On 2020-08-30 at 05:53 +0900, Koichi Murase wrote:
> > It's a bit more complicated than that; if, for example, some
> > excerpt ended up in regression tests, there would be a question
> > about whether or not there was a copyright violation. As I
> > understand the GPL (IANAL), it requires all parts of a "work" to be
> > GPL'd, and that wouldn't be possible for any parts of the script
> > that ended up in bash regression tests.
>
> That's an interesting discussion. I don't know how you define the
> "work", but basically GPL only affects the derivative
> programs/software but not all the "work" including the output of the
> programs or the knowledge obtained in running/developing the code.
I think it would be the other way around.
Suppose someone posted here:
> I found this nice telepathic shell that works better than bash:
> /* Copyrighted by Random Guy. All rights reserved */
> int dwim() {
> ...
> }
and then Chet ripped it copying that code into bash (or readline),
_bash_ would be infringing the copyright of "Random Guy".
Not the other way around.
Second, it is not true that "it wouldn't be possible for any parts of
the script that ended up in bash regression tests" not to be GPL.
A shell script is run by the shell using OS standard functions, and no
part of it is linked with GPL code. As such, you may have a shell
script under any license you wish, even if your shell is GPL. A
regression script is no different to that (and noone brought up that it
should be included in the regression tests, even!).
So you could for instance have a GPL shell and regression tests under a
different license, such as BSD. You could even have a script testing a
regression under a license such as CC-BY-SA 3.0, which is incompatible
with GPL.
Having a mixture of licenses would be undesirable, as it's simpler to
state "everything on this repository is GPL-3.0" or, "the test folder
is under license X and the rest is under license Y". But possible? It's
perfectly possible and legal.
The point that could be made is that bash regression tests are one work
on its own, that it is GPL and that adding your test to the regression
tests links them in a way incompatible with the GPL (I don't think this
would be the case, but ianal). As before, the copyright violation would
be to add such code to the regression tests.
And about test cases, as mentioned in the thread, a good test case will
be short and clear, testing one feature (or bug). As such they are
generally very simple. And often ineligible for copyright.
Best regards
