Interesting note on attribution and OSD #10

Ben Tilly btilly at
Thu Jan 4 12:30:00 UTC 2007

There has been a lot of recent discussion about whether various
attribution clauses are OSD conformant or not.  Opponents have done a
lot of pointing at OSD #10 (licenses must be technology neutral) while
proponents have done a lot of pointing at the Attribution Assurance
License.  (And opponents point out how much better the attribution
assurance license is, plus indicate that it was borderline then and
would likely be rejected now.  I haven't seen a good response to this
from proponents.)

Well I was looking at something else and made a curious discovery.

According to the sidebar on, OSD #10 was the _last_
addition to the definition.  No timeline is provided there, but
looking at*/
I can see that the clause was added on the website somewhere between
Oct 4, 2002 and Dec 4, 2002.  Looking at I see that the Attribution
Assurance License was approved in August, 2002.

So the Attribution Assurance License _never had to satisfy OSD #10_.
It comes much closer to satisfying that clause than the current crop
of attribution licenses do, but in no way shape or form can its
approval be taken as a guide for how the current definition should be

Conversely people who want to claim that various attribution licenses
should be considered open source have a somewhat reasonable claim.
The original definition under which the phrase "open source" was
promoted and popularized did not include OSD #10 or anything
resembling it.  Most of the community is blissfully unaware of OSD #10
and I don't remember its addition being widely publicized.  Therefore
one can somewhat legitimately question how much OSD #10 is part of the
popular understanding of "open source".

That said, OSD #10 has been part of the definition for years, must be
part of the Board's deliberations, and is clearly not satisfied by any
of the attribution licenses out there.


More information about the License-discuss mailing list