Hacker News Reader

Comment by dewitt | Hacker News Reader

Before anyone mocks this guy (sadly, partly too late), please try to keep in mind that he (a) does have a reasonable right to what he says he wants (to have some say in what he believes are his inventions), and (b) it's an illustration that the rules of intellectual property and open source are not always clear, even to those following well-established patterns.

While he's inarguably incorrect about quite a number of things, it's always better practice to assume the best of a fellow engineer and treat this as a teaching opportunity, not a cause for pitchforks and belittlement.

The package maintainer, threedaymonk, handled it perfectly, imo, by respecting the desires of the individual and not porting it, and closing the issue out quickly without further escalation. While it may have been technically "right" to copy the code, it wouldn't have been worth the damage it would have caused to the person behind it. Credit for handling a unfortunate situation with grace.

View on HN

> While it may have been technically "right" to copy the code, it wouldn't have been worth the damage it would have caused to the person behind it.

Well, I will play the devil's advocate. If the code is useful to enough people, then there is damage to everyone else who might have benefited the port.

What I understand from his comments is that he does not only not want you to use his implementation but also "his" algorithm as a basis to the port. Now assuming that he has an efficient algorithm to solve the problem, then he basically forbids other people from using the efficient algorithm.

So the needs of the community's damage may in fact be more than the individual's. Also say the exact same thing happened with GNU toolchain, does the individual's damage still outweighs the community's?

I would not bring this up if his tone in his comments wasn't arrogant despite the fact that he is at fault by not reading making any effort to understand the licence.

EDIT: Can down voters care to comment?

> If the code is useful to enough people, then there is damage to everyone else who might have benefited the port.

So what? There exist enough codes which are licensed under some kind of "open source" license for which you still have to pay to the patent holder if you actually use the code. While "somebody else" can benefit from using it, he/she still has to pay if the patent holder demands so.

Not every "open source" license is automatically a "GNU" one.

It's not only the moral issue (who benefits) but also the legal one. Even ignoring the legal one, imagine you being in his shoes: you earn money by selling your invention, and for convenience to your clients, you sell them your code. Then somebody asks you to allow him to include your code in some other project which is published under some other "open source license." You agree not knowing what that actually means, for example, because he who asked used misleading language. Later you find out that you misunderstood the effects of that action. Morally, the agreement where one agrees only because he was mislead is certainly not the agreement which has to be enforced. It should be annulled.

> You agree not knowing what that actually means

Exactly.

Firstly, kudos to project maintainer for handling the situation gracefully. I respect that he respected M3 creator's wishes.

You're quite right in an ideal universe. Ours is far from that. The code for the project is out there, so if someone does not want to pay the programmer, then they will reimplement it in house and no one will ever know about it period. If on the other hand they prefer the port, then there must be a good reason for it such as better compatibility/maintainability/performance/etc.

Now let's go to the ideal universe. Well, if everyone abides the rules, then the project creator effectively contaminated other programmers. If someone interested in solving the same problem examined the code thoroughly, then they are influenced by the solution and that is perfectly reasonable because they looked at the code thinking it is under BSD licence. Now that they know they are influenced by the program, they are ethically obliged to avoid solving the problem or they would be infringing author's "copyright" (that is despite the fact that you cannot copyright an algorithm).

The fact that other things are protected by patent is irrelevant, because it appears this creator has no patents for this project.

It's hard for me to imagine myself in the shoes of someone with views on intellectual property which are fundamentally very different from my own. I believe that views on IP like his harm the software community immensely. The fact that his specific take on IP (and his interpretation of relevant US law, which may or may not be accurate) is one which he believes benefits himself, combined with the "put yourself in his shoes" argument, is not sufficient for me to change my views.

He has applied for a patent but appear not to have been issued yet: "System and method for phonetic representation"

https://www.google.com/patents/US20090043584

EDIT: Downthread someone looked it up and determined that it failed to issue

Even with all what you say, do you really believe that there should exist any agreement that can never be annulled, no matter why? That if you agree with something not knowing full effects, somebody actually has the right to find you "guilty" for not knowing better? Do you really agree with the world where everybody who ever manages to trick somebody else is absolutely guaranteed to win?

> do you really believe that there should exist any agreement that can never be annulled, no matter what?

Definitely. I believe that most everyday transactions, like when I buy a loaf of bread from a store, should be like that. In general, most simple transactions that occurs within a very short time period without any stipulations for future action from any present party need not be annullable. My claim is that code licensing is an example of this of course any given license only applies to the state of the code when it was released with that license.

Licensing is certainly not an "everyday transaction" by any means. Ask Google and Oracle.

Well, I am sorry to have misunderstood the fact that a BSD license swears away any rights of control. In practice it turns out to be nearly indistinguishable from Public Domain, so what's the point? Let's just call it Public Domain and quit pussyfooting around.

Why the outpouring of hostility and contempt? I'm sure I sounded 'arrogant' - perhaps I was not too pleased with the tone of voice of many of the commenters here either.

Yes, I am perfectly aware of the difference between copyright and patent. What I was not aware of is that the copyright notice required by BSD is useful only in asserting the right of the copyright owner to swear away virtually all of the protections normally granted by a copyright notice. It is a legally castrated copyright.

"then he basically forbids other people from using the efficient algorithm" Did I ever say that I was trying to place any restrictions on the use of Metaphone3.java as is or however you might want to privately re-implement it? I donated the my work in the open refine project and it is there for all to see. I had a notion that by allowing the source code to be made available I was doing something good that would make Open Source advocates happy!

I'm glad you find my work useful enough to make a fuss about. Metaphone 3 is the product of my work alone - all the research and resulting 'rules' and logic were developed by myself from primary source data. Metaphone, Double Metaphone, and Metaphone 3 are not derivative works.

Again, my apologies if I misunderstood your "right" to do whatever you want with my work without restriction. Your method of asserting your rights, however, has certainly done nothing to make me feel welcome in your "community".

> Well, I am sorry to have misunderstood the fact that a BSD license swears away any rights of control. In practice it turns out to be nearly indistinguishable from Public Domain, so what's the point? Let's just call it Public Domain and quit pussyfooting around.

Yes. BSD is pretty close to public domain. That's what it's for.

1) It's not a "legally castrated copyright", it's a tool that we use to enforce a behavioral pattern that encourages growth of the commons.

2) Why should we feel "happy" when you've pulled a bait-and-switch, and acted as a bully? Your behavior and mindset are grossly at odds with the purpose of open source. Your effort was a trojan horse that's disruptive to the community. In short: You literally missed the point of "open source" entirely.

3) You explicitly stated that the algorithm is "not in the public domain", so falling back to a position that one can merely use Metaphone3.java is disingenuous. Java source is not useful or suitable to those of us in the Ruby world. We don't want to have to deal with the added overhead of a JVM.

4) Don't expect to be welcomed when you act like a bully, because you didn't know what the hell you were doing. It's not our job to teach you what the BSD license means, or what the difference between a copyright and a patent is.

5) Don't let the door hit you on the butt on your way out!

would you say that you are representative of the type of person i would tend to encounter in your ruby open source community?

No, they're vastly friendlier, by and large -- as demonstrated by the project maintainer of text being nice acceding to your demands despite having precisely zero legal obligation to do so, and despite you being an utter jackass about it.

I, on the other hand, am seriously considering taking Metaphone3.java and reimplenting it in several languages, then making those publicly available.

if you have any questions or comments about the code or algorithm, my email address can be found on my company website

Oh, hey you're in the thread. You're really confusing matters by using the term "Public Domain" in a way that really doesn't mean what you are trying to say (though I'm not at all sure what that would be). Please see the post I made elsewhere in this discussion that may clear up matters a little: https://news.ycombinator.com/item?id=8792829

(FYI, when I speak of "rights" in this post, I mean "legal rights", which I personally distinguish from the ethical ideas of "right" and "wrong"--I'm just explaining legality, there is no "ought" in what I say here)

> What I was not aware of is that the copyright notice required by BSD is useful only in asserting the right of the copyright owner to swear away virtually all of the protections normally granted by a copyright notice. It is a legally castrated copyright.

Um, what did you think a license would do? That last sentence, it's pretty much exactly the definition of what a copyright license does (if a little strongly worded).

Copyright is automatic. It only affects works (in your case, written code), not ideas (such as algorithms). You start out with all these rights about copying, distributing and broadcasting that particular work. These rights are exclusive, they belong only to you. That is default copyright, when you don't license anything.

Indeed exactly what a license does, is "castrating" (if you will) your copyrights. You can (partially) grant these rights to others, under certain circumstances, thereby diminishing the value of your own exclusive rights. That's what it does, a license defines under what circumstances you give up your right to exercise your copyrights, with respect to other people.

You start out with full exclusive rights, and a license formalizes how much and under what circumstances you are willing to chip away those rights and share.

Yet again, I must stress, copyright only applies to a "work"--a legal term, in this case referring to the written code of your project, and in general specifically not referring to ideas.

Thank you for your explanations. Actually, I understand the difference between copyright and patent very well thank you, and i also understand what public domain means. Unfortunately a number of people have misunderstood me to think that i asserted that copyrights protect algorithms.

threedaymonk behaved as a nice, well-intentioned, polite person ought to. however, the copyright claim was completely bogus as was the assertion of control over the algorithm itself.

i think we can all agree that being nice and polite to each other is a good thing, as is respecting the wish's of a fellow engineer out of professional courtesy. we shouldn't be confused about the law though. none of those copyright claims were valid.

> threedaymonk behaved as a nice, well-intentioned, polite person ought to.

I work with threedaymonk and the reason that he came across this way is that he is a remarkably nice, well-intentioned, polite person.

The best way to deal with this is to quietly take this person aside and explain to them how this stuff works.

For one, people likely aren't going to be as charitable the second time he does this, and, for two, unless and until he understands what he's doing with licenses, he's a potential land mine to any project he contributes to.

CS courses need more courses on how licensing works, and by "more" I mean "some non-zero number". It's useful to know even if you never enter the FOSS world.

I disagree. While knowing what these licenses are, how they work, and some pros and cons of each is nice, it's decidedly not computer science. There's no traditional CS courses I can think of that this would fit in, except for maybe an ethics in CS course, which not all schools have.

Scientists have an obligation to understand the legal, ethical, and social context of their work. So if no "traditional" CS course would fit a discussion of software licenses and copyright, a new course is needed. This is even more true for software engineering, but even a scientist who only knows science is in danger of being exploited or of damaging the reputation of his/her community.

This can easily be accommodated in a Software Engineering course. Project (and library) licenses are as important a matter as project lifecycle.

I believe most CS students would benefit greatly from this type of class. The topic may not perfectly fit into pure computer science topics but it's something that anyone who writes code will run into during their career and should have a basic understanding of.

Another (albeit more absurd) example is version control. You can successfully argue that it's not a computer science topic, but anyone graduating with a CS degree will be much better off in the real world having learned about it.

"While knowing what these licenses are, how they work, and some pros and cons of each is nice, it's decidedly not computer science."

Reusing other people's work(1) is decidedly computer science.

Without it(access to source code) it would be simply impossible to teach ANY CS course. Good luck teaching compilers without accessing one, or OS without being able to read the internals of one.

(1)Like Linux,network stack, LaTex, gnu tools, llvm, valgrind, qt, kde, WebKit engine, grass, OpenStreetMap, Wikipedia, Arduino, OpenScad, freecad and RepRap...

Even if it isn't part of theoretical CS, it's certainly part of the practice of programming, which university-level CS programs claim to teach.

> There's no traditional CS courses I can think of that this would fit in, except for maybe an ethics in CS course, which not all schools have.

All CS programs should have an ethics course, at least, and I think this would work in that context. For example, deontological ethical reasoning is about navigating among conflicting duties, and licenses can certainly impose duties on people.

an engineer should be educated about the legal and ethical aspects of her/his field of work. these are pragmatic concerns, and engineering is a pragmatic discipline.

while it might not fit into an academic computer science research curriculum, a good university program should include courses on these topics anyway. if your school doesn't (or if you didn't go to school!) its still a good idea to learn about law and ethics through self study.

During my CS studies I heard a course about my local equivalent of copyright and patent law and while it was only a few hours it greatly helped me to understand the implications of licenses and when to ask a professional for help.

I think basic knowledge of laws concerning your work of field is an important asset to anyone working in tech. Especially if your laws diverge from US ones (e.g. no "fair-use") the internet is a dangerous place sometimes ;)

The package maintainer, as you said, handled this very gracefully.

Yeah, the author is likely in the wrong/gray (IANAL). But I can't help but be disappointed that posters below immediately jump to "Well, it should be free anyway!" "Why is he protecting his code?" "Here is an evil oppressor trying to keep something valuable to himself."

Here is a developer who tried to build a living using his own software, something many of us aspire to do. Why are we so quick to throw him under the bus? Is it because we instinctively always want to "defend the underdog"?

> Here is a developer who tried to build a living using his own software, something many of us aspire to do. Why are we so quick to throw him under the bus?

Because he published it under a license he apparently never intended to honor. It's like trying to charge someone for a free sample after they've eaten it.

He was mistaken about what that license meant, not dishonourable. He intended to honour what he thought the license meant, and wasn't performing some sort of shell game. Threedaymonk took the best course in honouring what the original author intended, rather than the mistakenly-applied licensing.

FOSS licenses are confusing to people who don't live and breathe them. One other example from quite a few years back was when SugarCRM open-sourced their code and put it under a GPL license. They thought it meant 'you get to see what we're doing', but after a couple of competitors forked off it (we used vTiger, from memory), they got a short, sharp lesson in 'what the GPL means'. Their code from that point on was not GPL'd.

> (b) it's an illustration that the rules of intellectual property and open source are not always clear, even to those following well-established patterns.

I don't think there is anywhere on the internet that will tell you that you can't copy a BSD licence. How did the original author decide on the BSD?

The problem seems to be that he didn't really decide.

It's easy to make a quick decision when releasing software that cannot be taken back.

And when you're finally realizing your mistake, you can only resort to what we're seeing here: pleas and threats.

And of course, on the whole huge Internet you will always have people who defy you either for legitimate reasons or just out of spite. So it's practically guaranteed that you end up with war.

I cannot make up my mind whether we should blame him (for being so nonchalant with the license) or Google (for luring him into this trap without properly educating him on the consequences).

Blaming Google is, in my opinion, patronizing the author. There's no evidence of them having tried to deceive him, and the BSD is a very well known license.

I sympathize with him, but that doesn't mean there must be someone else to blame.

If you wanted something like what he wants, are there any licenses that limit the scope of derivatives? Or are you best off using a No Derivatives license with a contact for permission notice?

Someone could still reimplement the algorithm based on a description. You can't copyright algorithms. You can (sometimes) patent them, which the original author apparently tried to do. The claim would have been based on that patent rather than on copyright.

(Edited some details I was wrong about)

>assume the best of a fellow engineer

Quit calling yourselves that: you're all web developers, not engineers. At the end of the day, 95% of what people on HN do revolves around making pretty websites for the internet.

You have not taken the FE exam, you did not study engineering, you are not certified as an engineer. This is just the latest way that CS graduates are trying to self aggrandize and de-trivialize their profession.

I would say "making pretty websites" is still engineering because they are working with machines to design and manufacture [0]. This makes anyone who does it an engineer.

I find it rather offensive to generalize everyone on HN into one category. I have read comments from people in all stages of life and many different professions and we do not need this kind of talk.

[0] http://www.thefreedictionary.com/engineering

There is nothing to stop a software developer from taking the "other disciplines" FE and the "software engineering" PE.

Absolutely, but in the context of HN's "silicon valley hacker" culture, that seems pretty unlikely here.

How much of the hardware and software you're using now was originally conceived and designed by credentialed PEs and FEs?

You can walk up to Woz and tell him he wasn't an "engineer" when he designed the original Apple hardware, but you can't possibly expect anyone with a clue to take you seriously.

Unless you are in some specific locations such as Canada, where the term "engineer" is tightly controlled, then you can call yourself whatever the fuck you want. "Software engineer" does not imply any of the things you said in the United States.

I can call myself a software ninja if I wanted to. That doesn't mean I practiced the art of ninjas. I can call myself a code monkey. That doesn't mean I'm actually a monkey. It's just a fucking title, so get over it. If that's my actual title at a company then I'm going to use it publicly too.

Or Texas. Or, as I have heard, Florida. Now that there is a proper software engineering PE exam, maybe more states will follow.

It makes sense for safety critical areas to hold people to professional standards and legal liability.

Edit: speling

Don't forget all of Germany.

Updated to be more general, I thought Canada was unique due to some of the ceremonies engineers go through (eg. Iron ring)

NC State University has adopted the same ceremony, though it is not well known and leads to really weird conversations occasionally, until I mention that the ring does not mean I'm homosexual.

No, and the US has an "Order of the Engineer" modeled after that. It always seemed silly to me to wear a ring on your working hand when the first thing you do when working with your hands is to take of jewelry.

"Engineer" is a protected job title in the US, but most are able to operate under an industrial exemption and so legally use the title.

The reason why some are picky about the title is that it is a licensed profession, and has some expectations associated with it. If someone tells me they're an engineer, I assume they know how to use a Laplace transform, because that's common across disciplines at the undergraduate level. Same as applying perturbation theory at the graduate level. Same as having a common corpus of basic science knowledge. Same as with professional code of ethics. Etc.

To use your own logic above, calling yourself a software ninja doesn't make you a ninja, and calling yourself a software engineer doesn't make you an engineer. It might be legal, but it's crass.

> Laplace transform, because that's common across disciplines at the undergraduate level

At your specific institution. Every school has different curriculum, both within the engineering programs themselves, and when compared to other schools. There is no unified curriculum nor is there one unified thing that all engineers, regardless of their schooling, would know because I'm sure, at one school or another, that one thing would not be taught.

And what, exactly, is so crass about it? As I said, it's just a goddamn title. A made up word used to describe yourself. There's nothing inherently disrespectful about using a title. It doesn't instantly diminish the title of every other engineer on the planet just because someone decided to call themselves an engineer without getting a B.AS. I could call myself the Queen of Rotunda, my made up domain. That doesn't suddenly mean that I'm disrespecting Queen Elizabeth, or any other queen in the world.

I'm not sure you're informed on this topic.

The whole point of having ABET accredit programs is to standardize a basic body of knowledge. Curricula in mechanical, chemical, electrical, and biomedical all include solving DEQ and PDE systems. Solving those without Laplace transforms is cruel, once the basic principles are understood of course.

If it's just a goddamn title, then you wouldn't be so insistent on using it. You want to use it because it implies a level of education and professionalism. It's crass to use the title because there is such a thing as software engineering, and the people most strongly insisting to use the title don't practice it. Like I said, you can legally use it, but licensed engineers are rolling their eyes when your back is turned.

The very fact that you have a body accrediting programs means that there will be many schools serving up engineering degrees that are not accredited by ABET or one of its member societies, especially internationally. Are these people suddenly not engineers just because a licensing body has not bestowed their program the honour of accreditation? Just because there's an accreditation board does not suddenly make it the only ruling body on who can use the title "engineer". Unless it's actually illegal to be an engineer without being accredited by ABET, then their word really means jack shit.

> there is such a thing as software engineering

Which differs from place to place, yes. Not every program has been accredited but still provide software engineering degrees, perhaps with elements that would not be found in another's curriculum.

> and the people most strongly insisting to use the title don't practice it

Just because you don't have an accreditation does not mean you "don't practice it".

> Like I said, you can legally use it, but licensed engineers are rolling their eyes when your back is turned.

Good for them. They can roll their eyes all they want, it doesn't make them more right to do it.

Sort of; they were likely never engineers in the first place. Engineer is a licensed profession, like physician or lawyer. You can graduate law school, but that doesn't make you a lawyer; passing the bar does. In the US, it depends on the state, but the easiest path towards licensure is to graduate from an ABET-accredited undergraduate program. Some allow you to proceed to licensure without that, but it requires extra experience to compensate.

ABET does accredit international programs.

> ABET does accredit international programs.

Some international programs. They don't have nearly enough capacity to accredit every single international engineering program all over the world. I'm sure they have the major ones down, but there's plenty that are probably not which still teach engineering.

> Engineer is a licensed profession, like physician or lawyer.

The license only gives you the right to call yourself a Professional Engineer. Like I said before, unless it is illegal to call yourself an engineer without being a PEng and being licensed, then the "licensed profession" part of it really does not matter. A company can require you to be a PEng in order to receive an "engineer" title in your role, but there is still nothing, short of law that prevents you from using it yourself.

> like physician or lawyer.

And they're both different from engineer, because those licensed professions are actually legally enforced, whereas "engineer" is not legally enforced everywhere.

No, the license allows one to offer engineering services to the public and the PE title is granted in recognition of that right. It is generally illegal to call yourself an engineer, but most people are able to use the industrial exemption. Nevertheless, in using the industrial exemption one is not offering engineering services to the general public. I don't know what happens if you tried to use the industrial exemption but offered services to the public, but I think it's prohibited.

This all varies between states, but in general, if you for example started a software consulting firm called Vlad Engineering and the principals were not licensed, your local licensing board would likely have beef with you. If you were a software engineer at Vlad Consulting you would probably be okay.

The bottom line is that it's tricky to navigate, and so it was recommended at my school we avoid using engineering anywhere unless we had our PE license.

http://www.growthengineering.co.uk/

"Engineer" is legally enforced everywhere. If you don't believe me, try to get a building built without having a licensed Civil Engineer sign off on the blueprints.

That has nothing to do with enforcement of calling yourself an engineer. That's saying that people enforce who can build buildings.

> Curricula in mechanical, chemical, electrical, and biomedical all include solving DEQ and PDE systems

Civil?

Don't know about them.

It says Master of Engineering on my diploma, I can call myself an engineer if I want to.

Reply on news.ycombinator.com