Capaigning against software patents

Tue Dec 3 23:00:23 UTC 2002

El Tue, Dec 03, 2002 at 08:04:41PM +0100, Niall Douglas deia:
> 
> The solution here is to produce some very conservative numbers ie; 
> the lowest cost it could be. Obviously, if a politician should ask 
> what's the basis for these numbers you should pull out a sheet of A4 
> illustrating the basis. Most importantly of all, the numbers have to 
> be so conservative the proprietary lobby groups can't dispute them.
>

It's all too vague. I wouldn't try. Any number can be disputed.
I think it's better to say SMEs are at risk, and then say over 70 %
of workers in IT work in companies with less than 200 people (this is 
only counting coimpanies with more than 10 employees, and only for 
Catalonia in 2001, I'd like the figure for the EU, should be similar). 

http://www.gencat.es/csi/pdf/cat/estadistiques/Oferta_TIC_2001/2_Situacio_actual.pdf 

This is not saying you are going to lose 70% of employment in IT,
which is very hard to prove.

> 
> Accuracy is relative and subjective. Most people accept presented 
> evidence until proven otherwise.
>

I don't think we'll get that luxury. As you said the other side 
has deep pockets to make sure they prove otherwise. Also, 
confusing concepts is their strategy, we should be accurate
in contrast.

> > (v) try that it looks true too, not only that it is.
> 
> I'm not saying we lie, but I would argue for simplicity. I'd want 
> someone to look at our case and in ten seconds have a good idea. I 
> don't think reading through any quantity of text will help this.
> 

Maybe I've been looking at it too much. But I don't think you 
can get that. It's not so simple. Was it Einstein who said
"make it as simple as possible, but no more".

> In the end, these guys are busy. If you leave them with just an 
> impression it could be a vote-loser, you're 50% there.
> 

Yes, of course. I thought that impression was a function of how many
people complain...

> 
> Writing letters costs little and can make a huge difference eg; 
> getting star trek put back on air. Writing letters to Euro software 
> CEO's also I think is an excellent use of time.
>

We wrote a letter for Software companies but we finally didn't send it 
because it was too close to spam. Maybe paranoid, yes. It's not a big
deal, and it's in Catalan only, but it's at 
http://patents.caliu.info/emrpeses.html

> > I'm a little sick of people thinking they can understand everything in
> > just an hour and then making so much noise that politicians can't hear
> > the signal. I agree there must be introductory texts, etc. but just as
> > a first step. You can't hope to go talk to a politician just after a
> > couple of hours reading something. How are you going to ask her to
> > spend time on what you care about if you haven't spent it yourself?.
> 
> I wasn't proposing you send such people - the primers are for those 
> who know something vaguely about it (eg; anyone reading slashdot or 
> the register) who wants to know enough they feel safe in writing a 
> letter.
>

Ok, sorry. Maybe I was too sleepy. But for writing a letter I think
it's not to much to ask to read some 10 - 15 pages. Otherwise we'll
get lots of letters with unclear concepts which will be dismissed as
uninformed, and we will appear just complaining without knowing
anything about it. Which is by the way a classic argument of the
proswpat people "SMEs are worried about swpats because they're
uninformed, they need education. Computer professionals weren't taught
software patents at Unviersity, that's why they don't like them". This
is absurd, as Sweetman explained very convingly in Brussels, but it's
what they keep saying. And sending letters which confuse patents with
trademarks, or complain of patents invalid in Europe, or propose a
shorter term for patents without much convincing arguments for
changing or abandoning the TRIPS agreement is not as useful as some
slightly better informed contributions. Maybe uninformed ones are
better than nothing, but I don't really think I'm asking for so much.

I hope I'm not turning too many people down. It is not so utterly
complex and hard. It simply requires a few hours. That's all. And 
it's well worth it, because the impact then is much stronger.

But now it is me who is telling people what to do, and that's wrong. 
So thank you for any efforts. This is just my way of doing it, nothing
else.

> 
> If I came across as telling people what to do, then that was not my 
> intention and I apologise for any offense caused. I've been 
> campaigning against software patents long before joining this list 
> with articles written on the matter and letters sent to the UK patent 
> office. I was merely applying what had worked well for other popular 
> protest movements, and hoping it to be a source of good ideas.
> 

Thanks for all that campaigning. Could you send me or post URLs 
for those aricles or letters?. Maybe I'll understand better with 
some examples, and maybe I can use your ideas.

> > 
> > For some. Others possibly read it. And anyway, it's the kind of place
> > one should study before starting to teach politicians. You can't just
> > go to one politician and throw him a panphlet you hardly understand
> > yourself. You have to be ready to back your arguments and provide
> > constructive alternatives.
> 
> There aren't many constructive arguments for alternatives to software 
> patents except "just don't have them". It's a pity, because that can 
> make us look closed minded.
> 

But of course there are. Maybe they're not new. But there are a few 
arguments to use. You can't just say "let's don't have software patents". 
That's what's already in the law and we have software patents. The problem
is that the EPO is not going to use common sense, so you need sound 
legal definitions of what one is allowed to patent and what not. Otherwise
patent lawyers will find the holes and sneak through, that's their job.

Some traditional criteria that IMHO should be reinforced is that 
of technicity (the Taube test, I think it's called). 
The idea is that you can only patent an invention if it is an invention, 
i.e. it gives us new knowledge on the use of controllable forces of nature.
This means that if your innovation is only in the sphere of mind, a
logical deduction, an intellectual achievement, like math, software, 
literature, bussiness models, rules of organization, social games, 
whatever, you may have a lot of merit, but the economics make it 
bad to grant a patent for that because you only need your mind, time 
and preexisting ideas for it, so that kind of innvoation will happen 
if we allow free exchange and reuse of ideas. If you needed 
laboratories and experimentation to verify your innovation works, 
because it is not merely a consequence of a formal system of some 
kind but a piece of matter and energy with physical properties 
and any formal model is only a model, not the invention, this is 
what requires expensive experimentation that might not happen 
without a patent (at least argueably). 

More on this 
http://swpat.ffii.org/analysis/invention/index.en.html
http://swpat.ffii.org/vreji/papri/bgh-dispo76/indexen.html

Another traditional criteria worth inforcing is industrial applicability.
The current EPO interpretation is "you can sell it". But the correct one
would be that the invention should be applicable to the industry, 
that is, to the serial production of material goods. This insures 
the costs are not marginally zero and the cost of the patent system 
is maybe not so much when added to the production costs. This also 
ensures the effects will be on manufacturers, not directly the public. 

Another one is disclosure. If a patent does not show how to do something
it shouldn't be allowed to claim (monopolize) that something. For inventions
with software in it, this means a listing of complete source code for a
 working program should be included with the patent description. Otherwise
you get people which only patent but don't even take the trouble of 
implementing the invention. 

And we come to an odd point here. You may wonder why I speak of programs
in patents if I don't want software patents. Computer program patents 
are forbidden by Art 52.2 of the European PAtent Convention. But Article 
52.3 says only programs for computers "as such" are excluded from patentability.
This makes sense, but it has been so abused that we could well erase Art 52.3.
Why makes sense, though?. Because the correct interpretation is that you 
cannot patent software as such, but if you have a patentable invention
(say a chemical plant) when you add software to it (maybe it's the only 
practical way to control valves so that they react to pressure sensors) 
it does not become patentable. The criteria is what is the new knowledge.
Is there some teaching on forces of nature or not?. Did we know that 
changes in input in response to pressure would produce that process or 
we didn't?. If we didn't the fact that some software is necessary does 
not render the chemical plant unpatentable. But the software is not 
patented as such, just a chemical plant using software. This means that
anybody can use the same software (respecting copyright) or rewrite an
equivalent program, as long as she does not set up the chemical plant 
(for instance anybody can use it for simulations, etc.). 

Likewise if you have something that is not patentable (say a general 
computer, old invention) adding software does not make it patentable
(that would stop all those patents claiming a computer or processor 
or network equipped with a program or configured to do X). 

This is only common sense, but we need a clear codification in law of
this (the only feasible) interpretation to stop the EPO from
misbehaving (not really likely without reforming it) and to ensure
that EU courts disregard wrongly issued EPO patents (more likely). Of
course the Comission and Council proposals are trying to codify the
EPO interpretation, in which a program for computer as such is some
esoteric concept that does not include any software at all, so that
all software is patentable.

The reason of all this rant is that if you go to a politician and 
just ask him to vote on software patents to "just don't have them",
he may well tell you he agrees but vote the Commission proposal, and 
then tell you it does not allow patents for software, just for 
"software with a further technical effect". You really have to ask for 
strict criteria, and that requires constructive alternatives to the 
directive proposal. 

Luckily we have a very good one, the FFII counter-proposal. 
HTML http://swpat.ffii.org/papers/eubsa-swpat0202/prop/ 
PDF, better typesetting, included in a bigger dossier quite 
convenient to give to politicians (although it's 80 pages or so)
http://swpat.ffii.org/events/2002/europarl11/cfa-en.pdf

If you are confused by all these details, then suffice to say we 
need clear limits to what can be patented. Limits are not clear 
now because people is twisting interpretations to their convenience, 
so we should use a test suite to benchmark any legislation. Whatever
the technical details, before approving a directive we should make sure 
there is no way one of these example patents would be enforced under
the directive. This is one of the things that we include in 
our call for action 

http://swpat.ffii.org/papers/eubsa-swpat0202/demands/

So we need businesses, university departments, NGOs and assocations, 
labor unions, etc, etc, etc to sign it (the call for action is 
included in the dossier above)

But we must defend all this, not just push a headline "software
patents are bad".  But most of you already knew all this, so I'm sorry
for preaching to the choir once more. And you are probably right we 
cannot push all of this in the first meeting with a MEP. 

-- 
Xavi Drudis Ferran
xdrudis at tinet.org