Software patents was Re: [extropy-chat] Care Economy?

[Emlyn]

On 18/05/05, Brett Paatsch <bpaatsch at bigpond.net.au> wrote:
> Rik van Riel wrote:
> 
> > On Tue, 17 May 2005, Brett Paatsch wrote:
> >
> >> If a single AI could be constructed in that time period and protected
> >> by a patent, then copied ad-infinitum, what effect do you imagine that
> >> would have economically and geo-politically?
> >
> > Then there'd be an even better reason to get rid of
> > software patents ;)
> 
> I'm curious, do you really think that it would be a good thing to get rid of
> software patents or to change the software patenting laws where you
> work?
> 
> I worked in the IT industry in Australia for something like 10 years but
> I was a business manager rather than a programmer.
> 
> I tend to see patent law as a potentially good thing but it has to strike
> the right balance between creating incentives for people to be practical
> and to innovate and not creating disincentives or other side effects that
> produce a net or suboptimal social benefit.
> 
> Its is of course always going to be possible to balance conflicting goods
> suboptimally in particular laws, and technology and other things can
> shift the optimal balance point over time.
> 
> I know that Microsoft is often kicked about and has faced anti-trust
> actions but I don't really know whether those that complain about
> software patents generally are complaining because they see that the
> balance in patent law is wrong in some area or whether it is just because
> they personally want a different set of laws that in the short term would be
> in their personal interest.
> 
> I would like to hear from someone that knows what they are talking about
> with respect to software patents and can make a reasoned case that the
> existing law in their area is suboptimal and has got the balance wrong.
> 
> It is possible that the software ip laws are suboptimal in some respects
> but it is also possible that those that are complaining about it are not
> looking at both sides of the quid pro quo that patents are supposed to
> balance.
> 
> A very good programmer that has other skills as well might want to have
> some software patent laws so that they can create a path to wealth based
> on their personal efforts and their merit.
> 
> Brett Paatsch
> 
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo/extropy-chat
> 

Sorry, I don't know enough about the laws to comment in a detailed
legal respect.

However, from the point of view of someone working in some very fast
moving technical areas, I can tell you that software patents are way
more trouble (to the industry) than they are worth.

An important first point is to distinguish patents from copyright.
Source code is subject to copyright, and distributable software is
also subject to it too I think (in that, for example, you can't make
an exact or close to exact replica of my software's user interface
without infringing my copyright I think, or of my manuals, etc etc,
although I'm not completely sure about that).

So without getting into patents, I am reasonably protected by
copyright law from someone just taking my software and onselling it as
their own (creative commons licensing notwithstanding!).

But patents let people protect entire methods of doing things, which
is a whole different idea. With patents in the picture, I can stop
people from doing something in the same way that I have done it. In
isolation, this is of course attractive to business, but in aggregate
it creates a minefield for developers to crawl through, always risking
that something they are making is violating someone's patent because
they have inadvertently used some technical or methodological approach
which violates a patent.

I get the impression that for large companies, this has lead to an
arms race, where they gather as many patents to themselves as
possible. This can be for malevolent purposes, or just as protection,
hoping that if they violate someone's patent, they will then be able
to find corresponding violations of their own patents to use as
bargaining chips. I imaging that large war chests of patents allow
large companies to come to mutually advantageous compromises in some
circumstances, licensing various technologies to each other as needed.

For small companies, however, which I believe are the driving engines
of change in software, the environment is difficult and dangerous.
Small companies can't have the war chest, and so can't play the legal
game. Moreover, small companies can't even afford the reconnaisance,
to discover whose toes they are treading on. So, in practice they have
to develop away regardless, and cross their fingers with respect to
patents. I think that many small companies do hit patent landmines and
get taken out.

The other result of the hostile legal environment is that people keep
their products as closed up as possible; every possible avenue is
taken to hide code, hide techniques, and generally make a product
appear to work as a magic genie in a bottle, because you (hope that
you) can't be sued for violation of a patent if no one can find out
that you've violated it in the first place. So the system of patents
contributes directly to the tightly closed software environment, and
the clear inefficiencies and problems related to that (just ask any
security expert).

The killer with the patent system is that it demands behaviour at an
individual level which is deleterious to the industry as a whole, and
so ultimately deleterious to the invidual. I am positive that many
large patent holders hold major portfolios not because they want them,
but because not to hold the patents would be suicide in the hostile,
low trust environment.

After all that grumbling about patents, is there a good that balances
the bad? Do they encourage innovation enough over and above that which
would happen without them? I'd say it's pretty clearly not the case,
if only by pointing out that there is a considerable history of
pre-patent software development (or at least before there were enough
patents to be a problem), and there is no evidence to show that
innovation increased since patents were taken up. For most of the
software industry, patents for the purpose of protecting your
innovation aren't worth much anyway, because the industry moves so
fast; grabbing a monopoly and sitting on it isn't going to serve you
for long because the market will move away from you in short order;
it's a moving target.

If you don't believe my ranting, here's a good article I plucked out
of the ether at random, saying that the system of patents is stifling
innovation:
http://news.zdnet.com/2100-3513_22-5342291.html

In summary, the system of patents is damaging the software industry.
Would I get a patent in an area of novel development for my own
company? The costs of the process notwithstanding, I very well might,
even though I believe they are bad, the same way that I might possibly
travel armed in an environment of personal risk of violence, even
though I believe that citizens going about armed in a civilian setting
is a bad thing for the group. However, I'd understand fully the
problems I was contributing to in that context.

I believe that the regulators must remove the patent system in its
entirety (though I don't think anyone is going to do it in a hurry).
Copyright will more than suffice for protection. Actually, copyright
needs some strong pruning too, but that's another story. But patent
law just seems to have unintended consequences which lead to
unforeseen outcomes far different to those intended. Of course it is
supported by the people to happen to win in that environment, ie: big
capital, whose winning and support is usually a sign of legislative
failure imo.

Emlyn