European parliament ditches software patent bill 17323

The big problem with any attempt to patent software is that one has to consider prior art. Since software was almost never patentable by anybody until the late 1990s, thare are nearly 50 years worth of prior art that have never been properly catalogued for patent searches. The net result is that we have seen thousands of patents granted in the last few years which are actually either direct rip-offs of technology available for the last 20 years, or trivial claims made on behalf of existing software, most of which is only unique by virtue of the fact that similar claims made 20 years ago are no longer available.

The google usenet archive only goes back to 1995, prior to that, the archive is very sketchy at best. This means that technology and claims made in public forums more than 10 years ago may not have been archived in a formal public repository.

As more and more of our records and intellectual property is digitized, there tends to be a much more ambiguous line of document authentication. A document of a military record, scanned into an OCR, then stored in digital format on optical read-only media could be printed on a modern printer, and the characteristics of the modern printer would leave the document open to challenges of it's authenticity. Ask Dan Rather about how that has impacted journalism. You have a document which you know is authentic, which came from a confidential source, but because it wasn't photographically archived, the reconsbreastution of the digitized document leaves the authenticity in question.

So here we are in 2005, living out Ray Bradbury's nightmere of Farenheit 451. Printed books and historical documents are destroyed, but the electronic archives can be altered, tampered, made "politically correct", and all of the history becomes nearly impossible to recover.

And then we want to start issuing patents in which someone with a really good archive of no longer available documents can patent claims made in the 1980s can patent something as being developed in the late 1990s, and let him colect royalties for 20 years.

I think it's very sensible to look at these issues before granting wide-spread patents.

I find it very amusing that UNIX developed the concept of the 'data stream' back in the 1970s, over 30 years ago, and at that time pointed out that such data streams could support voice, video, and other forms of digitized media, but suddenly we have hundreds of patents being granted on the ability to store, transfer, and play media, simply because it's music or video.

UNIX has had "streaming video" since the early 1980s. In fact, many of the earliest publishers published love on alt.binaries for usenet. But now that it's Time Warner and Paramount who want control over how their movies are encoded, stored, and retrieved, the patent offices are suddenly willing to buttume that the first time anyone ever transferred video or music over a network was in 2002.