The /bin/true Command and Copyright (2009) | Hacker News Reader

true Command and Copyright (2009)

trillian.mit.edu

142 pointskmstout83 comments

Adding copyright notices to every source code file is reasonable because it costs next to nothing. Figuring out what is copyrightable, on the other hand, is expensive, because it requires lawyers, and a waste of time because lawyers' answers are always "it depends". You never actually know how enforceable your copyright claim is until you actually litigate, which is REALLY expensive.

> a waste of time because lawyers' answers are always "it depends".

Sorry for the little OT rant, but I'll have to object to that. That's IMO less of a lawyer thing, and more a "asking a field expert a generic question and expecting an immediate, brief answer" kind of thing.

In my experience, if you straight up ask any field expert, be it a lawyer, a doctor, an engineer, etc... a vague, but obviously case specific question, the real experts will rather try to point out that they lack necessary background information and that the underlying subject matter is not that simple ("It depends", "I'd have to look into this", "Well uhmmm, possibly/maybe/yes/kind of, but...", "It's not that simple").

If you're not particularly familiar with some subject matter, it is kind of hard to gauge that the "simple question" might be very vague, maybe confuse some subject matter and might be edging on a very complex topic.

I take it most people here have some sort of programming background? Just try to remember the last family gathering where someone started pestering you about their computer problems, or tried to start small talk about something they picked up in the headlines. Or maybe a meeting with management where you were asked a "but couldn't you just [simply]...." kind of question (or even worse: "exactly how long will this take?") and they tried to press you for a swift, immediate answer.

From my experience so far, I'd even go further and suggest to be skeptical of any "expert" who has a habit of confidently answering briefly and to the point.

You're right. I am a lawyer, so you'd think I'd be more careful. I should have written "lawyers' answers are usually 'it depends'". You know, because it depends.

It’s almost as if matters that go to court have to be figured out on a, well, case to case basis.

Adding copyright notices to every source code file is also pointless, since copyright notices are no longer required to obtain copyright protection.

It might help to prove intent when somebody does violate a license (they cannot say "oh wow this was copyrighted? i totally didn't know!")

I don't think that follows naturally.

The default if code does not have any notice anywhere is that it is copyrighted, and no license is given to anyone to use it, so any use of it is a violation.

At least in the US, all creative works are copyrighted by default (such as this comment, which does not include a notice of such, but is copyrighted).

Because of that, the default assumption of anyone would be that anything is copyrighted and cannot be used for any purpose outside of fair use unless there is a specific license granting additional usage.

A copyright notice is relevant to willful infringement, which may result in higher damages. Willful infringement requires that the infringer actually knew or should have known (or deliberately avoided finding out) that the copied material was copyrighted.

When you're talking about source code, I don't think anyone is going to get away with "I thought the author died more than 70 years ago".

What I look at answers on Stack Overflow, the lack of copyright notices on the code makes me feel safe to copy it. Same goes for grabbing code from random person's blog.

If they had put explicit copyright notices in the code snippets, I'd assume otherwise.

Your assumptions are likely wrong for those cases.

In the case of a random blog, the code likely does not have an associated license, meaning it is copyrighted and any derivative use of it (i.e. copying it) is a violation of their copyright.

In the case of stackoverflow, their ToS require people who post code to share it under a Creative Commons license [0]. Unfortunately, if the author of the post didn't realize that, it's possible they did not have the original copyright and couldn't make that assignment, so you still may have to do your own due diligence.

Where things work as intended on stackoverflow, you are required to provide attribution.

Your assumptions are backwards, and it sounds like that results in you violating copyright.

[0]: https://stackoverflow.com/help/licensing

I know the assumptions are wrong. But the point is the lack of explicit copyright notices on blogs or SO posts is probably enough to convince most courts that you did not willfully violate copyright. The parent I responded to implied you’d have to convince a court that you thought the copyright expired due to death.

In general I suffer from NIH complex so don’t really copy code from those places ;)

> Where things work as intended on stackoverflow, you are required to provide attribution.

Most(*) small snippets of code (e.g., 1-10 lines) on Stack Overflow are probably not copyrightable, at least not under U.S. law. You don't need a license to copy and use something that's not copyrightable.

(*) But not all. As usual, it depends.

If any creative work is copyrighted by default (and ignorance of the law is not an excuse) then surely all use of a creative work without a license is willful infringement?

U.S. government works are not eligible for copyright.

Some works do not have sufficient creative element to be covered by copyright, or the other user may be using only the non-covered elements of the work (factual collections, or see recipes).

But probably most important, fair use is a limitation to the exclusive rights of copyright, in spite of copyright holders' frequent assertions that "any other use without written consent is prohibited". A use may be fair even against the express wishes of a valid copyright holder, as fair use is not infringement. Factors to be evaluated in fair use include, but are limited to:

  (1)the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;

  (2)the nature of the copyrighted work;

  (3)the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and

  (4)the effect of the use upon the potential market for or value of the copyrighted work.

https://www.copyright.gov/fair-use/

https://www.law.cornell.edu/uscode/text/17/107 see 107-112 for other limitations of copyright.

That's the catch though isn't it? It's only wilful infringement if there was intent. Just because you don't know the law, that doesn't make it allowed but it does make it unintentional.

A license notice provides a clear explanation of the rights and responsibilities of the user. This dissuades most any notion of "lack of intent" as the restrictions are put in your face from the start.

Hypothetical situation:

I have a bunch of code in a repository or set of repositories. I got confused and thought this file was one we created/had a license to reuse rather than a piece of third-party code, and thus copied it into a different project.

For my hobby projects I add a one-line comment at the bottom refering to the license and its file.

    // Please keep this line & LICENSE: <license> <year> <name> <email>

Indeed. That said, it's worth noting the files in the article predate that. The US didn't join the Berne convention until 1 March 1989.

Copyright notices are useful the same way written contracts are: you legally do not need a notice (written contract) to establish copyright protection (enter into a contract), but if a dispute arises it helps establish the fact that the user was notified (the other party agreed to this contract).

A question of 'law' vs 'facts'

Those aren't at all analogous.

The automatic state for content is "protected by copyright," unless a different arrangement is made.

The automatic state for contract terms is "not agreed to," unless a different arrangement is made.

tl;dr You need proof (written or otherwise) that a contact happened; not so with copyright protection.

Oral contracts are legal and valid, yet most people write down their contracts for good, practical reasons outside of strict legality.

Same goes for copyright notices. For example, the notice makes it hard for a violator to use a negligence defense, effectively increasing the penalty for violations.

Negligence is not applicable in this case, as, like the comment above pointed out, the default state is copyrighted.

It's a bit like putting a "stealing this car is illegal"-sticker on your car - sure, you made that clear, but "I didn't know stealing this car was illegal" is not going to fly as a defense either way.

People write down contracts for proof. Without that, it's dubious if the contract even existed.

But proof is not relevant for a copyright. Even without writing, the existence of the copyright is not in doubt.

But the owner of the copyright is.

Yeah this is funny, but it's not stupid. It's also a lot simpler to have a company policy that says "every code file should have this boilerplate at the top of it" than to try to specify exactly what does and doesn't. You could even check for this in your CI pipeline, to avoid wasting time debating it in code review.

To expand on that idea, purely functional expressions are not copyrightable in the first place. How and when software becomes copyrightable is undefined in law.

Why not? Do you have a reference for that?

US Copyright law: https://www.law.cornell.edu/uscode/text/17/102

  (b)In no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work.

Thus, if it is simply the description of the algorithm to accomplish a process, it isn't copyrightable (functional processes are the domain of patent law, not copyright). But creative ways of describing an idea or process could be copyrighted...

Further discussion https://www.copyright.gov/circs/circ33.pdf https://www.iusmentis.com/copyright/software/protection/

I, too, miss GrokLaw…

> Adding copyright notices to every source code file is reasonable because it costs next to nothing.

Yes, if your integrity is worthless.

My company doesn't require this, but I also don't understand in any way how adding copyright boilerplate to each source file has anything to do with the writer's integrity. Are the people in the GNU foundation lacking in integrity, since they have this policy?

Placing copyright notices on things you don't have a copyright on amounts to making a false claim that they are not only copyrightable but that you own the copyright, hoping to intimidate other people out of exercising their legal rights. Lying in order to intimidate people, particularly to intimidate them into paying you, is not only dishonest but contemptible. It's no different from the mugger who says he owns the Golden Gate Bridge and demands $10 from you to cross it.

The GNU foundation, whatever its failings, does not have a policy of claiming copyright on blank lines.

I thought your comentarii was referring to the practice of adding copyright notices to files in general. Specifically adding a copyright notice to an otherwise empty file is indeed scummy.

> So if you use blank lines in any of your files, you are in blatant violation of AT&T's copyright claim

Independent derivation of a work is defense against copyright infringement.

Even if the copyright on these blank lines were to be valid, it would be reasonable to claim that your own use of blank lines was independent, not a derivative work copying from theirs.

Patents do not require actual copying to infringe, but copyright does. If you can prove you did not know of the original work, or otherwise would have come up with the same thing on your own, that's enough to absolve you of copyright issues.

Proof of direct copying may not be required, if AT&T can instead prove Access and Substantial Similarity.

For access it could be a little tricky, certainly every one that has posted a reply here has had access, and if this post is popular enough it could be argued anyone could or has had access (this happened in some music cases in the last few years).

For Substantial Similarity your definitely in trouble if you have exactly 3 blank lines and probably in trouble if you have 2 or 4 blank lines, but you may be fine if you have like 6 or 12 blank lines.

While I'm being a bit sarcastic about the number of lines being "substantially similar", the legal standard in the US is not if you actually copied something, but if the plaintiff can convince a judge/jury that it's more likely that you copied it than it is that you created it on your own.

> Patents do not require actual copying to infringe, but copyright does.

In other words, if you use Ctrl+C, Ctrl+V to copy these blank lines, then you infringe AT&T copyright. If you don't and add blank lines by other means - you don't ;)

No. You can copy without copy-and-paste. But, a blank line isn't enough to be a creative work.

As long as it wasn't a blank line typed by John Cage ;)

> The /bin/true (or /usr/bin/true) command is now nearly obsolete, because most extant shells now have a builtin "true" command. But it's still useful occasionally, for various silly reasons [...]

This make it sound like this is useless, but I disagree. For example, recently I had some games crash because they couldn't find pulseaudio, and making a symlink from pulseaudio to /bin/true fixed my issue (and yes, they had sound). So there are certainly legitimate uses for it.

Instead of a symlink to true you could have made an empty script.

And violate AT&T's copyright? Perish the thought!

Well the symlink command was also easier to share with others having the same issue, whereas a one liner that creates an executable script which returns 0 is less ideal.

An empty script returns 0 by default, thus `touch /path/to/missing/library` is a sufficient (and shorter!) one-liner fix to this problem.

I know, but you still have to make it executable, hence it's not as convenient.

And on top of this, an unlink is less error prone than a rm -f.

“Note that there is one less blank line here; it has been replaced by the #! line. But otherwise it is identical. Sun has merely passed on the copyright notice. I wonder if Sun has written permission from AT&T to use blank lines in their code?”

Fun fact: SVR4 was a joint AT&T/Sun project, so this was a SunOS file originally — I believe the “SMI” in the #ident line stands for “Sun Microsystems Inc.” and was a standard feature of all our SCCS identifiers.

On macos, /usr/bin/true is 118K, not counted the linked shared libSystem.dylib. I wonder what did they put in there? Doesn't seem to have --version or any GNU bells and whistles of that nature.

Based on a quick check of the disassembly, _exit is dynamically linked. It appears that because of that dynamic linkage, there's some information stored for dyld to use. That data seems to be split across a few sections which are loaded with padding.

It's also a fat binary. It contains both an X64 and AArch64 version.

I should add that it really is doing nothing other than calling exit:

    EntryPoint:
        push       rbp
        mov        rbp, rsp
        xor        edi, edi
        call       imp___stubs__exit  
    imp___stubs__exit:
        jmp        qword [_exit_ptr]

How big is a basic hello world compiled with -O3?

> So if you use blank lines in any of your files, you are in blatant violation of AT&T's copyright claim.

I think it has to be exactly these three blank lines, not just any three blank lines.

Does someone need to write an essay titled "What Colour is your whitespace?"?

ref: https://ansuz.sooke.bc.ca/entry/23

(1990), or perhaps as late as (2009) if you have the right sense of humour.

Previous discussions:

• https://news.ycombinator.com/item?id=3528663 in 2012

• https://news.ycombinator.com/item?id=15457281 in 2017

That's fair use, not a copyright violation.

Google: "(in US copyright law) the doctrine that brief excerpts of copyright material may, under certain circumstances, be quoted verbatim for purposes such as criticism, news reporting, teaching, and research, without the need for permission from or payment to the copyright holder.".

The one confusing thing to me about this is that you're not supposed to append copyright years to a notice when the contents have not changed, yet this script got copyrighted all the way to 1989.

We used to do this at work. Then someone bumped the date in every file. I can point to PRs that copypasta'd old dates in from people who copied the notice from deeply ancient source files and claimed a copyright date ages in the past. Or files that were updated and nobody bothered to bump the date.

As a result I ripped out all the dates, with the blessing of our legal department, because they were next to useless. We've got a git history and while the dates there aren't perfectly tamper and error free, they're vastly better than all the nonsense we had our headers.

Pretty sure these dates are a software development legal cargo cult and would encourage anyone who disagrees with me to actually sit down with their legal council and go over the issues and publish a nice medium article if you think I'm wrong. And make sure to spell out issues like this one which opposing council would be sure to use where the date bumping on /bin/true is a clear violation of policy. At best it seems legally useless and at worst you've got a policy that opposing council can probably slam dunk prove that you are not following.

I’ve heard that, but I’ve also heard that the year is supposed to be updated whenever it’s published again.

But by updating the copyright year, haven't you thus changed the file's contents?

(Please read this comment as written in the same tone as the original article.)

> you're not supposed to append copyright years to a notice

It shouldn't matter because the date of actual first publication would control, not what is on one particular notice.

The date of revisions extends the copyright beyond the original cutoff date.

A later date on a subsequent publication including revisions doesn't extend the copyright term of the original material.

See also https://www.telegraph.co.uk/culture/music/rockandjazzmusic/3... (https://archive.is/sjYJG), where the rumored settlement payment was £100,000.

Sometimes parties pay what seems like a lot of money to settle dubious claims because it's cheaper and less time-consuming than fighting them. That can be completely rational. But paying £100,000 for infringing one minute of silence is just ridiculous.

Well, he sold a lot of copies of that one minute of silence, so I think the real question is whether it was copyrightable. I mean, if the answer is "yes", then the following question is what reason there is to suspect that the government that made it copyrightable is anything more than a criminal racket.

> so I think the real question is whether it was copyrightable.

Whether a particular thing is copyrightable or not is a question of law and fact. The government will accept your $25 for a copyright registration on anything you submit, but they don’t check if it’s copyrightable or not, if you own it or not, etc. So, no one knows for sure whether that one minute of silence is copyrightable or not until someone challenges it in court. Until then, all we can say is probably not, based on prior cases.

The liner notes included an attribution to John Cage, making the case somewhat more complicated.

He said it was to "Clint Cage", a "registered pseudonym" for himself, but obviously he was paying homage to John Cage.

rewrite in C++ to avoid copyright

https://lists.freebsd.org/pipermail/freebsd-questions/2012-J...

I would like to point out the obvious fact, which is newline characters. Even if those empty lines were empty, they still had \n's , maybe \r's too.

> This prevents AT&T, Sun or SCO from taking the GNU code, claiming it as their own, and suing the linux crowd for infringement.

you see, no matter how trivial the matter, the lawyers manage to have their share.

The court transcript of a copyright lawsuit would be hilarious:

--What is the copyright content? Nothing.

--Are you saying that that you have no copyright? No

--Okay, can you then give me a summary of the content? It's nothing

And so on.

I'm sure expert witness Bud Abbott could clear this all up for them.

Well damn. I guess I can’t use blank lines in my code anymore. This is going to be rough.

Also are there any copyright koans? As in a copyright notice that copyrights itself?

Just copy them from public domain works! Or, come up with the independently - an absolute defence to copyright infringement.

I vaguely recall seeing a license that claimed the license text itself was under copyright and it included a license to use the license. I’m not a lawyer though so no idea if you can actually copyright a license text.

Companies have to vigorously protect their intellectual property and copyright of source regardless. If they fail in one instance to assert those then it can badly impact them when it matters.

This is false in the United States. Failure to assert copyright does not relinquish it. Trademark does, however, because it is connected to brand recognition concepts.

Reminds me of a CEO I used to work for who realized that old blog articles (which no-one was reading anyway) were copyrighted of the years they were written, and made a fuss that "no-one is paying attention, these details are important for our image, we need to show that we're up to date". There was no explaining him that this is not how copyright works, what's important was that we remind people of the current year.

I heard this about brands. Do you have examples for copyright?

/bin/true is backwards then since 0 usually means false

> Zero is used to represent false, and One is used to represent true. For interpretation, Zero is interpreted as false and anything non-zero is interpreted as true.

Not in this context. In POSIX exit codes, 0 means “no error”, i.e. success. && and || in the shell even work like that.

Shell commands use an exit status of 0 to indicate success, and any other number indicates failure

Reply on news.ycombinator.com