Why are developers still limiting themselves by posting their code on the clearweb in a manner they can be traced and held liable for? Host everything on Tor from a server outside USA jurisdiction and this should be a total non issue. Code is speech. It’s time to stop letting the government get away with trampling on our 1st amendment rights to free speech.
My failed startup operated in a similar space: SDRs & military applications. I dunno how people don't plan from Day 1 with a knowledge of Uncle Sam's heavy-handed export restrictions in their mind.
I was using existing open-source software as a basis (GNU Radio), all of my engineers were foreigners in their home countries, my SDRs and single-board computers were dual-use hardware from multiple other nations, and my company was in Hong Kong. All because I knew I primarily wanted to target foreign countries with behind-the-power-curve militaries, not the admittedly-huge US defense market with its obnoxious barriers to entry. If you operate in the US, just keep your stuff closed source until you can afford expensive lawyers to tell you what you can share.
> It’s time to stop letting the government get away with trampling on our 1st amendment rights to free speech.
That needs to be a legal fight, since classified information is specifically exempted from free speech, among a long list of other things (https://en.wikipedia.org/wiki/Freedom_of_speech#Limitations). Moving to Tor might skirt the rules, but does little to challenge them, and won’t prevent any legal trouble for someone who gets caught. (It could make things worse, as it might demonstrate intent.) If you believe that free speech should be absolute, that needs to be litigated and voted for. Just remember Chesterton’s Fence: all the free speech limitations we have now have already been litigated and fought for. There are good reasons that freedom of speech is not absolute.
If someone independently invents something using the available resources at hand it shouldn’t be able to be considered classified or copyright restricted, if it was really that advanced and sophisticated then nobody should be able to discover it unless it leaks. If there’s no proof it leaked to the public in violation of a government employee’s oath then the information should be legal. In that case I agree anyone who leaks classified documents should be charged for treason. But there’s a major difference between a software developer accidentally inventing a banned algorithm and getting slammed with the full force of the government and secret information the government has being leaked.
I can agree with everything you just said, but there’s a bit of a misconception of what free speech means tied up in this. The government isn’t claiming ownership. Freedom of speech is a protection the government offers to protect citizens against itself, and the government defines what freedom of speech means. It’s probably best to leave copyright aside, introducing that now and mixing it up with free speech is going to muddy the discussion. This isn’t a copyright issue.
It doesn’t matter if I independently invent nuclear weapons, I’m still not currently allowed to open source them for other people, possibly in other countries, to use. That isn’t because the government thinks they own my ideas, it’s because the government believes that sharing information on how to build nuclear weapons is bad for us and threatens our safety. (Edit) BTW, it’s also important here to recognize that claiming “independent” invention is risky and problematic, if you received any benefit from your environment in the form of education, ideas, collaborators, parts, market conditions, etc. There are very few, if any, truly independent inventions.
Note I’m not making any arguments on whether ITAR should or should not be classified. What I’m pointing out is that that is what needs to be debated - whether ITAR is classifiable (or otherwise export controlled), and this isn’t otherwise an issue of free speech failing to be absolute. It’s a simple fact that freedom of speech is not absolute, and therefore demonstrating perceived abuses needs to be demonstrated based on the specifics of the case. Why should ITAR be declassified/open? That’s what needs to be shown.
> The government shouldn’t be able to classify scientific information that the public is able to discover on their own
It doesn’t matter if I independently invent nuclear weapons, I’m still not currently allowed to open source them for other people, possibly in other countries, to use.
It's funny you mentioned nuclear weapons. Nuclear design information is purportedly "born secret" in the United States (that is, purportedly restricted based on what it is, not where it came from).
However, the one time that this looked like going to trial, when a magazine called The Progressive intended to publish Howard Morland's article describing the operation of thermonuclear weapons in the 1970s and the DoE tried to stop them (United States v. Progressive, Inc.), the Government ended up backing down before a final resolution of the case. The article was published, but whether or not this prior restraint is legal is still undecided.
It is also worth noting that the argument for suppression leaned on the severe consequences up to thermonuclear war, something which the argument for suppressing passive radar technology is going to have a harder time with.
> the argument for suppression leaned on the severe consequences up to thermonuclear war, something which the argument for suppressing passive radar technology is going to have a harder time with.
Why would you assume radar needs the same justification as nuclear weapons, when there are lots of export controlled products and ideas already that aren’t justified by the specific or immediate threat of thermonuclear war, e.g., encryption, weapons, chemicals, software, etc.? That’s what ITAR is…
The Morland/Progressive story is quite fascinating. It’s worth pointing out that it happened more than 30 years after the original designs, after other countries had their own nukes, and the case was dropped due to all of the info Morland shared already being in the public domain. It’s not really an example of free speech winning against the government, and isn’t precedent for how we’re handling defense related technology today.
Why would you assume radar needs the same justification as nuclear weapons, when there are lots of export controlled products and ideas already that aren’t justified by the specific or immediate threat of thermonuclear war, e.g., encryption, weapons, chemicals, software, etc.? That’s what ITAR is…
The first amendment arguments in the case (and related cases like the Pentagon Papers) in part hinged on the immediacy and degree of harm to the United States that publication was likely to cause. "Thermonuclear war" is an easier sell in this balancing of concerns than "better radar".
It’s worth pointing out that it happened more than 30 years after the original designs, after other countries had their own nukes, and the case was dropped due to all of the info Morland shared already being in the public domain.
Not exactly. The Teller-Ulam design was about 26 years old at that point - it was first tested at Ivy Mike in 1952. The Soviets and British developed similar staged thermonuclear weapons over later part of the decade, the Chinese and French not until the late 1960s.
Morland and The Progressive argued from the start that the article was based on information in the public domain, when the DoE first came to them. The government on the contrary argued that the article contained still-secret information. The government then did say they were dropping the case because it was mooted by events, but dropping it did also avoid the risk of an adverse ruling - as I said, the legality of these kinds of prior restraints on publishing independently-derived material remains unclear.
> “Thermonuclear war” is an easier sell in this balancing of concerns than “better radar”.
Why? That doesn’t explain the entire rest of everything that is currently export controlled, right? Threat of Thermonuclear war is not the basis for the ITAR program.
Why? Because it's a more grave and imminent threat, and the jurisprudence around first amendment vs prior restraint against publishing of classified material is around balancing of concerns, with prior restraint justified in cases of "direct, immediate or irreparable harm to our Nation or its people". It's not a black and white line, the degree of harm matters.
"Everything else that is currently export controlled" has not been litigated (but note that we are only concerned here with limits on the export of expressive text, not for example actual armaments). The one case where restrictions on the export of software by ITAR was litigated - Bernstein v. United States Department of States (District Court of California) (1997) - the District Court ruled the regulations in question were an unconstitutional prior restraint on speech and issued a declaratory judgement preventing the government from enforcing the ITAR in question against DJB or anyone else seeking to use, discuss or publish his encryption code.
Oh I see what you mean now, thanks for the explanation. Yes I agree that when challenging in court whether some tech can be classified or not, then nukes are an easier sell than radar. I was more referring to what happens before that, how something becomes classified and/or export controlled by the government. It doesn’t have to be litigated in order to legally limit feee speech, while it does have to be challenged, litigated, and won in order to become free speech.
I'm not a fan of github for other reasons, but how the heck would your solution work for searchability and discoverability, two of github's largest values?
This could work for distribution but it's not a solution for shielding the developer(s). If you have already publicly published code with attribution I would not consider tor + forward pawns to be 100% invulnerable to forensics to determine authorship. So now you're looking at tackling code transformation without obfuscation to cover provenance, which sounds non-trivial.
(Your comment made me wonder if coPilot can also be used to fingerprint developers based on their existing code.)
