Tuesday, February 22, 2005

letter to the Times on 12 Aug 2004 - unpublished.

12 Amberley Court,
Overton Road

(020) 7274 3989
07710 864 775


Your business supplement yesterday on the proposed European Community Directive on intellectual property omits to mention that the directive will permit the patenting of "computer implemented inventions" or, in a word, software. Under existing law software is basically protected by copyright - you write it, you own it, and then you get to use it or even be paid for it.

Under the new proposal software will be patentable as an invention provided it has a "technical effect". This could apply to any aspect of a system, from screen layouts and interfaces to algorithms and file formats, and at any level of abstraction, from byte compression to business protocols.

Defending the ownership of "your" software against EBP (expropriation by patent) involves research, legal advice, and course much time and money. There is no presumption of innocence. You may try to prove that the words of a patent do not apply to your code, or that the claim is invalid because it is obvious, or there is relevent "prior art". None of these steps is as simple, clear or cheap as simply proving that you or your supplier wrote the code.

Who benefits from the introduction of software patents? In the US, which has had software and business process aptenting for some time, large companies like Microsoft have substantial patent portfolios and the resources to offer their customers IP indemnities. Smaller beneficiaries like PanIP are not software developers, but lodge or buy patents and then target larger or smaller businesses depending, it seems, mainly on how plausible a threat they can muster.

Who loses? Any business which relies on a clear title to software which they have developed or licensed, and on which they daily depend.

The potential costs of software patenting to British business are, I suspect, almost unquantifiably vast. This doesn't mean they should be ignored.

Francis Norton.

Free Mojtaba and Arash Day

If you're spending your valuable time reading this this blog, you either share my belief in free speech or at least know what to expect.

Democracy without free speech is a recipe for government by spin and deceit. Is Iran truly a democracy? Its censorship of media and the internet suggests that democratic values don't run as deep in the government as amongst the people.

And is the blogosphere is as overblown and ridiculous as its critics suggest, or maybe something that will make a change for good? Now is the time to find out.

Wednesday, February 16, 2005

Perseus gets a hand

Forget Sisyphus and Oedipus, my personal favourite re-interpretation of a Greek myth would be Perseus, using his shield as a mirror to kill the snake-haired gorgon Medusa without letting her eyes turn him to stone. As a software guy I think this could be a founding myth - the use of a model (incorporating real-time position and motion data, but abstracting out the metamorphic gaze) to solve a particularly "hairy" problem.

Two problems with models. One, they are always approximations, simplifications. There's not much point to a model that is just as complicated as the reality it models. Two, the modeller gets to determine the purpose and value of every component. And sometimes these components are sentient, are in fact us.

But every public policy, and every law, incorporates some kind of model of the field in question. All laws and policies are potential instruments of ruin to real people. So accurate and timely feedback on their consequences is essential for humane and effective governance.

Now, from the 1st of January 2005, we have the Freedom of Information Act here in the UK. It will be evaded and abused, of course. But that is why I think it's a step in the right direction.

Monday, February 07, 2005

Straining at gnats? Trying to define "technical contribution"

I want a definition of technical contribution which will exclude most forms of software. Trickier than it sounds. Here's my best bet so far:
technical contribution

A technical contribution is that part of an invention which cannot be defined as a purely logical or mathematical extension of an existing technical environment because it depends on properties of the physical world which did not contribute to the existing arrangement.

All comments gratefully received.

[Immediate feedback - See the FFII version, which is based on the EP version, at

the reluctant activist, part 2: how to recruit your boss

Dear <boss>,

I have been invited to a workshop on the implementation of the EU "Computer-Implemented Inventions Directive" (http://www.patent.gov.uk/about/ippd/issues/cii-workshops.htm).

As I'm sure you know, copyright protects a software author against unauthorised copying of their work, whereas patenting protects against independent re-invention.

My concern is that as a company we can be reasonably certain that we are not infringing anyone's copyright, but there is no obvious way of checking that we are not infringing anyone's patents, as several companies and standards bodies have been discovering over the last few years. And since (like most software SMEs) we hold no patents, we have nothing to gain and plenty to lose by this step in the legalisation of software patents.

There is a loophole. In principle "pure software" is not and will not be patentable, but "Computer Implemented Inventions" - anything involving software but which also involves "technical effect" - are.

I intend to argue that this "technical effect" should be defined in such a way that the kind of systems we write will be classed as pure software rather than as computer implemented inventions, and therefore not at risk of expropriation by some patent-swapping cartel or litigation enterprise.

I am happy to do this workshop on my own time, out of my vacation allowance, but the application form invites me to state my affiliation. May I state that speak for <us> - our shareholders, customers and employees - on the lines that I have outlined?

Thanks -


Thursday, February 03, 2005

Telegraph | News | 'If you don't take a job as a prostitute, we can stop your benefits'

A 25-year-old waitress who turned down a job providing "sexual services'' at a brothel in Berlin faces possible cuts to her unemployment benefit under laws introduced this year.

I recently read a comment to the effect that to tolerate something is to make it compulsary, which I thought at the time was obvious nonsense. Now I'm starting to understand how the problem can arise.

A friend raised the interesting question - is religious sanction all that comes between us and compulsary prostitution, as an inevitable consequence of tolerated prostitution?

My boring answer would be, in a word, no. The penalty cuts in benefits are presumably intended to prevent the benefits going to people who did are not genuinely seeking employment. The fact that a person has turned down work as a prostitute is not compelling evidence that she is work-shy, any more than a vegan declining work in a slaughter-house or, similarly, a pacifist refusing military enlistment.

So what happens when there is a gap between the acceptable moral argument and an unacceptable adminstrative reality? If you believe, with Protagoras, that man is the measure of all things then we - and she - should stand side-by-side with the vegan and the pacifist in a common fight (obviously non-violent, involving no harm to animals and with no trading of sex for favours) for the law to be amended to permit all to exercise their various credible moral convictions.

But what happens if I have a credible conviction that I should not pay tax? Or that there should be compulsary eugenics (as opposed to merely supporting unrestricted advertising for fast food and tobacco companies) - perhaps some of these convictions are more tolerable than others?

Ah, tolerance again. Like the dust under the carpet, you can move it round but the bump remains. The price of freedom may be constant vigilence, but price of tolerance seems to be constant moral choice and debate.