That anti-patent pamphlet I mentioned

[Alex Hudson]

On Mon, 2002-12-16 at 18:03, Arnoud Galactus Engelfriet wrote:
> Alex Hudson wrote:
> > I also don't think that re-implementation of something is necessarily
> > something that should be covered by a patent. 
> 
> But surely all imitators are re-implementing an invention?

I don't accept that software can be an invention, so the question is
meaningless - it misses the point.

> Well, I also think that if you find a different implementation
> of the same inventive idea in hardware, the patent holder should
> be able to stop you. Just like with software.

No, not like software at all. Software is generally covered by
copyright, and often by contract law also. There is nothing that says I
cannot re-implement things: that's how the GNU project started.

> Ok, but when you say "implemented in software", you mean you
> have a working echo chamber purely in software? Why should
> you then not be able to patent that? 

For the reasons I spelt out in my previous e-mail : you *cannot* compare
the technical effects of software versus technology like electronics.
Patents should be given on the basis of an achievement; and my example
of a software echo chamber showed there was no achievement there. The
only thing 'new' was the effect itself, but the effect is not what is
patented, it is the means of achieving the effect. With electronics,
that may well be inventive, with software it is not.

Software cannot be lumped together with electronics, and biology, and
chemistry, and medicine, and all the other 'patentable' fields. It must
be treated as different, because it is able to achieve such vastly
different effects. 

> Well, what you're saying is that 'inventions' in software do
> not actually exist until they're built in hardware.

No, that's not my point. My point is that building something in software
is easily and not challenging. Building it in hardware isn't necessarily
so. Devices created in software are not restricted: there are no bounds
which cause a programmer to think, other than those created by the
hardware he is running the software on. Hardware is very different. A
novel piece of hardware achieves something others could not, and is new
and inventive - the inventor works with the limitations of the physical
but acheieves his goal anyway. A software programmer has no theoretical
limitations (other than what is computable); it is difficult to see how
something in that domain could ever be inventive.

> I do not believe the field of software development itself
> needs patent protection. I believe patents should be available
> for technological progress, and should protect a patentee
> against people making imitations of his inventive idea.

You cannot have one without the other: there is no way you can make
patents available for 'technological progress' without making them
available for software development. 

The AFFS (a UK Free Software organisation) went to London Expo in
October, and we had a number of posters with us. One of which was a
"Danger: Software Patents!" poster, which had two quotes which I think
expose the risk neatly. They are taken from the “Software Patents in
Europe” Conference
http://www.patent.gov.uk/about/ippd/softpat/index.htm) that was held by
the UKPTO over here a year or two ago, IIRC.

       "Up to now software-based inventions can only be protected in the
 European Community - with the exception of one Member State - disguised
 in the form of processes, systems or devices controlled by software.”
              - Arno Koerber, Director Intellectual Property, Siemens AG

 “The Commission might be prepared to support a revision of the Munich
  Convention to improve it and to make it clear that computer programmes
                       are no longer excluded from patent protection.”
                                  - John Mogg, Director General of DG XV

People are not interested in this "patent reform" because they want to
make sure people with patents are protected from "softare imitators".
They want it because they want to extend patents to software in general.
They see the patent for what it is - an extremely effective economic
device - and wish to use it against their competition.

> In my opinion, it should not matter *how* you imitate someone's
> patented invention. If it is covered by the claims, you
> infringe and you need a license. If you have a different solution,
> great, you're in the clear. 

This is where we disagree - your assumption that software and other
fields are comparable is, I think, wrong.

> To name an example, let's take MPEG again. When you make a
> video compression technique, you can build it in hardware or
> in software that runs on a general-purpose processor. The
> patent should cover both, or none. Not just the hardware variant.

I don't know of anyone building MPEG in hardware - I think this is a bad
example. Even the specialised broadcast hardware systems are essentially
computers in boxes / on chips. Just because it resides in silicon does
not make it hardware :)

I also think MPEG is a prime example of how damaging software patents
would be. I believe - I heard at RMS' last talk on the subject - that it
has taken longer for them to form licencing agreements for the (30+)
patents involved with MPEG than it took them to agree the specification.
And MPEG, although complex, isn't nearly as complex as most modern
software. The standard doesn't even give you a way of encoding video
into that format!

Cheers,

Alex.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://lists.fsfe.org/pipermail/discussion/attachments/20021217/d566d1d8/attachment.sig>