (presently, there is nothing lighter than “2”), would it not still be ownable by Monotype and not me? Or do you mean that nobody can own something that hasn’t been made?
@Nick Shinn I am just objecting to your choice to use the phrase “derivative work” which has a legal meaning in copyright law, to mean something else. It seems to me to be a Bad Idea, on multiple levels.
Call it something that has no technical meaning in this specific area. Spin-off or offshoot have legal meanings, but only (as far as I know) in corporate organization terminology, so one of those might be a fine phrase.
In Nick’s defense: I’m the only one who used the phrase “derivative work.” I still believe it applies to variable font instances, even though it has not been tested yet legally, nor am I in a hurry to test anything legally.
A discussion of ease of copying need not address whether copying or exclusive ownership is legal.
A discussion about ease of copying is not necessarily a discussion of intellectual property. As soon as it becomes a discussion of intellectual property, I think the legal aspect forces itself upon the discussion because property is overwhelmingly a legal concept in modern society.
But I would say yes. For instance, if I were to produce a Univers font of “1” weight (presently, there is nothing lighter than “2”), would it not still be ownable by Monotype and not me? Or do you mean that nobody can own something that hasn’t been made?
The latter. If you were to make a hitherto non-existent weight of Univers, that would be the instantiation of that idea. You spoke of intellectual property in terms of “an idea that belongs to a designer”, and my point is that the idea itself is nobody’s property, but the instantiation of the idea is where intellectual property subsists. This is why I think it is incoherent to try to frame intellectual property independent of the nature of the made thing in which it subsists, e.g. treating software as a form of literature for copyright purposes.
The latter. If you were to make a hitherto non-existent weight of Univers, that would be the instantiation of that idea. You spoke of intellectual property in terms of “an idea that belongs to a designer”, and my point is that the idea itself is nobody’s property, but the instantiation of the idea is where intellectual property subsists. This is why I think it is incoherent to try to frame intellectual property independent of the nature of the made thing in which it subsists, e.g. treating software as a form of literature for copyright purposes.
I find what you are saying here to be confusing. It sounds as though you mean something which, for obvious reasons, can't possibly be true.
It is true that if I simply think about a lighter weight of Univers than was ever made available as a font, no violation of intellectual property rights has taken place. It only takes place when I create a font.
But what you said sounds like a claim that a font in a lighter weight of Univers than was ever made could be made and sold legally by anyone (at least as long as they don't call it Univers and violate trademark rights) because only instantiated ideas, i.e. fonts, can be protected from copying - not the "idea" of Univers.
That would be ridiculous, because if there's a Univers 50 and a Univers 60, and we are in a jurisdictiion that protects font designs, if anyone could make and sell Univers 51 or Univers 59, then this protection would become meaningless.
So the idea you're trying to express is probably complicated enough that it will be difficult to get it across clearly.
Obviously, since mind reading is not possible, an idea must become manifest outside of someone's head in some physical embodiment before it can be copied in violation of patents, copyrights, or whatever.
But what is stolen in an intellectual property violation is not the physical embodiment itself. If someone steals books from a bookstore, the law of theft is violated, not the law of copyright.
So intellectual property rights protect information about the physical embodiment, which is data, or an 'idea', rather than any material object.
Furthermore, the information is at a certain level of abstraction. Copyright law allows an author to control not only verbatim copies of a book, but also paraphrases, the Reader's Digest condensed version, and unauthorized translations into other languages of that book.
So if you're not contradicting that, your point is not clear.
Another way to misinterpret what you've written would be to claim that a typeface, heretofore only made in lead type, if copied in digital form, or on film for phototypesetting, isn't protected, because the nature of the made thing into which it is copied is completely different. Again, that wouldn't make sense.
But on the other hand, all the embodiments of type as type end up producing text that is ultimately read by the human eye, whether on paper or on a screen. (Well, there is optical character recognition, but let's not get too picky.) So the physical embodiments are all linked by one thing in common.
This philosophical debate could go on a while, so I’d like to try and explain my theory with an illustration.
Three type genres which may be interpolated between, with axes of contrast and serif thickness. They correspond to the traditional categories of Didone, Clarendon, and Geometric Slab.
The personality of each is quite distinct historically and graphically, being defined as a100·b100, a50·b50, and a0·b0, where a and b are the two axes.
What kind of a typeface is a parametric font that can create all three of these different genres? It is a shape-shifter, the opposite of what I mean by robust.
(Thomas, ‘derivative’ is a very useful, because precise, term when it comes to discussing design. That legal discourse makes use of the term too, in its own way, is good to know but that fact cannot mean that the word cannot be used any more in non-legal discourse. The other terms you offer just do not have the exact meaning of ‘derivative’, or give it one or another twist, so do not help much.)
Translations are derivative works, and digital revivals of metal types are a form of translation. As such, they are derived from the existing metal type, which has certain characteristics, and also constitute new works that include some of those characteristics that transcend the original medium—which is what makes translation between technologies possible—but may also include new characteristics that are particular to the digital format.
Translations, as derivative works, must be authorised by the original copyright holder; but depending on the jurisdiction may obtain their own copyright under the name of the translator. It makes sense to me that something similar should obtain in the case of typeface design revivals, extensions, modifications, etc.: they are derivative, so require authorisation, but may also contain novel work in themselves, including characteristics that are particular to new technology.
But what you said sounds like a claim that a font in a lighter weight of Univers than was ever made could be made and sold legally by anyone (at least as long as they don't call it Univers and violate trademark rights) because only instantiated ideas, i.e. fonts, can be protected from copying - not the "idea" of Univers.
The ‘idea’ of Univers is only protected in that it subsists in the typeface family Univers. I certainly did not intend to suggest that this means any slight variation of that design is outside of that protection, because, as you note, things like paraphrases can be considered derivative works, not only exact copies. My point, contra the possible implication of Nick’s identification of intellectual property with “an idea that belongs to a designer”, is that uninstantiated ideas are not property. It is not a corollary, however, that every instantiated idea constitutes intellectual property (this varies by jurisdiction, but most place limits on kinds of works and kinds of novelty that may constitute intellectual property, e.g. some jurisdictions extend copyright to typeface design and some do not), nor that newly instantiated ideas may not be derivative of existing intellectual property.
I’ve often had ideas for typefaces and then discovered that either someone else has already made them or someone else comes along and makes them subsequently. My imagination is neither a derivative work based on the types I have never seen, nor something that can be protected as intellectual property against being made by someone else. The idea is not property until it is instantiated. After that, it becomes subject to intellectual property law that is a) specific to jurisdictions and b) includes notions of derivative works that include various kinds of derivation, not just direct and exact copying. And as I argued, and this exchange seems to confirm, as soon as the discussion becomes about intellectual property, it becomes a discussion about law, because that is how our societies frame property.
@John Hudson there is the creation of derivative works from source material which is what you're referring to and then there is the generation of derivative works from font software which is what @Thomas Phinney is referring to.
In following the thread closely, any possible output derived from font software is a derivative work which includes raster images, postscript outlines.
Accordingly, variable fonts ARE font software which is simply capable of generating considerably more width/weight output possibilities, all of which would still be considered derivative works under IP law so I think we can breath a sign of relief.
I was off-track in mentioning intellectual property in a legal thread, when what I was really concerned with was identity.
The consensus of this discussion is that a font with an identity that is polymorphic, schizophrenic and flexible, with regards to typeface classification, is no less robust intellectual property than any static font which it is capable of closely imitating.
There is a point at which parametric fonts cease to be typefaces, becoming in effect megafamilies with a comprehensive list of tags and keywords.
All the more reason to criticize those who call typefaces fonts!
There is a point at which parametric fonts cease to be typefaces, becoming in effect megafamilies with a comprehensive list of tags and keywords.
I charactersise variable fonts as a packaging format for generative type systems. In that respect, I would agree with you if you are saying that what is instantiated in a variable font is not a typeface, but then I already use the term typeface to refer to different members of a type family. Typeface designs are one of the things that go into the making of a variable font, but what is also being designed is the generative system that will instantiate those typefaces and also all the typefaces between those typefaces and the way in which they are generated. I’m okay calling the design of the system a new aspect of typeface design, since the processes are not really separable, even though the outcome is something more than a typeface design, or even a type family design.
I already use the term typeface to refer to different members of a type family.
There is also an “uninstantiated” kind of typeface, a Platonic ideal which exists in the ether of the collective conscious. We all know what “Bodoni” or “Garamond” look like and can draw those from memory (more or less).
what is also being designed is the generative system
There may be a few kinks in the middle of the axes, but “generative system” is rather a grand term for something that is no different than how we already make a family of static fonts as interpolations.
Erik van Blokland’s Superpolator was multidimensional, in that respect, before the variable format was introduced.
The variable format adds nothing to the type designer’s conception of a typeface or type family, that was not previously possible as “Blend” in Altsys Fontographer 30 years ago, before even the Multiple Master format was implemented. However, I don’t think I would have come up with the idea for Beaufort in 1996, as a two-axis typeface family, if the MM format had not already existed. Well perhaps, inspired by Univers (1964).
Nick, you need to take a closer technical look at OT font variations, which are based on Apple’s GX delta model, not on MM-like interpolation. The key difference is that the space between the axes is extrapolated, meaning that you can generate designs in parts of the design space where no ‘masters’ exist. You can also distort the design space and have non-linear axes, and with the recent avar v2 spec you can create virtual axes that manipulate other axes. I think the term ‘generative system’ is simply an accurate description of what one can design and build with the technology. Sure, there are a lot of pretty simple OT variable fonts that don’t do much more than one could do with MM-like interpolation and old-fashioned blend functions in font tool, but the capabilities are a lot more extensive.
John, you need to un-contradict yourself about what goes on in uninstantiated design space. First you stated that it is not legally the responsibility of the typeface designer, now you say that the typeface designer is designing a “generative system” which determines it.
Once the masters are designed (with a single axis, at least, which is the main use of variable fonts, for weight), there is no difference in the “design space” between static and variable fonts. In both one seeks compatibility of outlines at the poles. What difference is there between a system that has eight static weights, generated by interpolation, and a variable system that covers the same range? The system is the same whether there are eight or eight hundred instances. In both, one’s design addresses how interpolation will take place between the poles of the axis, with perhaps a kink or two along the way.
I’m not comparing OT variation with MM.
The extrapolated space between design axes was, as I mentioned, in Superpolator previously, used to make static fonts.
Non-linear design axes were a feature of Gill Sans and Futura in the 1930s, with their topologically distinct weights.
However, I’m beginning to detect a certain quality in variable designs that have novel axes. It’s hard to say just what it is, but I think it might be a sensibility that emerges when type designers look at “slider animation” of a variable font and gain a visual awareness of the axis as a whole.
John, you need to un-contradict yourself about what goes on in uninstantiated design space. First you stated that it is not legally the responsibility of the typeface designer
Where do you think I said that? I’m not even sure what you mean by ‘uninstantiated design space’.
now you say that the typeface designer is designing a “generative system” which determines it.
Correct: the way the OT variations design space is designed and implemented determines what happens within the design space, which includes both interpolation and extrapolation. I call the system generative, because depending on one’s approach, one can generate large areas of the design space between axes without explicitly controlling the corners. Yes, you are right that this is similar to what was possible with Superpolator, which I would also classify as generative: there is a good reason why Erik was part of the OTvar working group, and why the .designspace file format is an integral part of many OTvar build processes. And just as with Superpolator, the fact that some of the generate instances of the design space are not made by the designer in the same way that masters are made does not mean that they are not designed, because the design is in the generative system itself.
“…the instantiation of the idea is where intellectual property subsists.”
Generating design space does not instantiate glyphs. Therefore, if a font varies between Condensed and Extended, nothing in between is intellectual property. Although it has been designed—that is the paradox I’m calling to your attention.
…the design is in the generative system itself.
That may be true for multiple axis fonts, but isn’t it a stretch to call a single axis font a generative system? In both static and variable fonts, one’s design addresses how interpolation will take place between the poles of the axis. Does this mean that designing a static family with interpolation is also generative design?
Would you say that class kerning is also generative system design?
Nick, trying to conflate my comments about uninstantiated ideas with something like unnamed or unmastered locations in a variable design space not only doesn’t make any sense, it doesn’t look like good faith discussion.
A variable font is an instantiated idea—or collection of ideas—the fact that it exists is what makes it instantiated, unlike an idea that no one has given form outside their mind.
Any specific unnamed or unmastered locations in a variable design space may be uninstantiated in a technical sense—may even remain so forever if no person or system chances to use that location—, but it is still part of the instantiation of the idea(s) that subsists in the variable font as a thing.
That may be true for multiple axis fonts, but isn’t it a stretch to call a single axis font a generative system?
Perhaps, but I was talking about the characteristics of the technology, not particular implementations, which I acknowledge may be simple and use only a tiny portion of the potential of the format. A simple generative system is still a generative system, even if what it can generate is pretty limited.
In both static and variable fonts, one’s design addresses how interpolation will take place between the poles of the axis. Does this mean that designing a static family with interpolation is also generative design?
I didn’t use the phrase ‘generative design’ anywhere in this discussion, but I would say that yes, interpolation as a design tool—or, indeed, any number of other design tools that programmatically affect shapes according to instructions or settings determined by the designer—are examples of generative design. But if these processes are only producing static fonts, then what the designer is making and putting in the hands of users are not generative systems. The process is not the product.
Would you say that class kerning is also generative system design?
Well, it’s more generative than fixed-size subsets of non-class pairs were, which is why we use it: it provides a way to kern more, more efficiently, and to extend kerning from pairs that you know exist to pairs that might exist or that actually don’t exist.
Small generative systems permeate digital technologies, simply because writing a rule to produce a superset of outcomes is often more efficient than creating a specific subset of outcomes. GPOS mark positioning is another example: a combining candrabindu may never occur on top of a Q in any writing system, but having a common above mark anchor that generates the display of this combination is much more efficient than trying to identify all the combinations of Q plus above mark that do occur and create them as precomposed diacritic glyphs. Q̐!
Any specific unnamed or unmastered locations in a variable design space may be uninstantiated in a technical sense—may even remain so forever if no person or system chances to use that location—, but it is still part of the instantiation of the idea(s) that subsists in the variable font as a thing.
Which, to me in my ignorant simplicity, sounds like it means that all typeface designers must hasten to offer variable fonts for sale, in order to ensure that their intellectual property rights (if, of course, they have any) extend to additional weights of their typefaces.
Of course, you have already clarified that this notion is mistaken in your previous reply to a post of mine.
But I re-iterate this point, although it's already been addressed, because I think that the reason the discussion between you and Nick Shinn is unproductive is that you are discussing these issues from an angle that most people haven't thought of, or aren't interested in discussing (perhaps because they see it of little practical importance) and therefore assume you are discussing the aspect of these issues that matters to them, thus taking your words in the wrong way.
So it seems to me that you need to find, or even invent, language that makes it clear what you actually intend to talk about, so that your point can be appreciated.
From your reply to me, I can see one part of the picture. To me, what you said in that reply seems to mean:
"When I say that a typeface idea needs to be instantiated to be eligible for intellectual property protection, what I mean is that before the first Univers font was out there, if someone else came up with Univers independently, there would be no way to claim he stole it, because there was nothing out there to steal. I did not mean that intellectual property protection is exactly limited to those things which have a physical embodiment - once one embodiment of the idea of a typeface is visible, of course related things which still include that idea, like intermediate weights, are also potentially able to be protected."
All right, that's great, but that seems to mean that it's such an obvious truism that there can be no genuine disagreement between you and Nick Shinn. Nobody is claiming font piracy by mind reading is possible.
Since the discussion is continuing, presumably there must be something substantive at issue.
It occurs to me that designing a typeface is ispo facto generative design, in as much as the design space is all the different combinations of characters that may be made in (or by) text, which one can’t possibly see and assess during the design process.
Digitization elucidates the methodologies of the analogue world—I noticed that when translating the “joining rules” of calligraphy into <calt> coding for a handwriting font, they seemed much more logical as such, compared with how I had been taught.
Nick, trying to conflate my comments about uninstantiated ideas with something like unnamed or unmastered locations in a variable design space not only doesn’t make any sense, it doesn’t look like good faith discussion.
They have similarities which puzzled me, I was interested to find out how you would describe the difference, which you have done with your usual erudition, thank you!
It occurs to me that designing a typeface is ispo facto generative design, in as much as the design space is all the different combinations of characters that may be made in (or by) text, which one can’t possibly see and assess during the design process.
Again, I have not used the term ‘generative design’, and it doesn’t seem to me a usefully meaningful term except perhaps in terms of particular tools and processes. But I would say that moveable type is inherently a generative technology, for the reasons you state and in the manner that all modular systems are generative: it is designed to be able to generate instantiations of arbitrary texts from a set of prefabricated elements.
To John Savard’s point, Hudson and I are splitting hairs, but that is a very typographic pursuit.
As we all attempt to come to terms with new technology, new perspectives emerge. I doubt I would have come to the realization that, as Mr Hudson puts it “moveable type is inherently a generative technology” without this discussion, and I think that helps put the new stuff (and the old stuff) in context.
I need to know just what it is that new tech adds, that previous generations of tech could not do, rather than just adopt the latest round of “upgrades” in the tech rat race willy-nilly.
To John Savard’s point, Hudson and I are splitting hairs, but that is a very typographic pursuit.
As we all attempt to come to terms with new technology, new perspectives emerge. I doubt I would have come to the realization that, as Mr Hudson puts it “moveable type is inherently a generative technology” without this discussion, and I think that helps put the new stuff (and the old stuff) in context.
I need to know just what it is that new tech adds, that previous generations of tech could not do, rather than just adopt the latest round of “upgrades” in the tech rat race willy-nilly.
No doubt it is because I am not a type designer that my eyes have tended to glaze over at such a discusson on a rarefied plane.
To me, it doesn't seem meaningful to look for the gestalt of variable fonts.
What they do, so far, in their most common and straightforward application, is let one choose the precise degree to which a font has been made bold. Since some fonts have demibold weights, perhaps this does have some value in a typographic application; not being a professional graphic artist either, I hadn't really been all that sure of that.
Of course, what I do know is that Douglas Hofstadter was spot-on in his criticism of an offhand remark by Donald Knuth.
So let us all turn to the chapter on Fluid Concepts and Creative Analogies in our copies of Metamagical Themas, and we shall find all the answers laid out before us...
EDIT: Actually, chapter 13, titled Metafont, Metamathematics and Metaphysics: Comments on Donald Knuth's Article "The Concept of a Meta-Font". Previously published in the August 1982 issue of Scientific American.
Perhaps it's not quite that simple. But he did cover a lot of ground on the larger questions associated with parametric fonts.
So, Laatz sued Zazzle. Zazzle made a counterclaim. Laatz made a motion to dismiss the counterclaim. The court has denied Laatz’ motion to dismiss. This court decision has a several interesting elements, worth reading for those who are trying to keep up on the status of font copyright.
Comments
1) Are all ideas ownable? If not, who determines whether a particular idea can be owned? Insofar as that question demands an answer, I wonder whether there is any discussion of intellectual property possible that is not also a discussion of the legal aparatus that exists precisely in response to that question.
2) Are uninstantiated ideas ownable? If not, is there any sense in which intellectual property exists other than as instantiated in made things? That being the case, doesn’t it also make sense that what is owned must be reckoned in the nature of the thing made, in which case both the USCO analogue that a piece of software is intellectual property in the same way that a novel is and your notion that a static typeface design is more robustly intellectual property than a dynamic typeface design seem to me equally incoherent, as neither takes proper account of the nature of the things made.
But I would say yes. For instance, if I were to produce a Univers font of “1” weight (presently, there is nothing lighter than “2”), would it not still be ownable by Monotype and not me? Or do you mean that nobody can own something that hasn’t been made?
Call it something that has no technical meaning in this specific area. Spin-off or offshoot have legal meanings, but only (as far as I know) in corporate organization terminology, so one of those might be a fine phrase.
Three type genres which may be interpolated between, with axes of contrast and serif thickness. They correspond to the traditional categories of Didone, Clarendon, and Geometric Slab.
The personality of each is quite distinct historically and graphically, being defined as
a100·b100, a50·b50, and a0·b0, where a and b are the two axes.
What kind of a typeface is a parametric font that can create all three of these different genres? It is a shape-shifter, the opposite of what I mean by robust.
Translations are derivative works, and digital revivals of metal types are a form of translation. As such, they are derived from the existing metal type, which has certain characteristics, and also constitute new works that include some of those characteristics that transcend the original medium—which is what makes translation between technologies possible—but may also include new characteristics that are particular to the digital format.
Translations, as derivative works, must be authorised by the original copyright holder; but depending on the jurisdiction may obtain their own copyright under the name of the translator. It makes sense to me that something similar should obtain in the case of typeface design revivals, extensions, modifications, etc.: they are derivative, so require authorisation, but may also contain novel work in themselves, including characteristics that are particular to new technology.
I’ve often had ideas for typefaces and then discovered that either someone else has already made them or someone else comes along and makes them subsequently. My imagination is neither a derivative work based on the types I have never seen, nor something that can be protected as intellectual property against being made by someone else. The idea is not property until it is instantiated. After that, it becomes subject to intellectual property law that is a) specific to jurisdictions and b) includes notions of derivative works that include various kinds of derivation, not just direct and exact copying. And as I argued, and this exchange seems to confirm, as soon as the discussion becomes about intellectual property, it becomes a discussion about law, because that is how our societies frame property.
In following the thread closely, any possible output derived from font software is a derivative work which includes raster images, postscript outlines.
Accordingly, variable fonts ARE font software which is simply capable of generating considerably more width/weight output possibilities, all of which would still be considered derivative works under IP law so I think we can breath a sign of relief.
The consensus of this discussion is that a font with an identity that is polymorphic, schizophrenic and flexible, with regards to typeface classification, is no less robust intellectual property than any static font which it is capable of closely imitating.
There is a point at which parametric fonts cease to be typefaces, becoming in effect megafamilies with a comprehensive list of tags and keywords.
All the more reason to criticize those who call typefaces fonts!
Erik van Blokland’s Superpolator was multidimensional, in that respect, before the variable format was introduced.
The variable format adds nothing to the type designer’s conception of a typeface or type family, that was not previously possible as “Blend” in Altsys Fontographer 30 years ago, before even the Multiple Master format was implemented. However, I don’t think I would have come up with the idea for Beaufort in 1996, as a two-axis typeface family, if the MM format had not already existed. Well perhaps, inspired by Univers (1964).
Once the masters are designed (with a single axis, at least, which is the main use of variable fonts, for weight), there is no difference in the “design space” between static and variable fonts. In both one seeks compatibility of outlines at the poles. What difference is there between a system that has eight static weights, generated by interpolation, and a variable system that covers the same range? The system is the same whether there are eight or eight hundred instances. In both, one’s design addresses how interpolation will take place between the poles of the axis, with perhaps a kink or two along the way.
I’m not comparing OT variation with MM.
The extrapolated space between design axes was, as I mentioned, in Superpolator previously, used to make static fonts.
Non-linear design axes were a feature of Gill Sans and Futura in the 1930s, with their topologically distinct weights.
However, I’m beginning to detect a certain quality in variable designs that have novel axes. It’s hard to say just what it is, but I think it might be a sensibility that emerges when type designers look at “slider animation” of a variable font and gain a visual awareness of the axis as a whole.
Generating design space does not instantiate glyphs.
Therefore, if a font varies between Condensed and Extended, nothing in between is intellectual property. Although it has been designed—that is the paradox I’m calling to your attention.
Would you say that class kerning is also generative system design?
A variable font is an instantiated idea—or collection of ideas—the fact that it exists is what makes it instantiated, unlike an idea that no one has given form outside their mind.
Any specific unnamed or unmastered locations in a variable design space may be uninstantiated in a technical sense—may even remain so forever if no person or system chances to use that location—, but it is still part of the instantiation of the idea(s) that subsists in the variable font as a thing.
Small generative systems permeate digital technologies, simply because writing a rule to produce a superset of outcomes is often more efficient than creating a specific subset of outcomes. GPOS mark positioning is another example: a combining candrabindu may never occur on top of a Q in any writing system, but having a common above mark anchor that generates the display of this combination is much more efficient than trying to identify all the combinations of Q plus above mark that do occur and create them as precomposed diacritic glyphs. Q̐!
Digitization elucidates the methodologies of the analogue world—I noticed that when translating the “joining rules” of calligraphy into <calt> coding for a handwriting font, they seemed much more logical as such, compared with how I had been taught.
They have similarities which puzzled me, I was interested to find out how you would describe the difference, which you have done with your usual erudition, thank you!
As we all attempt to come to terms with new technology, new perspectives emerge. I doubt I would have come to the realization that, as Mr Hudson puts it “moveable type is inherently a generative technology” without this discussion, and I think that helps put the new stuff (and the old stuff) in context.
I need to know just what it is that new tech adds, that previous generations of tech could not do, rather than just adopt the latest round of “upgrades” in the tech rat race willy-nilly.
https://law.justia.com/cases/federal/district-courts/california/candce/5:2022cv04844/399632/176/
Also, last I saw, if this were to make it to trial, that trial was scheduled for July 14, 2025.
... reading the PDF, well, what is this Twombly 2007 case against Bell?
And, interesting elements, indeed. 🧨