Hi
I thought I'd forward a copy of the email I just sent the Eastern MEPs to the list - I was in both a hurry and a blind rage when I wrote it so it's probably not me at my most coherent. Please let me know if there's anything obviously wrong with it.
While I'm here, are people aware of this:
http://swpat.ffii.org/papers/rms-cam020325/index.en.html
Whatever you think of rms (personally I think he's a god and will defend his honour with my life) this is a great article about patents.
Joe x
---------- Forwarded Message ----------
Subject: Re: Software patents Date: Wednesday 02 Jun 2004 14:18 From: Joe Button joe@joebutton.co.uk To: rhowitt@europarl.eu.int
Hello again.
I'd just like to follow up my email with the information that Network Associates (a large american company) has just been granted a patent on various spam (junk email) filtering techniques in the US. By way of exploring the wisdom of software patents I would like to point out some of the implications of this:
* One of the techniques the patent covers is 'Bayesian filtering'. The patent was applied for in December 2002. A very cursory search enabled me to find public discussion of Bayesian spam filtering from February 2002, which referenced material on the topic from dating 1999.
* Another technique covered by the patent is apparently 'MD5 paragraph hashing', which simply means the creation of a shorter 'digital ID' code for a paragraph of text. I'm not a spam filtering expert, but as a reasonably capable programmer it seems to me that that is a blindingly obvious thing to do if you're trying to identify spam.
* If I write a program that uses either the first established and public technique or the second extremely obvious one, I can be sued. In fact if I'd written one before the patent was granted or applied for I could still be sued. My options would be to get out of the spam-filtering business, to pay Network Associates for the right to use 'their' technology, or to launch a separate case to try to get the patent invalidated, which would be too costly and difficult for virtually any small business.
* As an email user, I will be denied the use of free spam filtering software such as the popular 'spamassassin' as the distributors will not be able to license the technology.
* Due to the decrease in competition that will be caused by this patent, spam filtering will be generally less efficient than it could be. The world will be a better place for spammers and Network Associates, and a worse place for the rest of us. According to the European Commission spam cost European businesses around two-and-a-quarter-billion euros in 2003
* It's also worth noting that many other spam filtering techniques are patented, some of them under patents that are so broad that they are simply widely ignored on the assumption they won't stand up to legal challenges. Contrary the UK rep's statement of support for de-revised text, permitting software patents will _not_ lead to a simplified legal situation, they will simply create a minefield for software developers and businesses.
I hope this helps to clarify some of the implications of the impending legislative changes.
Regards,
Joe Button