[eu-gene] Psst...Wanna Buy an Algorithm? (Working Title)

[Rob Myers]

On 16 Oct 2005, at 14:20, alex wrote:

> On Sun, 2005-10-16 at 12:43 +0100, Rob Myers wrote:
>
>> So assuming that a generative *performance* has no fixed form and is
>> live, you couldn't claim copyright on it.
>>
>
> Ah, then I think this is where the disagreement lies.  In my view, the
> sourcecode is the fixed form of the composition, which is performed  
> by a
> computer.  The sourcecode is the musical score, it just happens to  
> have
> if statements as well as loop constructs.

Whilst this is aesthetically and ethically correct, the law probably  
won't see it that way.

> I think to state otherwise is to deny the programmer their rightful  
> role
> as a musician.

It's a technicality, but a useful one because of the "program that  
generates every four-note-combination" problem I mentioned.

If your code dumps a score then plays that there shouldn't be a  
problem. It's the fixed form criteria that's the problem.

> However, problems comes for me when I consider the relationship  
> between
> a musical work and a musical style...  A work is an instance of a  
> style,
> and both can be described in sourcecode.  There's danger here - it  
> would
> be too easy to sign away rights to your whole musical style by  
> accident,
> while selling rights to a particular composition.

I think the law does protect against this: certainly if I sell a  
painting of a flower I can make other paintings of flowers in the  
same style.

> I guess the trick is to keep your musical style in separate libraries
> from the code relating to a particular work, and making the boundary
> clear in contracts...

I don't think that helps unfortunately.

Do ask artlaw about this.

- Rob.