Jason Wharton wrote:

> Go back and carefully take into consideration the context of my comments. I
> was talking about the possibility of unfair snitching of sources (e.g.
> algorithms) from my code (not my public posts) in order to accelerate IBX to
> close the gap between the two products. My point was either it would be a
> long time before IBX would be able to catch up to IBO or else there would
> very likely be some underhandedness going on.

Source code is the expression of an idea, the algorithm is the idea.
Copyright protects your source code. They are different things.

> I certainly do not think this is an unreasonable concern nor is it
> unreasonable to want to protect myself wherever allowable by standard
> industry copyright protections.

Of course it is reasonable to want to protect yourself as allowed by
law. The issue is that "standard industry copyright protections" does
not protect you in the way you seem to expect. Copyright protects the
expression of an idea, not the idea itself. Patents offer a mechanism
for protecting an idea, but that protection is limited to the country in
which the patent is filed, and the USA is the only country I know of
which allows software patents in any case.

So, Jim's point is clear and valid: Your claim of ownership of a
particular algorithm is without basis unless you have a patent on that
algorithm, and even then that protection doesn't stop someone outside
the USA from reimplementing the algorithm and distributing it.

See http://lpf.ai.mit.edu for more on the evils of software patents.

A minor side issue on one of Helen's points: I don't see how software
patents harm anyone outside of the USA. A software patent is
unenforceable outside the USA, but the information about the algorithm
must be made public. Surely more information is better than less? The
crypto issues are to do with export controls and whether or not a
cryptographic algorithm is patented or not doesn't tend to cause
problems outside the USA.

Jan Mikkelsen