evacide: I see that Discourse is happening among people who think US courts have not, in fact, found that code is speech and is protected by the 1st Amendment: www.eff.org/cases/bernst...

I see that Discourse is happening among people who think US courts have not, in fact, found that code is speech and is protected by the 1st Amendment: www.eff.org/cases/bernst...

Bernstein v. US Department of Justicewww.eff.org EFF established that computer code is speech and shielded the developers of privacy-protecting software from government censorship. In 1995, researcher Dan Bernstein planned to distribute an

Fun fact: long before there was an established term for it, Ada Lovelace called programming "poetical science": taking abstract ideas and turning them into simple beats anyone can follow along. Programming as expression is so old, it predates computers.

Some of my code would be considered obscene, thank god it’s protected.

Obscene speech isn't protected. Perhaps you should self-censor.

That’s what code review is for.

You'll traumatize your reviewers.

"But Russia!!1!1!1!!"

I would counterargue that there is no law, there's only enforcement. It's only protected until a judge (or panel of judges) says it isn't

Wait this has nothing to do with what Masnick was posting about, where he claimed that compiled software sold as products are also speech.

I don't think Mike is wrong in that. Compiled software is as much speech as the original source code. Some things can only be properly expressed in binary format. But if the US government sanctions 12 people in Russia I can't pay them for their speech, or guns, anymore.

I was not aware of this. It would seem similar challenges would undo ITAR, yes?

No, and I say this as somebody that has owned t-shirts that were ITAR violations if I took them out of country because they had code on them. Code is speech. Money is speech. And while I sometimes disagree with ITAR classifications they are well within the fed's power.

You're wrong about money, but you have august company in the error. As for code, you can write code that's dual-use under any DSGL classification. If it's free speech, how can ITAR prevent you from saying it wherever you want?

I can publish rocket guidance software on Github. I can own a shotgun. I cannot publish ICBM guidance software on Github. I cannot own a nuclear weapon. Hope that helps.

not really

The wikipedia article says the gov was forced to loosen regulations. It said nothing abt revising the ITAR regime.

Interesting, because I just read something that was more about model weights, but also got into the whole "code is speech" thing. And the idea that it is settled seems a bit more murky. www.lawfaremedia.org/article/ther...

There Is No General First Amendment Right to Distribute Machine-Learning Model Weightswww.lawfaremedia.org Unlike source code, which humans use to express ideas to each other, model weights function primarily as machine-readable instructions.

So if I read the summary correctly, *source code* is speech, but there's a chance that compiled binaries not accompanied by their source might or might not be. That makes sense to me as a software person, but also feels kind of weird, legally.

Specifically what feels weird is the way copyright attaches to things that under this view aren't speech: If you write source code, you own the copyright. If that's then compiled, it's a derived work of your copyrighted thing, so people need a license from you to distribute binaries.

So if compiled binaries aren't generally speech, you've got this thing that can be a derived work for the sake of copyright but somehow exempt from First Amendment protections and so subject to much stricter control, the way physical objects are.

Oh man there was a book on the history of cryptography that came out years ago that was super interesting and went into how the government wants to control all encryption. I think it was Crypto by Steven Levy. Kind of surreal to see that this legal fight went on for another 7 years after the book.

I have terrible news for you about governments' opposition to strong encryption, my dude.

I still remember a time at Computers Freedom and Privacy in the 1990s, not long after the Clipper Chip battle, where the government representative basically said "you got us this time, but we'll be back." It's kind of like a horror movie with too many bad sequels. "It's over ... isn't it?" Nope.

I was phrasing myself very poorly, apologies. Just trying to remembering the book from 20+ years ago and that I thought this was a topic in the book and it dragged out for years beyond that.

How do you find the dividing line between machine tool commands (punched paper tape that tell the machine which motors to activate and for how long) versus computer programs that basically do the same thing at a much larger scale?

Post