|
Boost : |
From: Matt Hurd (matt.hurd_at_[hidden])
Date: 2004-12-09 18:21:52
On Thu, 09 Dec 2004 22:09:36 +0100, Ian McCulloch
<ianmcc_at_[hidden]> wrote:
> Matt Hurd wrote:
>
> [...]
>
> > I saw another interesting note about GCC and type traits the other
> > day. The only probably is that it is a one way door to the GPL world
> > from boost due to the GPL licence restrictions, even looking at the
> > GCC implementation and implementing any improvements is questionable
> > :-(
>
> Why is looking at GPL code a potential problem? I know that some software
> companies try to claim all sorts of trade secret and patent rights over
> their code, but from a pure copyright perspective, there is not and cannot
> be any problem with examining an existing code to figure out how it works
> and then reimplementing it. AFAIK it is entirely analagous to, say,
> reading a chemistry textbook and then trying out some of the reactions at
> home, or even writing your own textbook. Unless you actually plagarize
> parts of it, copyright law doesn't care.
>
> Granted, this case may be a bit more complicated if both codes come from a
> common source and are substantially identical to begin with. But copyright
> law wasn't designed to cover such cases anyway. It might be worth asking
> the FSF (as the copyright holder) what they think, my guess would be "don't
> care".
>
> Cheers,
> Ian
Perhaps you are right, I don't know, IANAL. I'm sure you're right
about the "don't care" part. I'd imagine it is a risk thing. I
seem to remember one of the original BIOS clone vendors, Phoenix I
think, in the 80s hiring people that had never seen a BIOS before as
part of "clean room" implementation team. That's one extreme...
W.r.t. the original copyright holder granting rights, that's fine, but
other contributions, fixes, to GCC CVS and the like are not
necessarily part of that unless they are fed back. It is just a
complication w.r.t. licenses. I'd be quite happy if FSF took over the
world, license incompatibilities dissolved...
matt.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk