If you have questions that aren't answered here, please either login and add a question, or join the discussion.

Q. What qualifies as Prior Art?

A. See Prior Art .

Q. I think software patents are evil. Shouldn't we focus on eliminating them instead of trying to help the USPTO improve their approval process?

A. No, that's beyond the scope of this project, which is about how to handle problems presented with the status quo given that software patents aren't going away anytime soon. The USPTO must implement whatever the US legislature dictates -- they have been assigned the daunting task of reviewing software patent applications and issuing patents when legal requirements have been met. Faced with problems finding prior art, they contacted IBM and reached out to OSDL and others in order to get help in doing their job better. Whether or not you're a fan of software patents, everyone can agree that the state of issued software patents is not good. By developing ways to help the USPTO find prior art, and making Open Source Software more available as prior art, fewer software patents will issue. The project's basic goal is improving accessibility to the very creative, innovative code developed by the open source community as a source of prior art. Everyone wins, proprietary and open source developers and businesses alike, because fewer bad patents are issued.

Q. I wrote programs under DARPA contracts in school may years ago. Are the algorithms and processes in those prior art that can/should be registered?

A. Anything that has been publicly published can be considered prior art, and once the "tagging" mechanisms have been developed, everyone will be urged to use them. Hopefully places like Freshmeat.net will add it to their categorization system. We will be able to elaborate on this answer once we get further details from the USPTO on exactly what can be legally considered "Prior Art". During the Dec '05 meeting, they gave a presentation on this, and we're waiting to get a copy to publish on the main site.

Q. Will something be considered prior art as of the date it was originally created or will only be considered to be prior art as of the date that it is properly tagged and indexed?

A. A document is considered prior art as of the date it was published. For it to be published, a document must have been accessible to those interested in it. With electronic documents available on the internet, tagging and indexing generally are not important in determining whether a document qualifies in the strict legal sense as prior art. This is because the law considers the internet to be uniformly searchable and accessible via search engines and other means (regardless of whether or not this is technically correct). The point being that this project is not about making sure the products of open source and free software development qualify as prior art under the law - they already do. Rather, this project is about making sure that those products - which already qualify as prior art - are readily available in useable form to patent examiners with limited time and resources available to them.

Q. How do you dig into source code to find prior art? Searching with key words won't help, and I do not see how you will extract concept and ideas out of source code.

A. Tagging is a limited way to extract ideas and concepts from source code. Beyond tagging, any form of extraction is probably beyond the scope of this project, at least for now.

Q. Is there an existing taxonomy of algorithms and software problems/solutions that can get us started? Is there a wiki?

A. There are a number of them. See Classifications schemes .

Q. Do language constructs constitue "prima facie" prior art? That is, if the construct is in a language syntax, the idea has been disclosed and hence prior art, whether it was actually used or not.

clarification needed: Do you mean documentation and discussion? i.e. would discussing a way of doing something in a mailing list or a design document count as "prior art" even if the code for it doesn't exist?

Q. It strikes me that we, the developers, need to put on our Patenteers hats, and start to make 'claims' about our software, imitating the claims made by real patent applicants (only less obfuscated). If we manage to emulate the language that the PTO is accustomed to, we'd perhaps fit better into their patent review process, and become a more comprehensible resource. Is it practical to devise a 20-questions script which might help developers to adapt to the PTO mindset, rather than expecting them to delve thru code ?

A. Perhaps. The PTO indexing, however PTO-centric, is at least a taxonomy, which developers presumably find more comfortable than the seemingly purposeful obfuscation of patent descriptions and claims. The trap in attempting to speak their language is that they take our claims literally, and we risk being too narrow in our claims.

Q. I have a proposal: let's use this wiki to start a taxonomy of software problems with links to example open open source projects that have solutions to the problems. We can also list out obvious solutions to problems. We can start with one of the existing taxonomies but since it's a wiki we can evolve it as necessary. (1) How many folks would help with this? (2) Is OSDL happy to host this? (3) Any other reactions?

A(1). one... so far, update this number if you want to help

Q In the Wikipedia article, Software Patent Debate, the point is made that the disclosures of software patents are not particularly useful. How much guidance should an inventor provide on a new software technique so that it would be worthy of a patent?

A The legal answer is that enough guidance should be provided so that another person could reproduce the invention without "undue experimentation".

Q For a document to qualify as prior art, the date of publishing must be provable. This can be difficult for internet publication. As a substitute, can the tagging system provide legally relevant time stamping (certified time) for the date of tagging ?

Q How does all this apply to Europe?