Re: [Fsfe-uk] Arlene McCarty on software patents

[Robin Green]

On Thu, Jun 12, 2003 at 09:25:34PM +0100, Nick Hill wrote:
> This gives some insight to the workings of Arlene McCarthy. I will 
> compose a response. Any input about elements to include, and references, 
> are welcome.

1. Please, please ask her what the fuck "software as such is not patentable... 
[but
software inventions are]" is supposed to mean. We're all waiting for a clear
explanation of "technical effect" here.

2. McCarthy writes:
"Even in the US, where the patenting system is more liberal, the Linux 
operating system proudly declares a 50% growth in world wide shipment of its 
operating systems, with the Apache system dominating the internet web server 
market with 66% market share. With a more restrictive EU law, the open source 
community has nothing to fear."

Given that the Open Source community has nothing to fear, will McCarthy then
gladly accept either of Red Hat's proposed amendments:

"If software patents are going to be accepted by Europe, then our position as a 
true Open Source company is that they should not be enforcible when used in GPL 
or other open source software. If that sounds too radical, then our position 
would be that Europe should not have software patents at all."
Ref: http://www.redhat.com/advice/speaks_openeurope.html

She cannot both claim that open source will not be affected, and that it must
be prevented from "stealing other's ideas".

3. Of course her argument is BS. Either Apache is the subject of ridiculously
overbroad patents which haven't been enforced, or it doesn't contain anything
patented. Either way, bad example, because Apache can't plausibly DEPEND on 
anything 
patentable for either functionality or performance, in my opinion. It's just
a web server.

However, Linux or GCC are different kettles of fish, being more complex than
Apache. It's usually rather pointless for corps to sue contributors to 10+ year 
old
open source code (unless the bit they are contributing to is really new),
because they won't gain significant restitution, royalties or competitive 
leverage -
the only point would be spreading FUD (hence the SCO case). No restitution 
because
the corp would be suing someone probably much much less wealthy than them.
No royalties because the offending code would be ripped out and replaced.
(They could try suing customers I suppose, but that would be bad for their 
corporate
reputation - again, see SCO.)
No competititive leverage because, uh, the offending code would be ripped out 
and 
replaced. Maybe with a suboptimal algorithm, but let's face it, people don't use
GCC because of its performance.

Now, by stark contrast, the chilling effect of patents preventing code from 
being
written IN THE FIRST PLACE should not be overlooked! That's where the real
anti-competitive action is. Not in 10+ year, well-established old pieces of 
code!

That's because you're not going to be able to take the likes of GCC or Linux out
with any patent suit. They would patch and heal themselves - because a hell of a
lot of people and corps now depend on them.

However, it's much more plausible for a new project or module to be tactically 
nuked
by a well-positioned patent lawsuit. The mere threat of that - especially if one
patent lawsuit is successful, and the company threatens more - will be enough to
make people who value their bank balances turn their attentions to something 
safer.

4. Please bear in mind when writing your response that she may have been simply
misled by the pro-patent side, rather than being intentionally deceptive.
(I know, I know, I give politicians too much credit sometimes... but one
can't be defeatist about these things.) Stay reasonably polite, obviously.

-- 
Robin