Features

net.wars: Patent pending

Technology is a little like Las Vegas: because it's constantly being reinvented and rebuilt, people are always claiming that they're first at something or other that's actually been done many times before. Still, hope springs eternal, so when you're asked to meet a company that's launching a system that's about to receive a Europe-wide patent, of course you hope it's something really new and clever.

The company was AllIsblue (don't go there unless you really, really love Flash and think it's the bee's knees even for displaying news headlines), and the patent-pending invention is a system ("SMS2mail") whereby you can text a word to a free short code and get back more information by email to the address of your choice. You're overwhelmed, right? Because that's the thing that's really lacking in your life: sufficient advertising, at least until Bluecasting ads from nearby billboards takes off.

The thing that's hard to understand is why on earth this system should be worth a patent. As another journalist pointed out, it's really just an updated form of printing a reply coupon in an ad, and not at all dissimilar to the signs you used to see in the Paris Metro advertising Minitel pages (although in that case of course you had to go home and get in front of the terminal before you could use the numbers). Advertisers may like it because using a mobile phone instead of a pen/paper/envelope/stamp means people will respond on impulse. But what is non-obvious or technically innovative about a database that collects email addresses via SMS and sends them files (magazines, audio, video)? No one knows, except apparently the EPO and the proud parents.

"In my opinion – and I have only made a quick review of the patent – this is a childishly obvious solution to the problem of sending files to mobile telephones," says Gary Barnett, a senior analyst at the research firm Ovum.

Downloading ringtones in response to an SMS, which you've certainly been able to do for some time, doesn't count, they say, because you have to have a compatible handset in order to do it. Their system works with any mobile phone, no matter how ancient. Even so, it just doesn't sound special.

Those campaigning against the Computer Implemented Inventions Directive have long said that the EPO is the most liberal patent office in Europe. That is in fact one of their objections when CIID proponents (such as the UK Patent Office) say that the law will harmonise existing practice and note that software is already being patented. Well, it is, and if the patents were being awarded to software that truly changed science (a dramatically new way of doing encryption, for example), there'd be less objection. But as far as I can make out, the Allisblue patent is purely software, and a perfect example of what's wrong with the system – and a perfect argument against harmonising everyone to what seems to be existing practice. The more so when you consider that the EPO's existing practice is at odds with existing law.

"The EPO systematically disobeys applicable law," says Florian Mueller, software developer and founder of the NoSoftwarePatents campaign, "which clearly excludes computer programs from the scope of patentable subject matter. That's why courts in key EU member countries such as the UK and Germany frequently declare those EPO software patents invalid within their respective jurisdiction." He points to a recent example, in which a UK High Court rejected software patents in the case of Halliburton vs. Smith.

More generally, he says, the structure of the patent system makes it more likely that a patent will be granted than rejected: "A patent examiner at the EPO only gets something like ten hours to check on a patent application. If he doesn't identify any prior art during that very small amount of working time, then he has to issue the patent unless there are any formal errors based on which he can reject. If he rejects the application, he has to prove his case, as the applicant will be very likely to appeal the decision." If the patent is granted, the applicant is happy, and while others may not be the burden of proof that the patent is wrong will fall upon them, not on the original examiner.

Now, you might ask – as we did – why it matters to Allisblue whether its system is patented. After all, the real barriers to entry for competition are going to be the business partnerships it builds. The company's CEO, Eric Delfosse, described it as "the oxygen" of the business. Not, he said, because it will give them a business monopoly, but because it will protect consumer privacy. Allisblue is not evil, he argued, and as the only operator it will make sure (under the guidance of FEDMA) that email addresses are guarded carefully and not spammed. If there are multiple operators, some of them might not be so honorable, and the entry of spammers would destroy consumer trust and kill the whole enterprise.

It's an interesting argument: but nowhere in law does it say that patents should be granted on that basis. Can we keep our eye on the scientific merits, please, folks?