Da Sie Deutsch können, lesen Sie hier:

www.sanskritweb.net/forgers/computerprogramm.pdf

Do you have an English version?
Thanks

Well, they're not hardware, so they must be software.
The distinction between program and data is another matter.

But are there any defining properties of font formats other than OpenType that make them software? I presume it must be defined somewhere since most EULAs call fonts software.

As Nick already explained, every computer file can be considered "software".
The distinction between program and data is more tricky. There is no general answer. It depends on the single font in question. A type 1 font will usually just be data, but an OpenType or TrueType font might contain elaborate hinting or OpenType code. Manual hinting was acknowledged as computer program in court before. I don't know of any cases concerning OpenType feature code yet.

Harvard Cyberlaw defines software as "the programs or other 'instructions' that a computer needs to perform specific tasks." In the case of a font, the specific task is to render letterforms of a certain design. Period.

Just a small comment, slightly off topic. In the early days of digital typography, Adobe had a Type 3 format. Type 3 fonts were true PostScript programs. You could do crazy things like erasing files in your directory by just "executing the font". Making parametrized font families was easy, if you knew how to program in PostScript. All the things people like in OpenType today (ligatures, replacement, etc.) could be programmed inside a Type 3 font.

I loved that format.

If you can kick it, it's hardware.
If you can corrupt or overwrite it, it's software.
Otherwise, it's firmware.

Ok I see. So by the same definition an MP3 is software, as is a text file?

oldnick:
Harvard Cyberlaw defines software as "the programs or other 'instructions' that a computer needs to perform specific tasks." In the case of a font, the specific task is to render letterforms of a certain design. Period.

The specific task to render a font is done by the font renderer of an application that takes font data as an input. So only the combination of renderer and font is software. As also wikipedia defines it:

Computer software, or just software, is a collection of computer programs and related data that provide the instructions telling a computer what to do and how to do it. (Wikipedia contributors, "Computer software," Wikipedia, The Free Encyclopedia, http://en.wikipedia.org/w/index.php?title=Computer_software&oldid=412140036 (accessed February 5, 2011).)

A font alone (if there is no opentype scripting or type3 involved) is only data not software. As a tire alone is not a bike.

So by the same definition an MP3 is software, as is a text file?

An MP3 file isn't human-readable on its own, so you might stretch the definition of software to cover it. However, your MP3 player (real or virtual) performs the specific task of interpreting the data in the MP3 file. MOST text files don't induce your computer to perform a specific task, so they're a definite no.

MOST text files don't induce your computer to perform a specific task, so they're a definite no.

Even plain text files are just sequences of bytes that need to be interpreted in a certain way to be human readable.

@oldnick - MP3s aren't really interpreted, since the player always performs the same actions for each file: They are decompressed and then blasted to the sound chip.

@pseudo - A text file can be thought of as software... the shell scripts on my Linux box are pure text, after all, as are the batch files on the Windows machine. I would not think of my fanfic database on its own as software, even though it can be (and has been) corrupted and scrozzled.

MP3s aren't really interpreted

Open an MP3 file in a text editor.

A text editor is an interpreter.

A text editor is an interpreter.

Who said it wasn't? But a text editor does a lousy job of interpreting MP3s...

Hmm... you clearly interpret 'interpreter' far more widely than I do when it comes to computers, since I use that word only about language interpreters, such as PERL, Python and command processors (shells). As I see it, then, if a text editor is an interpreter, then so are your font editors, which would make fonts interpreted code, thus definitely software.

Am I too hackerish for this site?

Even a TrueType font contains algorithmic elements: it may contain a sequence of instructions to carry out hinting.

It's true that the bulk of what a font contains is simply data, point coordinates, which another program reads in order to draw the shape specified for a given glyph.

But in some cases, data can be licensed in the same fashion as computer programs: for example, a company might license you a database of names and addresses - and supply you with it only in an encrypted form that could be read by the program that would print your allowed number of mailing labels.

The important thing is that by recognizing fonts as software, their creators gain an additional element of legal protection. Even in jurisdictions where typeface designs lack legal protection, the font, as a specific binary sequence to be given to a computer, still enjoys full copyright protection. While, even where typeface designs are protected, that protection runs for a much shorter term, similar to that for patents.

So I don't see how treating fonts as software is a bad thing - it seems to me it's a very good thing for typeface designers.

the bulk of what a font contains is simply data, point coordinates

There used to be a nifty little PostScript editor named NOAH (alas, I believe it was 16-bit) which allowed you to view (and edit, if you felt adventurous) the code used to render each glyph. What the human-readable information showed was not merely point coordinates, but also INSTRUCTIONS on how to connect them (line, arc, etc.), If we return to either definition of software listed above, both make it clear that such INSTRUCTIONS are what make a piece of code software.

…the bulk of what a font contains is simply data, point coordinates…

It is indeed "simply" data, because it is written directly as digital media, and that's what makes it akin to code.
It is not data which is a measurement of analog objects or events, or for that matter of digital events.
The font does not exist outside its digital being, so it is not the kind of "factual information about something" which we normally consider data to be. It is the thing itself, sine qua non.
Just as surely, a program is "simply" code.

Why are you asking the question? To clarify your own thoughts?

Pseudo> Ok I see. So by the same definition an MP3 is software, as is a text file?

Pseudo called my here by tweet, party because this topic is in the wrong place, and so I didn't see it otherwise.

Fonts range from pure data (bitmap fonts), to data and software, (hinted truetype gx fonts with opentype tables). So having any single definition of software covering all fonts is not appropriate.

Then you have the US Copyright Office. These hard-working people, and the courts have little interest in getting clogged with the finer details we all know exist, and looking at an evolving situation, they assigned a software type to all digital outline fonts, and defined the requirements for font patents.

What this means is that Harvard, Wikipedia, mp3 and text, or what "the bulk of" a font is, have nothing to do with the definition of fonts as software or not. It comes down instead, to where you live, how much you want to pick on this or that government or foundry or format, and what you're willing to risk.

First of all I will clarify why I asked this question; I am just starting to design typefaces of my own, which hopefully I will sell at some point in the future. As a result of this I have become more and more interested in aspects of the type design other than just the design part, including the retail of fonts. I have been aware for sometime that most foundries sell fonts as "font software", but I could never understand exactly why. It seemed strange to me that they were calling these things software, and I presumed that most end users did not necessarily think of fonts as software.

The only reason I could think of (not knowing the technical details of fonts) as to why foundries would sell fonts as software was for legal reasons, and having more protection of software than just data. If this was the case then it seemed slightly misleading to consumers. So I asked the question here to see if there really was one defining factor of fonts that made them software.

When I asked the question I knew that there was going to be a mixed response, because the definition of software is so difficult anyway. And as we have seen there are differences in opinion as to why fonts are software. I am neither against, nor for the way fonts are called software, I just think there should be a discussion amongst type designers about whether this is the correct, and best way to sell fonts.

From the response here, it seems that the current way fonts are sold as software seems the best model. However, with the introduction of webfonts, the appearance of many new smaller foundries, piracy, and the long predicted end of print, I foresee that the way fonts are sold will probably have to change, but that is another discussion entirely.

I would be interested to know in more detail the history of this subject; when were fonts first sold as software etc. Can anybody can point to some documentation?

pseudo:

> I would be interested to know in more detail the history of this subject; when were fonts first sold as software etc. Can anybody can point to some documentation?

I happen to have on my bookshelf many of the old catalogs and books of the 1980s, e.g. numerous Linotype catalogs and Compugraphic catalogs, and the old URW books, and I also have the old printed Adobe PostScript reference and cookbook manuals of the 1980s and many other similar books of the late 1980s and early 1990s.

If you look through these old books, you will find nowhere a text passage, where fonts are called "software" or "computer programs" etc.

For instance, in the Linotype catalog of 1985, which contained a description of how Linotype digitized its fonts with the URW Ikarus system, Linotype used the technical term "digital font" and described fonts are digital data, but Linotype never said that fonts are "software" or "computer programs".

The same is true for the original Adobe PostScript manuals. For instance, in the Adobe Reference Manual of 1985, page 62, we read this definition by John Warnock:

"font = set of graphic shapes that define the current typeface."

Nowhere the word "software" is used for Type 1 fonts by the Adobe managers Geschke and Warnock, the editors of these PostScript books.

At that time, for Adobe, a Type 1 font was a collection of graphic shapes or graphic outlines of characters (the word glyph came later), but not a computer program.

(Note: An exception were the "Type 3" fonts, which were full-fledged PostScript programs as already stated above by Luc Devroye).

When Adobe started to sell Type 1 fonts, however, the Adobe marketing department decided to cheat dimwitted font buyers by telling them the legal fairy tale that Type 1 fonts are "computer programs" and "software".

A few years later, clever Adobe shysters managed to make Judge Ronald M. Whyte believe that Type 1 fonts are "font software" and "computer programs".

If Judge Whyte, who had been taken for a ride, had read the old Adobe PostScript manuals written by Geschke and Warnock, he would have known better.

See also the Typophile thread www.typophile.com/node/78027

For more detail visit my website www.sanskritweb.net/forgers

In the original PS Language Reference Manual, 3rd printing, May 1987 (1985-1986), we read on page 31 of the printed edition:

"For example, a PostScript font is a dictionary that..."

In the most recent version of this PS Language Reference Manual, downloadable here
www.adobe.com/products/postscript/pdfs/PLRM.pdf we read on page 41

"For example, a PostScript font program contains a dictionary that..."

This means that the Adobe marketing department later replaced the word "font" by the word "font program" in order to make dimwits believe that fonts are programs.

It seems that software "can" be a text file but for me, it must do more than just present the captured text on demand. It makes sense that there are commands given and resources consulted for decision on what is to be presented or executed.
An ascii file of just typed text, like the text of an email, can only present the captured text as it was with the caveat of allowing formatting changes--size, weight, etc.. Even formatting changes require a savvy application which "reads"the text file and allows you to select options.
A typeface file is a set of captured data and selection system that has the intelligence to make itself usable to other software and hardware in many selectable ways. It may not be totally self-enabling and require other players to complete the task requested (InD, MS-Word, OS, Hardware...).
If you think of a text file as a cooked dish, it only allows eating and saving for eating later, without outside intervention. You may then think of a typeface file as a smart recipe book which can call up its ingredients and allow the user to combine them in countless ways. The application software is like the chef, who puts all the ingredients together in the way requested and presents them on demand.
The stove, refrigerator, utensils, etc.. are the hardware tools needed to do the job. The operating system is the software that allows all of the various players to interact with each-other in a selectable and logical way.

Oh, by the way, There Ain't No Such Thing As Plain Text.

>I would be interested to know in more detail the history of this subject; when were fonts first sold as software etc. Can anybody can point to some documentation?

As far as I am aware, there is not a history of SALES of font software in retail fonts for print. Font software is LICENSED there.And in web font "retail", the font software itself isn't usually licensed, the right to the SERVICE of font software is SUBSCRIBED to. There have been plenty of changes in foundry practice Pseudo, to accommodate the changes in the world.These changes can be dimly summarized as some sort of conspiratorial outrage, or one can simply understand that once upon a time fonts were only data and the typography was supplied by human "programs". Now, the typographers are almost all gone, but by plan and legal vision fonts can contain much of the software once provided by typographers, and it is copyrighted as software.

If you are really an interested type designer, I have now donated all you need to know.

>This means that the Adobe marketing department later replaced the word "font" by the word "font program" in order to make dimwits believe that fonts are programs.

Do you have documentation of this marketing department's central involvement?

"DTP" history, Adobe-centric:
http://www.amazon.com/Inside-Publishing-Revolution-Adobe-Story/dp/032111...

Font format history:
http://www.shinntype.com/Writing/BigThing.pdf

Snapshot of the emergent online font business, 2002:
http://www.shinntype.com/Writing/Cruisin.pdf

**

@Berlow: …all you need to know.

David, stop fretting over pixel display and pen the True Story of What Went Down by the Man at the Center of It All, —with lots of sex and drugs, just like Keef.

Well thanks for your comments everybody. I was hoping my post would start a hearty discussion and it did. I would like to reflect on some of the comments in the thread.

If the US copyright office says fonts are software then this only means that under US copyright law fonts are classified as software. To me that does not necessarily mean that they are software. At this point I am really trying to understand, from a purely technological point of view, whether or not they are.

I think it is important to make comparisons to mp3s, movies, even text files etc, because I would say these are the most closely related kind of "business model" there is; consumers paying money online and then downloading some data (which most of the time is only licensed, everything on iTunes and even ebooks from Amazon are only licensed not sold). By understanding a wider range of services we can improve the understanding of our own services.

I do not think that type foundries are trying to trick 'dimwits' into thinking fonts are software. If selling fonts as software gives greater copyright protection then it is understandable that they would currently be licensed like this. However, I do think it is important that the end user is also able to understand exactly what (and why it is software) they are paying for.

My own personal interest in this matter is driven by my inquisitive nature; if I am going to be licensing fonts then I need to understand what exactly it is I am licensing and also the best way to do this. I also think that this might make in interesting thesis for my master's degree (yes I really am an interested type designer!)

Thanks

My concern is not if fonts are software - it seems evident that they are. As with MP3's you can't make digital copies and sell them. However with an MP3 usually you can't take the licensed music it contains, record your own cover of it, and sell that. The "software" has changed, but the song has not. The creator of the song is owed. Not so with fonts (in the US).

So.. is a font like a song? Now that is an interesting question with blurry edges. At first I thought "it's easier to look at a song and see if it's a cover".. but what about a sample, or a single borrowed line, or a chorus? I'd be curious to know how these cases are handled as they are more akin to font - remixing, reviving etc.

Randy:

> So.. is a font like a song? Now that is an interesting question

A good example is the famous song „Still got the Blues“ (1990) by Gary Moore, who died two days ago on 6th February 2011.

The district court in Munich in Germany ruled in 2008 (LG München I, 03.12.2008 - 21 O 23120/00) that the song "Still got the Blues" was a copyright infringement of the older German song „Nordrach“ (1974).

www.justiz.bayern.de/gericht/lg/m1/presse/archiv/2008/01717/index.php

The highly interesting point of this court judgment is the fact that Gary Moore committed copyright infringement, because the song „Still got the Blues“ is very similar to, but not identical with the older song „Nordrach“.

Remember: Most fonts are very similar to, but not identical with other fonts.

If fonts were copyrightable, then most fonts were copyright infringements, because most fonts are very similar to, but not identical with other fonts.

The Munich court ruled:

"There are differences, but the similarity is striking."

("Es sind Unterschiede vorhanden, aber die Ähnlichkeit ist so frappierend.")

www.aufrecht.de/urteile/urheberrecht/still-got-the-blues-oder-auch-nicht...

A song is a brief piece of recorded entertainment that can be listened to numerous times, and it always sounds the same. No matter how many times you listen to it, you will never be paid to do so.

A typeface is a tool that can be used numerous times in man different ways forever and rarely looks the same. Being a tool, it can be written off as a business expense as well as charged to the client. You can more than make your money back with one professional use.

Isn't a great deal of this legal maneuvering with copyrights and licenses and subscriptions of "font software" also an outgrowth of the Eltra decision and ensuing copyright law since (ie, the lack of protection for actual designs of type)?

Luke will have to ask himself if being embroiled in a tedious discussion on legal-eeze terminology, with no possible conclusion, is worth the effort? Perhaps he will decide that his time is better spent just designing type and bringing it to market?

If you want to be a type designer, just do it and leave all of that other stuff to lawyers. Trust me, there is no equitable "understanding" to be achieved and there is ALWAYS someone to argue about it ad nauseam.

If you need a good lawyer, I can give you a recommendation.

Software shmoftware. You need an electronic device with a screen to read it, right?

ebspoony:

> Isn't a great deal of this legal maneuvering with copyrights and licenses and subscriptions of "font software" also an outgrowth of the Eltra decision and ensuing copyright law since (ie, the lack of protection for actual designs of type)?

Yes.

see www.sanskritweb.net/forgers/eltra.pdf

> Software shmoftware

exactly -- if it's all just ones and zeros in Silicon, it gets difficult to draw a clear distinction between programs and data (e.g., Uli: "Type 1 fonts are data"). Some have even gone a step further, saying that "there is no software" at all, which would be Richard's comment thought through to its (philosophical) extreme...

The distinction between character and glyph, which has emerged recently and come to exist as a well-understood philosophical concept, is, in a practical sense, politically predicated on both the Unicode Consortium, and the Microsoft-Adobe instituted OpenType, and dependent on the Internet.

The distinction between character and glyph says something about the relationship between text and rendering, software and hardware, but the meaning is elusive. And this is what makes type design so interesting, among other things.

Regarding the Eltra ruling I should like to mention that the Eltra Corporation was a merger of The Electric Auto-Lite Company and the Mergenthaler Linotype Company. The majority of stocks were held by the investor Gurdon W. Wattles. This means that the Eltra ruling may also be viewed as a Linotype ruling.

Who knows the name of the designer and the name of the typeface not mentioned explicitly by names in the ruling

www.sanskritweb.net/forgers/eltra.pdf

Oldnick explained it very well:

Harvard Cyberlaw defines software as "the programs or other 'instructions' that a computer needs to perform specific tasks." In the case of a font, the specific task is to render letterforms of a certain design. Period.

Essentialy what a piece of software does is to take the information given by the user, or by another piece of software, interpret it and display it in another way.

The stress here is "or by another piece of software" since many (most) programs' can only operate if they interact with other programs. That is where Christoph Knoth is wrong:

The specific task to render a font is done by the font renderer of an application that takes font data as an input. So only the combination of renderer and font is software.

Sure, you cannot use font file without, say a text editor, but you also cannot use text editor, without an operating system (and you cannot have an OS, without BIOS). Does that text editors are not software?

Picture it this way:

Bios and Device Drivers
|
Operating system and text encoding software (sometimes part of the OS)
|
Graphic software, such as Adobe Illustrator.
|
Fonts and Plugins

No, but thanks.