Goldman Sachs Created a Font, but You're Forbidden by Its License to Critique GS | Hacker News Reader

Goldman Sachs Created a Font, but You're Forbidden by Its License to Critique GS

techdirt.com

103 pointsTomte42 comments

This license doesn't work in Japan for the result of font rendering is not protected by the copyright.

It was decided in the court that in order the shape to be recognized as a character, that shape must follow the pattern we generally agree upon. The copyright is granted for the expression that reflect the human thought, if the shape can be recognized as a character, there is limited human thought left.

I think that's a fair argument.

Font copyrights are connected to the embedded hinting program.

My knowledge in font copyrights is limited, but what if one just removed the embedded hinting programs? Sure, it’d look bad, but would that fix the copyright issue?

More or less - this kind of thing does happen, but not by as technical means as extracting the hinting from the font (because you would need to provide new hinting anyway for it to look any good, and you get into a complex knowledge firewall situation doing that).

Instead, it's common for font foundries to design a new font based on the result of an existing one, creating an effective duplicate font divorced of copyright claim. It's pretty common for a few differences to be inserted though, I think less out of concern over copyright and more so that it's possible to determine which font was used for license enforcement of the new one.

This prominently happened with Arial, which is a copy of Helvetica designed by Monotype so that they could stop paying fees to license Helvetica from Linotype as part of their font offerings for early raster printers (back when fonts were embedded in the print engine). There are numerous other examples, to the extent that if one foundry offers a font which is particularly popular, the other major foundries probably also offer their own "version" of it.

I am not sure it's that fair: have you ever attempted to design a typeface?

FWIW, I am using "fair" in comparison with other copyrightable work: there's a lot of it that obviously did not involve a lot of "human thought" (Hollywood blockbusters included ;)).

The font data itself is copyrightable in Japan.

Did GS pull the page touting their font? Or is my browser acting up?

https://design.gs.com/d/story/goldman-sans/

I can access the page perfectly fine

Thanks. Still doesn't work for me, Safari or Brave. Or maybe they block access via proxies.

I get a blank white window on Firefox too

Not sure how you got your link. The less specific domain seems to have working links

Acting up on my end as well. Looks like a iframe 404ing.

I run pi hole as a data point.

My Android + Chrome has the center frame with the error page reading ""Requests to the server have been blocked by an extension."" - I don't even have extension in mobile Chrome.

Then a few seconds later my Android prompted to send a bug report

> Requests to the server have been blocked by an extension.

This is after disabling all of my extensions, including uBlock Origin.

I've come across a couple of instances of this over the past couple of weeks. Not sure what to make of it.

For me as well, it's not working in Chrome or Firefox. Seems archive.org has a earlier copy that is working (https://web.archive.org/web/20200623163509/https://design.gs...) Unsure if that's the actual font though, and I'm too lazy to verify.

Seems they either pulled it or something malfunctioned.

Edit: navigating to the direct URL seems to work too. http://view.ceros.com/goldman-sachs/type-story-ii-1-2-1-1-1/...

It's a CORS issue. The page is served in an iFrame (//view.ceros.com/goldman-sachs/type-story-ii-1-2-1-1-1)

>[Report Only] Refused to frame 'https://view.ceros.com/' because it violates the following Content Security Policy directive: "frame-src https://marquee.gs.com:* https://*.marquee.gs.com:* https://research.gs.com:* https://cdn.appdynamics.com https://*.qualtrics.com".

>Refused to frame 'https://view.ceros.com/' because it violates the following Content Security Policy directive: "frame-src https://*.qualtrics.com".

I would speculate that CORS is somehow implemented as an extension (kind of like the PDF viewer or the crypto engine) which is why it is reporting an extension error

> (C)(2)(d) The User may not use the Licensed Font Software to disparage or suggest any affiliation with or endorsement by Goldman Sachs.

> (E)(2) Further, Goldman Sachs may terminate this License, without notice to the User, for any reason or no reason at all and at any time, completely at Goldman Sachs’s sole discretion.

> The User may not use the Licensed Font Software to disparage or suggest any affiliation with or endorsement by Goldman Sachs.

I would like to make clear that although I may disparage Goldman Sachs this does not mean I disparage any affiliation with them.

IANAL but this seems a reasonable interpretation to me. So stick to not disparaging your affiliation with GS (or their endorsement of you) and you should be in the clear. Say what you like about GS themselves.

I think it's fairly clear that it's

> disparage or (suggest any ((affiliation with) or (endorsement by))) Goldman Sachs

Not

> (disparage or suggest) any ((affiliation with) or (endorsement by)) Goldman Sachs

And GS has more lawyers than you.

>And GS has more lawyers than you.

yet they're still not able to write a contract punctuated so as to avoid ambiguous wording.

I wonder whether E2 is actually enforceable? How on earth are you supposed to know the license is no longer valid without notice?

Can I just use a different font for the disparaging parts? e.g.

"Goldman Sachs has consistently demonstrated they are a

    bottom-feeding, tone-deaf, greed-driven

organization of the highest order."

The difficult to distinguish 0's and O's are still the worst part of this. It's supposed to be for numbers focused things, come on.

> It's supposed to be for numbers focused things, come on.

It's not supposed to be for programming. If the 0 and O glyphs were identical, where would that confuse a reader, in a numbers-focused document?

For programming, this is why I name variables "foo" and "bar" and not "FO0" and "F0O". Really opens up the range of fonts you can use.

Contracts. I’ve run into the issue trying to reference paragraph 2O vs paragraph 20. (Paragraph 2 is often definitions and tends to get lengthy.)

That's a good example. It makes me curious about such nested references, though -- in my mind, they usually look something like 2.III.O, or 2(III)(O), or the like.

That’s definitely how they would usually be referenced but when it’s crunch time and you’re firing emails back and forth trying to hammer out the details that tends to fall by the wayside. Especially on mobile. And 99% of the time it’s clear enough. 2(B) or 2B doesn’t leave much room for confusion but oh man did those missing parens on the O cause us a lot of confusion about version control.

I meant finance and investment analysis, not programming.

Apart from its strange provenance, I thought it is a beautiful typeface. Something ugly gave birth to something beautiful.

and then wrapped it in an ugly license blanket

How to get people to use your typeface to disparage your company 101