Re: Attorney fees

[Tim Smith]

In article <85bq13p9dh.fsf@lola.goethe.zz>, David Kastrup <dak@gnu.org> 
wrote:
> > What would actual damages be?  Typically, this is the profits you
> > didn't make, because people bought from the infringer instead of from
> > you.  But when you make your product available for free, you don't
> > really have lost profits.
> 
> But we are not talking about "making a product available for free", but
> "making it available under the conditions of the GPL", whether for free
> or not.  If the perpetrator could have made use of the software under
> those conditions, he would not have to violate the license.  So we are
> talking about a _different_ use then the one that is being licensed
> under the GPL.  A product that has _not_ been made available for $0.  If
> the author has not yet made this product available (namely a licensing
> under conditions the perpetrator would have used), then he is, of
> course, free to name the price he would take for said product.  Or

That's not going to fly.  It's too remote.  To determine damages under 
that model, the court has to figure out what P and D would have agreed 
to if P had been willing to sell, and D to buy, a special license to the 
work.  That's just too speculative, unless P has already sold such 
licenses to others, so there is some basis for settling on a likely 
price.

What the court wants to do is put P in the position P would have been in 
if the infringement had not occurred.  There are two ways to do that.  
The one you are suggesting is for the court to speculate what would have 
happened if the D had hypothetically purchased a license to do what D 
actually did.  The other way, which is the one I believe courts usually 
go with, is for the court to ask what would have happened if D had 
obeyed the actual license that D had.  That requires much less 
speculation.

Under that analysis, P expected to make $0 off of D's use of the 
software, so I don't think it is likely the court would award much more 
than that.

-- 
--Tim Smith