Legality of publishing code for a stackoverflow answer as an example implementation

Question

Consider this scenario:

• I was trying to implement my own version of some awkward functionality.

• The same functionality is implemented in at least one open source library (which allows copying).

• My code is written from scratch and not derived from any open source library.

• I wish to make use of this at work - so technically it becomes a "work for hire" and the copyright for my implementation belongs to my employer - even though I did it outside normal working hours.

• I could not get this work so I asked a question on stackflow which is basically "how do I implement this functionality?"

• The question was not answerable without reference to my implementation so I included the relevant almost functional code in the question*.

• With the help of an answer I was able to get the code working.

• The answer which helped does fully answer the question. However, I am now in a position to answer it myself.

• I would like to do so by publishing a complete annotated example as a project on github.

This would involve publishing

• a working implementation
• my unit test suite
• assigning a license - E.g. BSD

In principle this would be useable in its own right. Live code is better than 'copypasta'.

Q Is it 'safe' to do this without getting permission to open source from my employer? I would very much like to do this the proper way. However, I have been trying to get the company to define an open source policy for many years without success thus far. I hold little hope of this happening in a timely manner.

Some relevant points

• code to do this is provided in at least one open source library with a license that allows copying - so it is does not meet requirements for originality of IP
• There is essentially only one way to implement this functionality correctly - so fair use or Scènes à faire might apply
• It is sufficiently awkward to implement that a good Q&A pair on stackoverflow would be beneificial to the community
• my jurisdiction is the UK.

Assuming it is trivial (de minimis) would publishing my test suite affect that? Would it be safer to publish just the minimal example on stackoverflow (making it CC-by-SA)? How can I tell if it is sufficiently trivial that copyright rules won't apply?

What happens down the line if wish to make alterations to the open source project independently of work for my employer?

This is a general question but the one which prompted it is this one (where I was trying to implement a relative pointer type in C++, of which the offset_ptr from boost is perhaps the most well known example:

• https://stackoverflow.com/q/72356447/1569204

Relevant related questions:

• What is defined as trivial code when licensing a project
• https://softwareengineering.stackexchange.com/questions/216029/found-a-better-solution-to-a-problem-at-work-should-i-deter-from-posting-the-c?noredirect=1&lq=1
• https://meta.stackexchange.com/questions/12527/do-i-have-to-worry-about-copyright-issues-for-code-posted-on-stack-overflow
• Can I legally use code I found on a forum?

Relevant concepts:

• https://en.wikipedia.org/wiki/De_minimis#Copyright
• https://en.wikipedia.org/wiki/Idea%E2%80%93expression_distinction
• https://en.wikipedia.org/wiki/Sc%C3%A8nes_%C3%A0_faire
• https://en.wikipedia.org/wiki/Fair_dealing_in_United_Kingdom_law

*could even publishing the broken code in my question technically be considered a copyright violation?

Answer 1

You have wished to develop some code to do a particular job of work. Having read an open-source codebase that does a similar job, you developed your own code, with the assistance of others on a SE website. You wish to further develop this code as part of your paid employment, and include the result in your employer's proprietary codebase. Your employer has no clear open-source policy, and has shown some resistance to engaging with the idea. At the same time, you wish to place the work you've done on GitHub, under some kind of open-source licence.

On the basis of what you've told us, it seems to me that the code you've developed thus far is arguably copyright-entangled with the open-source library (NB I don't say it is; I only note that your admission has made your defence against any future copyright-violation lawsuit harder). Further, your code is almost certainly subject to the SE CC BY-SA licence. Any further solution to this problem you develop risks similarly entanglements.

You raise a number of legal defences, though I am unpersuaded by at least some (fair dealing exemptions in the UK apply only for non-commercial research or study, most of the jurisprudence on "only way to implement" is not from England and Wales and so not binding here, and as Bart says de minimis is a very low bar to have cleared). But the key problem with all of them is that they are defences: you have to press ahead with potentially-infringing behaviour, wait for the rightsholder to sue, then raise those defences in court. You may be willing to do that for the GitHub project (though the Boost licence is an MIT-clone, so a GitHub project under CC BY-SA or a mutally-compatible licence such as GPLv3 would be a strong possibility), but your employer may be very much less willing to shoulder that risk.

So at this point, I think you would be unwise to use this work (or any obvious derivative thereof) in a work scenario, and extremely unwise to do so without your employer's clear, written assent. If you get this assent, I would abandon any plans to publish the work on GitHub or elsewhere; it's very much an either/or situation. If you don't get the assent and wish to proceed with open-source publication, this becomes instead an ordinary licence compliance question.