Re: [Fsfe-uk] Software patent MEP response

[Chris Croughton]

On Mon, May 19, 2003 at 02:32:30PM +0100, Alex Hudson wrote:

> On Mon, May 19, 2003 at 01:33:26PM +0100, Mike Taylor wrote:
> > 
> >     Pick a random half-decent hacker and give him the
> >     patent application.  She has 24 hours in which to read
> >     and understand it, and produce a working
> >     implementation.  If she can do it, the patent is
> >     deemed NP-Obvious and thrown out.
> 
> Him/she? Some hacker ;) (I believe 'he' is the neutral term, isn't it? 
> Richard would probably know...)

The masculine gender has been accepted as inclusive in English for a
long time, some people try to be 'politically correct' and force either
the feminine gender, a mixed 'he/she' or invented pronouns (such as
'zie').

> Unfortunately, I'm not sure there any way such a test could get in the
> document, because I suspect the test needs to be conducted by someone
> who is not going to be 'versed in the state of the art'. 

In other words, a bored patent office employee.

> The idea (as I understand it) is that the idea should be non-obvious 
> before someone is shown the patent application. But to me, that doesn't
> mean that it would be non-obvious if J Random Hacker wasn't able to 
> conceive of it. I'm not sure how you could ever judge whether something
> is 'obvious', which is probably part of the problem.

Exactly, is it 'obvious' to me, to you, to RMS, to my mother (who's a
neo-Luddite who won't even have a computer in the house), who?

If their "must have a technical effect" test is taken to mean "an effect
on a specific hardware platform" then that is not too bad, most software
is more general, but the phrase could be interpreted very loosely.

There also needs to be a challenge mechanism for "prior art", so that a
software patent can be retroactively cancelled if evidence shows that it
was generally known before, and they need to be a lot tighter than US
ones (not difficult, admittedly).

Chris C