Amdahl's law and non-proliferation

Jason White jasonjgw at pacific.net.au
Wed Apr 13 07:03:35 UTC 2005


Forrest J. Cavalier III writes:
 > 
 > 4. I hold that the fraction of the 11% which is "easily influenced" is small
 >     because....most authors have some good reasons for choosing outside the
 >     "big 4".  "Good reasons" are not easily countered.
 > 

If major distributors/integrators decide only to disseminate code
available under licenses recommended by OSI, and to the extent that
developers and users take OSI approval into account in deciding
whether to contribute to or use particular software, then OSI policy
will place pressure on these remaining projects to adopt one of the recommended
licenses. Still, that might not be sufficient to outweigh the good
reasons for choosing a different license in the first place.

 > 5. That leaves some small fraction of the 11%: those who did not have a
 >     good reason. Perhaps they can be easily influenced, maybe they might
 >     even become aware of the OSI's new policies.
 > 
 >     Unfortunately, if an author hasn't thought enough about licensing
 >     to form a solid opinion on licensing based on "good reasons", they probably
 >     don't release much software anyway.  They aren't relevant.

I agree they are more likely to be responsive to OSI policies. Note
that their relevance should be judged not merely by the amount of
software they develop, but by the extent to which it provides
desirable components to be integrated into larger projects.
Integration issues are at the forefront of the concerns articulated in
this forum, most notably in connection with reciprocal licenses.
 > 

Even if the benefits of the policy change are marginal in terms of the
number of projects affected or the amount of code involved, both
current and future, in the absence of major drawbacks to the new
policy, its advocates could still argue that the change is, on
balance, better than the status quo.

I am not arguing for a position here, but only responding to the
analysis. The inference from "the benefits are marginal" to "the
change shouldn't be made" seems suspect, unless there is a good
argument to the effect that the new policy would also have costs of
such gravity as to overwhelm its advantages. A utilitarian analysis
requires consideration of both positive and negative utility.



More information about the License-discuss mailing list