Comment by crazygringo | Hacker News Reader

Comment by crazygringo | Hacker News Reader

Parent Full thread

Funny, I asked here on HN just a couple of months ago why nobody has done something like this:

https://news.ycombinator.com/item?id=37286891

The user 'pessimizer' replied that there was a company OptiFont that seemed to do something just like this, to sell cheap font clones:

http://abfonts.freehostia.com/opti/

http://luc.devroye.org/fonts-27506.html

It still seems so relatively trivial to automate this -- just render every character on a <canvas> at a fixed known font size at super-high resolution and autotrace outlines, and render every character pair to figure out kerning, and the common ligatures. The only major thing missing is hinting, but that's irrelevant for printed material and not very important on retina screens -- and auto-hinting gets you more than halfway there on non-retina. (And then there are all sorts of special font features like nonstandard ligatures, which can either be more carefully replicated, or just ignored.)

I'm not saying it would be a good/ethical thing to do this (I personally want designers to be supported) -- but I am kind of amazed that the market for digital fonts still exists, when in theory you can produce a perfectly legal copy and not pay a thing.

>but I am kind of amazed that the market for digital fonts still exists, when in theory you can produce a perfectly legal copy and not pay a thing.

There is a growing trend for style guides to use typefaces from no-cost or open libraries. This isn't limited to small companies either, I can think of a handful of industry heavyweights in tech, food and medicine who each have taken up this approach. This forms the first reason not to dupe fonts, because there's plenty of good ones already available for no cost.

I've found that companies that do this have an easier time of staying on brand since it's so simple to obtain the necessary fonts.

As for seeking out deliberate copies of popular fonts, there's a few reasons why an agency wouldn't want to propose this:

1. It's easier to sell in a well loved font, with a proven character from a foundry with heritage. Imagine talking to a board of suits, they're often looking for the safe option.

2. It's kind of a dick move for an agency to be ripping off the foundries, especially small modern ones. Creatives do enjoy what typographers make and can see the love put into their craft.

3. Typographers are not faceless. Modern font designers are still alive and active on socials, you might personally know who you are ripping off.

4. There's a bit of hypocrisy in hating on frankenfonts such as Arial, while deliberately seeking out dupes.

5. It's trivial to open a font in a font editor and resave it with different specs, if a creative wants to be sneaky they can just do this themselves anyway.

6. Adobe have also thrown their hat into this ring by offering a range of popular fonts available for use for no extra cost with their creative cloud subscription. It's not needed to seek out a dupe-font, when the real thing is available with a few clicks.

> It's kind of a dick move for an agency to be ripping off the foundries, especially small modern ones. Creatives do enjoy what typographers make and can see the love put into their craft.

But foundries sometimes do the same: they take samples of, say, 18th century font, and digitize it, and claim copyright. Do you think it is fair when someone copies Helvetica or 19th century Bookman [1] ?

[1] https://learn.microsoft.com/ru-ru/typography/font-list/bookm...

I don't feel your example is comparable here, but I'm interested in hearing maybe a different one that conveys your ideas better.

The reason for my thinking is that I don't see the problem in a foundry today building a font from a typeface that is neither commercially available, nor under any protection (or potentially even ownership). To me this is not comparable to dodging licensing fees from another foundry that is licensing a font. I actually feel that this foundry would be doing the design world a service by ensuring that pre-computing typefaces are adapted for modern use. Factually this process occurs frequently without controversy.

As for claiming copyright: There is a significant amount of non-trivial work involved into building such a font, including a number of creative decisions that the typographer will need to make to address shortcomings as well as fleshing out the wealth of missing characters that are commonplace in contemporary publishing.

Type from the 18th century or earlier (pre film/digital) is imperfect. Not every 12 pt 'm' slug in that font drawer is identical, and not every imprint from the same 'm' slug in that book is identical. It is a lot of work to create digital versions of these metal/wood typefaces.

This segues beautifully into a longer discussion about visual acuity testing charts.

It's relevant because even today many VA charts still feature the historical numbering systems which are deeply inconsistent from eye chart to eye chart. (The reason why this fault is of no concern is because in modern use, they are only stated to infer a vague acuity.)

Historical revivals are hardly the same thing. They often end up far from the original, and require a ton of work. It isn't just a matter of digitization. And they "claim copyright" on their derived work, not the original. With fonts they really just copyright the name.

Requiring a ton of work doesn't really matter one way or the other for copyright. Mowing the lawn is a ton of work. Doing the work doesn't let you prevent other people from mowing their lawn to the same height as yours.

Additional information on this point: "Sweat of the brow" [1] is a doctrine which grants copyright on a work due to the time and effort the author put into making it. The US Supreme Court rejected this doctrine, meaning that only the creative aspects of a work are relevant to copyright. For that reason, in the US, a phone book is not copyrightable unless the arrangement of its contents is creative [2].

The EU narrowly rejects "sweat of the brow" copyright for recreations of public domain visual art [1]. I have no idea whether the same applies to font faces, public domain non-visual art, or phone books, but I'm assuming that it doesn't.

[1] https://en.wikipedia.org/wiki/Sweat_of_the_brow

[2] https://en.wikipedia.org/wiki/Feist_Publications,_Inc.,_v._R....

That work I was referring to is a creative, artistic process of designing a font based on a historical starting point. That does matter. Hardly mowing the lawn or compiling a phone book.

They're not just scanning existing glyphs. They're performing copyrightable work to add de novo metadata like hinting and kerning.

Your approach would result in impossibly large binary font sizes unless a really good (smart) engine was present and utilized to separate the hinting logic from the underlying symbols. Then again, with most fonts these days putting in zero manual hinting work and just relying on the default font engine heuristics, it might not.

I was supposing zero hinting information would be included, so the file sizes should be perfectly normal.

Because while font glyphs/designs aren't copyrightable, I'm assuming that hinting instructions are, even if they could be reverse-engineered this way.

Reply on news.ycombinator.com