Open Stack

On Tue, Jan 21, 2014 at 02:46:29PM -0500, Rich Bowen wrote:
> 
> I would suggest language like what you see here: http://svn.apache.org/repos/
> asf/httpd/httpd/branches/2.2.x/server/vhost.c
> 
> So, perhaps:
> 
> /* Licensed to the OpenStack Foundation under one or more
>  * contributor license agreements.  See the NOTICE file distributed with
>  * this work for additional information regarding copyright ownership.
>  * The ASF licenses this file to You under the Apache License, Version 2.0
>  * (the "License"); you may not use this file except in compliance with
>  * the License.  You may obtain a copy of the License at
>  *
>  *     http://www.apache.org/licenses/LICENSE-2.0

I can't resist using this as an opportunity for a rant related to some
things I've said before, not entirely off-topic. 

It's my understanding, and this is reinforced by the existing practice
of copyright and license notices in OpenStack source files, that there
is an understanding that a contributor is both granting a license
under the relevant CLA and granting a license under the terms of the
Apache License. I could point out that it is potentially even more
complicated than that but let's keep it there for simplicity. I've
contended that this is a form of duplicative licensing unprecedented
in open source software development; if anyone has a counterexample
I'd be happy to know about it. The reason it's unprecedented is that
CLAs exist precisely so that contributors won't be granting in
licenses under the project license (leave aside whether that's a good
or bad thing). There are tons of Apache License projects that have
contributors licensing contributions under the Apache License, like
oVirt and OpenShift Origin, but these are the projects that don't use
CLAs.

While -- as was noted today in a Twitter conversation -- ASF projects
normatively accept 'smaller' patches without a CLA (the understanding
being they are licensed under the Apache License itself), the only
thing recorded in source code is the license grant from the ASF and,
in some cases, the note that the code was largely licensed in under
(nonpublic) CLAs, as in the ASF notice that you've adapted.

Your suggestion might make perfect sense, particularly to someone
coming from the ASF tradition, but it calls into question why this
duplicative licensing appears to be taking place. This is not the case
for ASF projects. As an arbitrary example, a lot of Red Hat copyright
licenses flow into Apache Camel, but Red Hat isn't granting an Apache
License on Apache Camel (via apache.org at least), because Red Hat is
the licensor of a CLA covering employee contributors to Apache Camel
and that's the extent of the licenses it is granting in to that
project. A copyright notice from Red Hat in Apache Camel would violate
ASF standards but would be accurate. A Red Hat copyright notice
immediately followed by an Apache License grant would be inaccurate,
because the Apache License isn't coming from Red Hat; it's coming
(mostly) from the ASF.

To understand my point here, look at any other project that uses CLAs,
and you will see that there are no contributor copyright notices other
than, in some cases, copyright notices from the main CLA licensee
(here the OpenStack Foundation). OpenStack is the only exception I'm
aware of, and I believe the rationale for having such contributor
copyright notices at all is to make visible that, separate from the
CLA, direct Apache License grants are being made by CLA signers.

Your proposal is one visible way to cease the practice of duplicative
licensing. However, OpenStack developers and fellow travelers who have
criticized the CLA regime won't like this, because it signifies full
reliance on the CLA regime. 

- RF