[R] Patents on algorithms harm data analytic services

[Jens Oehlschlägel]

Cologne, 10.12.00

Dear Sir,
Dear Madam,

We are concerned about the possibility that the European Commission might introduce software patenting into the European Community because we think this will harm our profession.

We make our living on data management and statistical analyses. Modern statistics crucially depends on algorithms [cf. Venables, W. N., & Ripley, B. D. (1999). Modern applied statistics with S-Plus. New York, Berlin, Heidelberg: Springer Verlag]. Furthermore modern data analysis goes far beyond applying algorithms contained in ready-to-use software packages, modern data analysis means customizing or developing algorithms suitable to specific data situations [cf. Chambers, J. M. (1999), Programming with data, a guide to the S language. New York, Berlin, Heidelberg: Springer Verlag; Venables, W. N., & Ripley, B. D. (2000). S Programming. New York, Berlin, Heidelberg: Springer Verlag]. Our work involves generating ideas about appropriate algorithms, turning them quickly into software ('rapid prototyping'), gradually improving successful algorithms in terms of performance, usability and documentation and sharing those algorithms with colleagues. This is a modern service oriented business: it's about selling intellectual service, not about selling 'intellectual property'.

However, software patenting threats this business, as nobody doing that type of work has the time or resources to check whether an algorithm might violate any patents. Checking for equivalence between algorithms is a non-trivial task, and often it is less time consuming to program an algorithm than to search and buy it. And once we have created it on our own, we might want to improve it and share it or even sell it. We do not see any contradiction between patent free algorithms and selling software. We are convinced that for good software there will always be a market, as can be seen with the statistics programming language S. Although there is a free open source implementation (called R) a private company can sell a commercial implementation (S+ by MATHSOFT) which offers certain advantages. It is rather the other way round: the existence of free algorithms and free software pushes software developers to improve their products, such free market competition pushes the new economy. The real big business and the real big challenge of modern software is achieving integration, not having unique or secret algorithms. It's an old economy idea to put a patent on an algorithm and to lean back waiting for money to come.

Software patenting does not only directly harm our professional business, but also indirectly threatens freedom of research, as research crucially depends on 'data analysis' and 'sharing ideas', which is incompatible with patenting algorithms. Furthermore patent restrictions on algorithms (and thus on data analysis) can harm society: It has been argued (by Edward Tufte, an expert on data visualization) that the challenger explosion could have been avoided given appropriate data analysis. Now imagine someone comes to you and says 'tomorrow morning we have this challenger launch and we suspect that ... we urgently need these data analyzed' and you have to answer 'well, let's go to the patent office'.

Best Regards

Dr. Jens Oehlschlägel & Peter Malewski
Research Consultants



petition at eurolinux.org wrote on November 14th 00:

> Let the EC know what you think about software patents
>
> Dear Sir,
> Dear Madam,
> 
> The European Commission is currently researching the economic impact of 
> software patents. For quite obvious reasons, many patent attorneys and 
> IP lawyers who earn money through the patent system are currently lobbying 
> the European Commission in favor of a broad extension of the patent system 
> to software, business methods, intellectual methods, etc.
> 
> Unless you express your own opinion, only their opinion will be taken into 
> account in the decision process, whatever the consequences on your business, 
> whatever the consequences on innovation.
> 
> It is therefore very important and urgent, if you consider software patents 
> to be more harmful than useful, to send your opinion by email to:
> 
> 	consultation at eurolinux.org
> 
> as soon as possible and, in any case, before December 15th, 2000. You
> can write in the official language of any member country of the European
> Union.
> 
> Your email will then be forwarded to the European Commission and published on 
> the EuroLinux Web in order to make sure that your point of view is taken into 
> consideration:
> 
> 	http://petition.eurolinux.org/consultation
> 
> There is currently a consensus among economists on the fact that software 
> patents tend to stifle innovation and harm small and medium enterprises 
> because they create tremendous juridical uncertainty which only benefits to 
> patent attorneys and lawyers. There is also a consensus among patent 
> attorneys on the fact that patents on business methods are just a kind of 
> software patents and that it is impossible to ban business method patents 
> once software patents become legal.
> 
> Please write serious (but not necessarily long) emails, with a consistent
> analysis based on economics, technology or real world examples from your
> everyday practice. Here are a few advice for your email to reach maximal impact
> within the European Commission:
> 
> 	1- NO POLITICS - Do not include in your emails any political analysis.
> Otherwise, certain civil servants at the European Commission will pretend that
> you are politically biased and claim that your arguments are irrelevant.
> 
> 	2- FREE MARKET RHETORICS - Use rhetorics based on free market,
> competition, innovation, entrepreneurship, SMEs and property, just as if you
> were the chief of the federation of enterprises in your country. EuroLinux has
> experienced that "free market economy" is currently the only common language
> which most civil servants at the European Commission understand. In order to
> let them understand your point of view and take it into account, it is
> compulsory to speak their language. Arguments based on epistemology, ethics or
> history are acceptable but have in general no positive impact on the European
> Commission because only few people will understand them.
> 
> 	3- DAVOS COMPATIBLE - Imagine that you are introducing your point of
> view at the Davos Economic Forum in front of CEOs who will only listen to you
> if your arguments mean more profits to them. Incidentally, many Commissioners
> at the European Commission used to be members of the steering committee of the
> Davos Economic Forum.
> 
> 	4- CONSENSUS AMONG ECONOMISTS - Always mention that there is a
> consensus among economists on the fact that software patents harm innovation.
> 
> Please understand that our advice does not represent any political point of
> view of the EuroLinux Alliance and is strictly designed at helping you to
> present your arguments in such way that they are going to be taken into
> account by the European Commission.
> 
> For more information on software patents, please read our knowledge 
> database and follow the links:
> 
> 	http://petition.eurolinux.org/consultation
> 	http://petition.eurolinux.org/pr/pr5.html
> 
> If you need inspiration to write your own statement, you may also access our 
> statements database where 100 European companies have already published 
> position statements:
> 
> 	http://petition.eurolinux.org/statements
> 
> Best regards,
> 
> EuroLinux Alliance
> petition at eurolinux.org
> http://petition.eurolinux.org
> 
>
______________________________________________________________________________
Die Fachpresse ist sich einig: WEB.DE 15mal Testsieger! Kostenlos E-Mail, 
Fax, SMS, Verschlüsselung, POP3, WAP....testen Sie uns! http://freemail.web.de

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._