THE world鈥檚 inventors are a prolific lot, publishing several thousand new patents every week. OK, so most of these are as dull as ditchwater, but a small number usually turn out to be gems: some are fascinating (Hitachi has a new patent to make electrical equipment release a warning smell just before it goes wrong 鈥� see 鈥淪niffing Circuits鈥�); some are genuinely world-changing (the patent for Pfizer鈥檚 Viagra); and some are just plain daft (one recent patent from Florida described methods of reincarnation).
What they have in common is that they represent a bargain between inventor and society. Society gives the inventor a limited-life monopoly, usually 20 years. In return the inventor releases full technical details for others to look at before they plan their own work. The information stays published forever afterwards and can be searched, read and copied freely without constraints 鈥� a public good open to all.
That鈥檚 the theory, anyway. The reality is rather different. For years I have scanned new patents for anything likely to interest New 杏吧原创 readers. Initially that meant travelling to a patent library and sifting through paper specifications. Then, about six years ago, the three main public patent databases 鈥� run by the European Patent Office (EPO), the World Intellectual Property Organization (WIPO) and the US Patent and Trademark Office (PTO) 鈥� went online, apparently promising effortless surfing and no more library trips. It seemed the answer to a patent searcher鈥檚 prayers.
Advertisement
But when people actually started logging on it was a different story. Taken together, the three systems are a dog鈥檚 breakfast 鈥� confusingly different and absurdly awkward to use. The British Library in London, where I used to go to read paper patents, does all it can to help, by providing an electronic gateway to the key databases and even running free courses on how to use them. But privately the tutors tear their hair out at the obstacles in the way of the would-be searcher.
When the EPO鈥檚 database is searched, patent numbers must be entered as a continuous string without spaces. So patent WO 2003/051033 must be entered as 鈥渨o03051033鈥�. The leading zero before the 5 is essential; entering 鈥渨o0351033鈥� gets no result and no helpful advice on why. But enter 鈥渨o03051033鈥� into one of the WIPO search sites and you will draw a blank: it needs a completely different entry format, 鈥渨o/03/051033鈥�. To trap the unwary, it defaults to searching only the current week, so a search can fail even when the number is correctly entered.
There are other problems. Each page of an EPO patent must be downloaded and/or printed separately, which makes speed-reading impossible 鈥� unless you pay for specialised commercial software. The PTO works differently again. Its patents are stored as a graphics file, making it necessary to install and use a browser plug-in. 鈥淥ur system was in place long before either EPO鈥檚 or WIPO鈥檚, and funding has not permitted improvement,鈥� says Larry Larson of the PTO. The US system also stops people downloading whole patents, unless they buy special software. 鈥淭hat is intentional,鈥� says Larson, 鈥渟ince some US patents consist of over 5000 pages.鈥�
The EPO gives a different reason for making searchers鈥� lives difficult. 鈥淭his was done on the basis of a voluntary restriction at the request of the commercial operators,鈥� it says. Jeremy Philpott of the UK Patent Office adds: 鈥淭here is no technical barrier鈥t is to allow users to choose between paid-for, added-value services of higher functionality, such as Derwent, or free services of more limited functionality.鈥�
For several months now I have been pestering Alison Brimelow, the ex-head of the UK Patent Office and now president designate of the EPO. If the EPO made users apply for free registration, I suggested to her, and then let them download a restricted number of full patents every day, plus any number of separate pages, this would protect the likes of Derwent, and benefit the rest of us. Derwent acknowledges that it is not selling primarily to people who search out only a few patents a day.
Patent officials say it is 鈥渘ot appropriate鈥� for Brimelow to say anything, but the EPO admits that users want a more user-friendly system. It also accepts that 鈥渢he download of complete documents is necessary to obtain a proper use of patent information鈥�, and has even developed a strategy that would allow this. All now depends on a meeting of the EPO鈥檚 Working Party for Technical Information, due any day. If it can agree on the new strategy, the switch from paper to electronic delivery might finally start to make patent information easier to gather, not more difficult.
In the meantime, patent searchers face only frustration. Last week, I went back to the British Library in London to try speed-reading on paper again. It was a wasted journey. Over the past few months most of the paper publication has stopped. It is now electronic searching or nothing.