[License-discuss] comprehensiveness (or not) of the OSI-approved list
rfontana at redhat.com
Wed May 22 14:05:03 UTC 2019
On Wed, May 22, 2019 at 9:15 AM Tzeng, Nigel H. <Nigel.Tzeng at jhuapl.edu> wrote:
> Let's clarify the history on CC0.
> Objection to CC0 was primarily you and Bruce which made it DOA regardless of the opinions of the rest of the list. There was no "quickly growing consensus" when they pulled the plug.
You're right, Nigel -- except that I can't see why that would have
made it DOA. Apart from that, I withdraw that characterization. I have
not reviewed the thread, but I do seem to recall being the person who
called attention to the patent language. I remember what bothered me
at the time was that if CC0 were summarily approved -- and yes, it
seems that that was the direction things were going in -- it would be
inconsistent with the hostile reception of the MXM license that Carlo
Piana had submitted, which I believe was based on MPL but with a
significantly narrowed patent license grant.
I was not troubled by the idea of approval of CC0 as such. Back in
those days I was a big fan of CC0. I used to recommend its use to Red
Hat developers, for certain purposes. I was known for saying "I love
CC0" (yes, I used that phrasing) so often that I remember Aaron
Williamson joking about this at an OSCON talk.
What concerned me, and I remember Carlo noting this as well, was the
possibility that OSI, or l-r, would treat similar licenses differently
based on varying sentimental attitudes toward the license submitter.
Creative Commons, in those days perhaps even more than today, was
viewed very positively in the open source community. (I feel that
today there is more distance between the CC and open source
communities.) The MXM license was associated with MPEG and more
generally with the controversial topic of media codec patent
licensing. Since you brought up NOSA, my concerns about that license
in a sense came out of the earlier MXM/CC0 experience. NASA is another
entity that, I had come to perceive, was viewed very positively in the
open source community for what seemed to be, at root, sentimental
reasons connected to NASA's relationship with tech culture (I was
seeing this at the time in the OpenStack project, whose early
popularizers made much of its NASA origins).
At least from today's perspective, we saw the problem play out a
couple of years later with the UPL submission. The hostile initial
reaction to UPL, on l-r and elsewhere, was obviously connected to
general community hostility towards Oracle, especially during that
So you've motivated me to say this: I think OSI should dispense with
the whole idea that it should passively react to any supposed
consensus that emerges from license-review. OSI has a responsibility
to determine whether a license meets the OSD and provides software
freedom regardless of what direction the l-r discussion is going in.
One reason for this is the history of inconsistent attitudes on l-r
towards submitted licenses based apparently on views of the license
> What was happening behind the scenes on the OSI board I cannot say but to say that “the Committee felt that approving such a license would set a dangerous precedent” is a significant overstatement.
> The fact remains that CC0 is a widely used Open Source license approved by CC, FSF and the USG. OSI approval, or lack thereof, has had minimal impact on its use.
I think things are more complicated than you suggest. The discussion
of CC0 (and the inconsistent earlier discussion of MXM) exposed a
policy problem surrounding open source and patents that had previously
been ignored or misunderstood. Or perhaps it showed the beginnings of
an evolution in community understanding of the issue.
One of the major open source policy issues of the present day, at
least from my standpoint, is the controversy over open source in the
FRAND licensing context. A couple of companies have been promoting the
view that (a) "copyright only" licenses that explicitly withhold
patent licenses can be open source, (b) widely-used permissive open
source licenses that do not explicitly mention patents should be
interpreted as "copyright only". The OSI has in recent years been
fairly clear and activist in its opposition to this campaign. The
approval of CC0 would have been inconsistent with the OSI's own
emerging view on interpretation of the OSD in relation to this issue.
More information about the License-discuss