[legal-discuss] Trivial contributions and CLAs
rfontana at redhat.com
Wed Apr 23 00:32:24 UTC 2014
For anyone on this list not accustomed to looking at such things, I
think it might be interesting to point out what this patch actually is
and what Stefano means by triviality (even though the CLA may not be
the relevant issue in this instance, the issue of contribution process
around trivial patches is the larger issue that Stefano was raising):
The patch would cause one existing line in one file:
options = sorted([(ip.id, ip.ip) for ip in ips if not ip.port_id])
to be replaced with this:
options = sorted([(ip.id, ip.ip) for ip in ips if not ip.port_id], key=lambda ip: ip)
That is: all this patch does is add the following text to one line of a file:
", key=lambda ip: ip"
The file itself contains about ~100 lines of code, and Horizon, the
relevant project, contains, I believe, about 2000 files.
> I have been notified of another very small patch that is left in a
> limbo, with the author not allowed to sign the CLA and the developers
> stuck in unknown legal territory. You can read more about it on
> From what I can see, the patch is trivial and shouldn't even be copyrightable but the person spotting the issue and fixing it is not comfortable signing the CLAs. Can any other developer copy the patch and put it into our trunk? Until when is this sort of behaviour safe?
> We're getting more of these small blockers and I think it's already a problem. Having to sign a Corporate CLA and Individual CLA for a trivial patch, from an operator (whose job is to run clouds, resulting in small and rare patches, not to develop large features) can conflict with our effort to get more operators involved in OpenStack.
> I'm not sure what solutions are available. If we can't change the CLA processes easily, what else can we do to get small contributions like these?
More information about the legal-discuss