Copyright assignment

[mjg59]

The fundamental problem with projects requiring copyright assignment is that there's an economic cost involved in me letting a competitor sell a closed version of my code without letting me sell a closed version of their code. If this cost is perceived as larger than the cost of maintaining my code outside the upstream tree, it's cheaper for me to fork than it is to sign over my rights. So if I have my own engineering resources, what rational benefit is there to me assigning my copyright?

Comments

Depends on your focus

[http://gould.cx/ted/]

Your engineering resources are working on something entirely different and you don't want to waste your time on maintaining the fork. So it's then cheaper for you to contribute the small fix in your platform and move on.

Re: Depends on your focus

[mjg59]

Yeah, for trivial patches that's certainly true. More substantial ones imply that you've got an ongoing interest in the project, at which point it's less so. And if there are several different entities in the same position, you're in a position to pool that maintenance. So it seems that the cases where copyright assignment is most going to benefit the upstream (ie, large features that would take a lot of work to reimplement) are the ones that are least likely to be granted.

Re: Depends on your focus

[http://gould.cx/ted/]

That's certainly true. But, if you're planning on contributing large amounts of resources it's probably the situation that you can negotiate with the person asking for CA. Depending on the situation, you might be better off paying them to do that work instead of doing it yourself as they'd have an understanding there. Of course, that is *highly* situation dependent.

I think there is also many cases where creating a new organization makes sense. I think a good example of this is Linaro which is staffed to do "general ARM work" where the individual chip makers then have their own staff provide specialized features for their individual SoCs.

At the end of the day, business agreements are more confusing than why we have so many programming languages :-)

Re: Depends on your focus

[(Anonymous)]

So copyright assignment is only sensible for projects where the only community contributions you encourage are one-liners?

Re: Depends on your focus

[cjwatson]

Arguments along the lines of "you might be better off paying them to do that work" often seem to rest on the idea that developers are interchangeable resources; I think we both know that isn't generally true. If you (as a company) are planning on contributing large amounts of resources to a project, then it may well be the case that you already have somebody with significant skills and/or enthusiasm that could be brought to bear, and those can't easily be transferred to another project nor brought into existence in the company you were planning to pay.

[stewart [launchpad.net]]

A contributing factor/problem is that for contributors working for most non-small companies this means talking to a lawyer. This greatly increases the amount of effort you have to go through to contribute a relatively small amount of code, thus making forking it yourself even more attractive.

I've also had company lawyers balk at signing the exact same contributor agreement that the company asks people to sign but with s/our company/other company/.

[https://profiles.google.com/askutt]

In most truly large companies you have to talk to a lawyer to contribute software, period. "Work for hire" rules nearly universally apply to software works.

[(Anonymous)]

For those of us in some companies, there's a related problem: I am permitted to contribute back under the license(s) I receive the code under by my contract of employment, provided my employer keeps the copyright. If I misuse this permission, I face internal disciplinary procedures, up to and including termination of employment and criminal charges, depending on just how severe the misuse of that permission is.

I do not have the power to sign a copyright assignment on behalf of my employer. If you want me to do that, I have to go to the Board of Directors and make the business case for signing this contract. Guess what I'm not inclined to do, ever?

Copyright Assignment in Germany?

[(Anonymous)]

I always wonder how Copyright Assignment legally works in countries like Germany, that don't have the concept of copyright like the US does. In Germany, according to the "Urheberrecht", the ownership of the work is bound to the author and cannot be re-assigned. After all, he is the author...

So no matter what piece of paper he signs, legally nothing can take away the author's ownership of the IP, so while he can assign usage right to third parties, he will always retain his own rights to the things he wrote, including the right to relicense.

Or so is my understanding, at least.

Re: Copyright Assignment in Germany?

[(Anonymous)]

Not quite. You can't ever give up authorship, because, well, you are the author, that isn't something transferrable. You can give someone an exclusive usage/distribution license, which would be the equivalent of signing over copyright. Being exclusive means of course that you can't license it to someone else.

What you can't give up/sign over is the rights directly tied to authorship, such as the right to be named as the creator, or the right to not be named as the creator, and a few more. There are some more restrictions on authorship rights for software written as an employee as part of your job, IIRC.

Besides, Germany is not a rare case here. In Europe authors' rights would be basically everywhere but the UK, I think. Copyright is a system born from the UK legal philosophy and is thus found mostly in those countries whose legal tradition is influenced by it, such as the USA.

[(Anonymous)]

And your fork has the advantage that it can take code from the upstream tree (as per licence) but the upstream tree can't take your code without dropping their copyright assignment requirement (or maintaining a seperate "copyright pure" fork for their closed releases).

If there are significant engineering resources outside upstream then that leaves upstream at a disadvantage and at risk of not being considered upstream any more.

Copyright assignment only makes sense for either side when those requiring it are (and will be) doing the vast majority of the work.

There may be other considerations (such as ease of relicencing to a different open license rather than releaseing a closed version) that make copyright assignment attractive in some cases but some licenses accomodate for that fairly well in themselves.

Nice case. Even stronger…

[mirsolutions.de]

… is the initial effort people have to go to. As an integrator, for example when working on BSD ports, or doing “mass” work on Debian packages, or one-off work on a package that doesn’t generally interest me, it’s always the same for me (and others): if I have to register an account at a bugtracker, or subscribe to a mailing list, or similar, in order to submit a fix, it’s too much effort, and the fix will be kept local, unless upstream takes it on their own time. (If I had a way to notify upstream, I’d use it. Or, I use it and upstream doesn’t respond.) A copyright assignment is just so much more trouble. I admit I have it standing with the FSF, but I’m very unlikely to do it with another organisation, no matter how good they present their case.

The point may be moot ...

[david.dasz.at]

... if the assignment is to a community instead of a company. No? e.g. if the assignment is conditional on a free-only relicensing policy, it would defuse your argument.

[simont]

Does that even make sense? I mean, copyright is precisely the right to be the final arbiter of what licensing may or may not be done on the copyrighted work. To assign somebody copyright, conditional on them only licensing it in certain ways, sounds a lot to me like "not actually assigning copyright after all, just granting a licence to redistribute under certain conditions".

Not really

[(Anonymous)]

It would for instance give them the right to enforce (at law) violations of the copyright on the code assigned. Right?

It would give them the right to put the code under a different Free license as per the agreement without having to get your further permission. Right? Say you had dropped out of sight for some reason and could not be contacted.

all the best,

drew

Assumptions...

[https://profiles.google.com/askutt]

Your premise isn't ipso facto true: they may only sell the closed source version to markets you don't have access to, or in markets in which you cannot compete. Few companies have the engineering resources to compete with Apple or Microsoft, if their intent is to put your software in OS X or Windows. Even when that's not the case, you may not actually be giving them an competitive advantage. For example, I don't develop print servers, so giving away copyright on patch to a print server really doesn't impact my business any. For lots of people, this is the case: the entire reason I downloaded your project in the first place is it does something I (my employer) cannot be bothered with.

Though yes, if you're in "competition" with upstream and have the resources to make a fork palatable, the value of assignment is minimal. You're left with externalities, like the political fallout of a fork.

website

[(Anonymous)]

distributed version control should make it very cheap to maintain a fork.

it would be great if there were a good hosting site (like sourceforge) specifically designed for keeping forks. maybe with support for auto pulling upstream, adding the patches, and making packages.

then anyone could make their own 'go-oo' type project.

at some point the company who wanted the copyright assignment would be so tempted by all the lovely patches in the fork, that they would have to drop the requirement.

Re: website

[mjg59]

It's very easy to maintain a fork up until the point where you have conflicting functionality, at which point it becomes a nightmare.

[reddragdiva]

I see what you mean (and wonder what in particular triggered it), but is it different assigning copyright for GNU contributions to the FSF than, say, OOo contributions to Sun was?

[mjg59]

I think so? The FSF commit to not doing anything that'd give them a competitive advantage over you - if they change the license then you get the code under the new license as well. There's still some cost to you submitting, but it's not as potentially large.

[reddragdiva]

Yes, that's what I thought. In game theory terms, you can be pretty sure the FSF is not going to defect; whereas the whole purpose of assigning copyright to a commercial entity who do a proprietary version is so they can defect.

(It was particularly egregious in the case of OOo, given it was LGPL/SISSL and they didn't need to hold copyright to release a proprietary fork anyway.)