Re: DFSG-freeness of Apache Software Licenses
[debian-legal folks, please cc: license@apache.org. Thanks.]
Anthony DeRobertis <asd@suespammers.org> writes:
> On Sun, 2003-11-16 at 12:15, Joe Schaefer wrote:
> > Anthony DeRobertis <asd@suespammers.org> writes:
> >
> > > On Fri, 2003-11-14 at 14:02, Roy T. Fielding wrote:
> > >
> > > > The act of running the Program is not restricted by the proposed
> > > > Apache license either. We don't need to list all of the things
> > > > that are not restricted by the license.
> > >
> > > It is if your patent license is revoked.
> >
> > Software that is licensed to you under the GPL, for which you
> > subsequently use in a patent-infringing way, is no different
> > in this regard. You are not free use GPL software to violate
> > the law, not even if you refuse to accept its terms.
>
> Quite. However, with GPL software, I don't have to worry about the
> copyright holder claiming patent infringment against me for patents used
> by the software as distributed by him. If my patent grants under the
> proposed Apache license are terminated, I do.
The freedom to worry is also unrestricted under the proposed license,
although if anything is to be gleaned from the SCO circus, it's that
just about anybody can make shit up and threaten the freedom to
use/copy/modify what was ostensibly free software. However, the "fair
use" provisions of copyright law, as applied to software, may provide
sufficient defense for us poor Linux users should SCO actually come
after us.
Note that Section 4(a) of the proposed license says your copyright
license is absolutely irrevocable, so it does give you a viable
means of defense against an infringement suit from a contributor.
I would imagine that clause [1] is a large reason why the FSF
characterized the IBM Public License as a free license [2].
> Please see my full rationale in
> Message-ID: <[🔎] 1068968797.3951.14.camel@bohr.local>
> which is archived at
> http://lists.debian.org/debian-legal/2003/debian-legal-200311/msg00145.html
>
>
> > My interpretation of the FSF's comments on the license@ list
> > are that, in their view, a GPL-v2 licensor _implicitly_
> > forfeits any patent claims that would apply to the licensed
> > software. However, that fact alone does not imply the proposed
> > license is _incompatible_ with the GPL-v2: an author of a
> > derivative work may be willing to forfeit such patent claims
> > without insisting that the authors of the original work forfeit
> > theirs also.
>
> I'm not following you. An author of a derivative work can certainly add
> exceptions to the GPL, just like they often do to allow linking to
> OpenSSL. However, that doesn't make the OpenSSL license GPL-compatible.
"GPL compatibility" has a very specific meaning, which AFAICT
is entirely unrelated to the situation you offer; see
http://www.gnu.org/licenses/gpl-faq.html#WhatDoesCompatMean
Since the proposed license has a better definition of "Contribution",
one GPL-incompatibility mentioned by the FSF (wrt the IBM license) is
eliminated. There appear to be others [3], especially related to the
first sentence of Section 5, but I'm optimistic that they can be fixed
without too much fuss.
Footnotes:
[1]- It's section 2(a) in the IBM license.
[2]- The FSF's notion of free software is here
<URL: http://www.gnu.org/philosophy/free-sw.html>
[3]- Read Eben Moglen's comments on the license:
<URL: http://nagoya.apache.org/eyebrowse/ReadMsg?
listName=license@apache.org&msgNo=69 >
--
Joe Schaefer
Reply to: