From: Beman Dawes (bdawes_at_[hidden])
Date: 2003-06-25 21:14:35
At 03:17 PM 6/25/2003, Ronald Garcia wrote:
>...
>
>In reading the license, I think the definition of "Software" needs to be
>broadened to explicitly include the documentation, test suites, etc.
>see: >http://www.opensource.org/licenses/mit-license.php for an example.
I'll pass that on to the lawyers. I'm making a list of issues to ask them
about.
> Additionally, the BSD license goes into even more detail
>with respect to software, binaries, etc.
>
>I feel that it's good to have a human-readable license, but it
>seems more pressing to ensure proper coverage of the legal issues, even
>if it means that the license gets a little longer. In this specific
>case where the license will (hopefully) cover much of Boost, I
>presume that it will be placed in the distribution and all files covered
>mentioned by reference. In that case, it seems fine to me for the
>license to be longer and more explicit.
One of the worries the lawyers express over longer licenses is that more
verbiage offers an opposing lawyer more opportunity to find a loophole.
Notice that the new license is sometimes referred to as the "short-form"
license. There was another proposed license first, which was longer. In the
end, the senior lawyers preferred the shorter license. There was also
explicit discussion (and a write-up) on some of the other common licenses,
such as BSD. It would have been a lot easier for these legal folks to just
say "use the BSD" license. The reason they went to considerable trouble to
research and then write a specific Boost license is that they believe it
will do a better job for Boost than other OSI licenses.
Thanks,
--Beman