[openstack-dev] [horizon] Do No Evil

Doug Hellmann doug at doughellmann.com
Mon Mar 9 14:54:25 UTC 2015



On Mon, Mar 9, 2015, at 08:52 AM, Radomir Dopieralski wrote:
> On 03/09/2015 01:59 PM, Michael Krotscheck wrote:
> > On Sun, Mar 8, 2015 at 3:21 PM Thomas Goirand <zigo at debian.org
> > <mailto:zigo at debian.org>> wrote:
> > 
> > 
> >     Anyway, you understood me: please *never* use this Expat/MIT license
> >     with the "The Software shall be used for Good, not Evil." additional
> >     clause. This is non-free software, which I will *never* be able to
> >     upload to Debian (and Canonical guys will have the same issue).
> > 
> > 
> > So, to clarify: Does this include tooling used to build the software,
> > but is not shipped with it? I suppose a similar example is using GCC
> > (which is GPL'd) to compile something that's Apache licensed.
> 
> To clarify, we are not shipping jshint/jslint with horizon, or requiring
> you to have it in order to run or build horizon. It's not used in the
> build process, the install process or at runtime. The only places where
> it is used is at the developer's own machine, when they install it and
> run it explicitly, to check their code, and on the gate, to check the
> code submitted for merging. In either case we are not distributing any
> software, so no copyright applies.

Not everyone realizes that many of the distros run our tests against the
packages they build, too. So our tool choices trickle downstream beyond
our machines and our CI environment. In this case, because the tool is a
linter, it seems like the distros wouldn't care about running it. But if
it was some sort of test runner or other tool that might be used for
functional tests, then they may well consider running it a requirement
to validate the packages they create.

That's not to say we need to let our tool choices be dictated by
downstream users, just don't assume that because a tool isn't used as
part of the runtime for a package that it isn't needed by those
downstream users.

Doug

> One could argue that since the review process often causes a lot of
> stress both to the authors of the patches and to their reviewers, and so
> in fact we are using the software for evil...
> 
> We are working on switching to ESLint, not strictly because of the
> license, but simply because it seems to be a better and more flexible
> tool, but this is not very urgent, and will likely take some time.
> 
> -- 
> Radomir Dopieralski
> 
> 
> __________________________________________________________________________
> OpenStack Development Mailing List (not for usage questions)
> Unsubscribe:
> OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev



More information about the OpenStack-dev mailing list