Den 28. aug. 2012 kl. 19:04 skrev Larry Colen: > > On Aug 28, 2012, at 7:17 AM, Charles Robinson wrote: > >> On Aug 28, 2012, at 5:13 AM, Larry Colen wrote: >>> I will agree that the goatse zoom is a clever idea, but being software >>> should not be patentable. >> >> "…but being software, should not be patentable"? Did you really just say >> that? > > Yes I did. I am firmly opposed to the concept of software patents. > Mathematical formulas are not patentable. Sentence constructions are not > patentable. All patent portfolios do is block smaller companies out of the > marketplace.
It is very easy to be opposed to the concept of software patents but it is difficult to define them and to avoid them. Also, it wouldn´t necessarily help to forbid them :-) First: A large part of the problem is outside the patent system. The "triple damages" system in makes it possible to make a living from owning patents without using them. This is why you have patent trolls in the US, but they are rare in Europe where you only get compensation for lost income. Second: you need to define what software patents are. In the US, a patented invention should be new and useful. In Europe there should be a technical effect. This way we avoid business methods and other problematic patents. Third: and why do we still have some software patents in Europe? Because it is difficult to define "technical effect". The Babbage computer was purely mechanical, http://www.computerhistory.org/babbage/ but was it patentable to make it perform some movements related to logic but not others? Or another question: it is patentable make a gear shift under certain conditions, but that could also be an algorithm and it would be unfair to the inventor to let someone make a cheap computer do the same. In the about 20 years I have been discussing this with people who are opposed to software patents nobody has ever found a fair definition. If you cannot forbid what you cannot define. :-) DagT The Evil European Patent Attorney -- PDML Pentax-Discuss Mail List [email protected] http://pdml.net/mailman/listinfo/pdml_pdml.net to UNSUBSCRIBE from the PDML, please visit the link directly above and follow the directions.
