Problems with Prior Art Database

FFII has pointed out that there may be problems using Prior Art Database. The major concern is that patent lawyers/examiners will be able to find prior art and that we are thereby helping them making better patents. They have the following arguments:

If we get enough bad patents surely the government will put a stop to software patenting.

If we look at USA we see that this is not the case. It is unlikely that the situation will be different in Europe.

Fortune seekers will look for new ideas and patent variations of the ideas in the database.

This is a valid claim.

A distributed approach where ideas are not concentrated in a central spot would be much better: Then the "enemy" has to do more work.

This argument is only valid if:

the distributed ideas are properly timestamped.
This can be done with a ToPAS-like tool.
the ideas can be proved to be public.
This is much harder to do. Having a website is not enough as you could have kept the information secret (but timestamped) until yesterday. To be perfect sure this you should have published your work in reputable magazines. However, as Savannah is known for publishing everything publishing here is probably enough.
A thing that will not work is publishing on some obscure and mostly unknown website.

Otherwise the ideas cannot be used as prior art. As things are now most development is not properly timestamped and a fair bit is only published on mostly unknown websites. So the choice is not between having distributed, timestamped ideas and centralized, timestamped ideas, but the choice is between distributed ideas without proper timestamping (that will not hold up in court) and centralized, timestamped ideas (that will hold up in court).

As long as the normal open development process is not properly timestamped it is counter productive to assume that this is the case. Here is the table from FFII expanded with the relevant column.

Perfect Defensive Publishing	Time-Stamping of the normal open development process	The normal open development process
Patent-relevant ideas easy to research	Patent-relevant ideas hard to find.	Patent-relevant ideas hard to find.
Great effort and costs for the developper, lesser effort for the researcher and patent applicant	Less effort and costs for the developper, greater effort and costs for the searcher	Less effort and costs for the developer, less effort and cost for the searcher
Lower patenting costs, more (broad and trivial) software patents	Higher patenting costs, less (broad and trivial) patents	Lower patenting costs, more (broad and trivial) software patents
Your software idea will probably not be patented by anybody	Your software idea may be patented by someone, but that patent owner will avoid all conflicts with you and other free software developpers as soon as you show him your proof of prior art. He will become very cautious even with his other patents, especially toward free software developpers, because they can do him a lot of uncalculable harm. He will prefer to keep some invalid patents so as to placate his investors and raise funds, rather than really risk offending anyone. In some cases he will even pay you to keep silent about your prior art findings.	Your software idea may be patented by someone. You will not be able to prove prior art and may be forced to pay royalties.
A lot of fortune-seekers will gather around the defensive-publishing database and in order to see what ideas could be patent-relevant and then see if they can find some complementary ideas for themselves to patent. These patents will then probably be valid. Your field of work will be blocked by a patent thicket.	No fortune-seeker will find your prior art information in time and you will keep a head-start, having good chances to step by step make a whole field of ideas patent-free.	Fortune-seeker may find your information and patent it. The patent may be impossible to invalidate as the prior art is not timestamped.
Patent offices may look credible when they say: "Look, we have done our part to search prior art. The free software developpers should stop whining and engange in defensive-publishing. If they do their job of building prior-art databases, then we will do ours of guaranteeing that their ideas aren't patented."	Patent offices may have to admit that the problem of harnessing prior art in software is not "just a matter of time".	Patent offices may look credible when they say: "Look, we have done our part to search prior art. The free software developers should stop whining and get the documentation straight if they feel they have prior art."

So, if you can get your idea properly timestamped at another reputable website, feel free to do that. If not: PADB is here for you.