Defensive Publications

Software patents are an evil thing which should die a horrible and painful death. Until that moment, recording prior art in a way that is understood by the system is an effective way to fight patents. By recording prior art in the form of defensive publications, we can make it much harder for a patent to be granted — and it does not have to be hard at all to do so.

Yesterday at FISL, I attended a panel on software patents in Brazil, the discussion revolved around the why and why nots. Unsurprisingly, there are a lot of good reasons why not, and very little pro. It’s an interesting topic, especially since I lately dived a bit deeper into this topic. Softwaree parents were also high on the agenda of this year’s Akademy, which happened last month in Tallinn, Estland. They are legal threats to Free software, so we need to figure out how to deal with them. Let’s first rehash why software patents are a bad thing:

  • Patents are valid for a ridiculously long time. This unrealistic timeframe makes them effectively stifle progress
  • For a casual developer, it’s impossible to know if she or he is infringing on any existing patents
  • Software patents are used in entirely wrong ways: Often patent claims are brought in after a product has shipped successfully, so in order to protect prior investment into research, they’re used as weapons of economic warfare.

So, software patents are bad, really, really bad. Unfortunately they are just as well a reality we will have to deal with for the foreseeable future. Make no mistake, software patents should not exist, they are evil and they should die sooner rather than later. Until then, however, the threat is real and needs mitigation.

The process of creating a software patent is, very roughly: You come up with a new idea, you write it down in a formal way, you register it with a patent office, the patent is reviewed and rejected or granted. One part of this review process is research for prior art. In order to grant a patent, it has to be an original idea, and it must not already exist.

The “must not already exist” is also known as prior art. Prior art must of course be known, so it has to be visible in the public space. Prior art that is useful in the War on Patents is right under the nose of the reviewers at the USPTO (U.S. Patent and Trademark Office), for example. In order to prevent patents from being registered, having as many technologies, ideas and concepts as possible registered as prior art is an effective way to fight the system with its own weapons.

The Open Invention Network sent us three smart people to Akademy, which was really useful. Apart from being cool folks, the OIN’ers also presented us a relatively easy way to record prior art as “Defensive Publication” and feed it into the system so it will be found, considered, and lead to rejection of similar patent applications. The nice thing here is creating a defensive publication is a lot less work than applying for a patent, it is in fact very well doable for an individual developer. A typical defensive publication fits on one page, has a few paragraphs of text with a generally applicable explanation of the idea and preferably a diagram or graphic that makes it easier to understand it. The key here is that it should be easy to understand for an alien to the specific technology (but technically savvy person nevertheless) to understand and find back in the case a related patent application ends up on the table.

There’s a few pleasant catches to this whole thing:

  • The process is lightweight, with a bit of exercise, writing a defensive publication about a good idea can be done in half an hour
  • The idea you’re recording doesn’t even have to be your idea: You can help a fellow developer by recording his work in this form. As you don’t derive any kind of license or copyright to the actual work by writing the defensive publication, the work can be spread.
  • The idea does not need to have been already implemented. Just being able to prove “somebody has already come up with that!” is enough, so even (consistent) brainfarts are eligible

By now, you might already get why I’m writing this: In order to effectively fight the war on patents, we need more people to write these defensive publications. As it’s quite easy to do and we have people who help us in this, I’d like to encourage you all to record prior art we’ve created.

Update: I’ve uploaded an example defensive publication of prior art about Plasma’s Activities, find it here.

3 thoughts on “Defensive Publications

  1. The OIN casts sand in your eyes. It offends me because they should know better. While it is true that in a patent case “prior art” is the most powerful means to win, “defensive publication” doesn’t really help unless you know about the specific attack vulnerabilility. Else it is quite a bit like getting permanent surveillance to avoid that you become the suspect in a murder case without an alibi.

    “Prior art” – patent professionals will dispute that – means de facto a prior document in the patent database. Patent offices do not search the internet. When your “prior art” is not documented in the form of patent legalese it is second class evidence and you rely on abitrary interpretation in Courts. Other issues, for instance, how is your document time-stamped and “published”? I am aware of so many anecdotal cases where professionals brought IT standard literature to the Courts and failed, because a combination of p.212-215 with p. 456 was considered “new”. It is hard to falsify novelty.

    A more workable method of “defensive publication” is to file a patent application (you can let the patent application lapse by not paying the renewal fees). Unfortunately the art of writing technical documentation and the art of writing patents are completely distinct professions. The perfect advise for “defensive publication” in the form described above would be a text generator that remixes existing patents ;-)

    1. hi!

      I work for Open Invention Network and I am one of the three people that was at Akademy. You raise some good points, but they are not new to us, so I’d like to comment on a few of the things you said and hopefully you will be less offended:

      * Regarding documents needing to be in the patent database, plus timestamping: the documents are put in a database on This database is searched daily by patent offices around the world to look for prior art. This has the same effect as being put in the patent database. Other entities (IBM, ISOC, Microsoft, NIST, Lenovo and many more) also put documents in this database. The documents are timestamped and trusted and (in some instances) also sent around in printed paper form. Digital timestamping is done by, who use a combination of digital fingerprints and Surety notarization for this.

      * You are correct that when you know of a direct patent threat it is much easier to deal with it. With a defensive publication you publish prior art and it might or might not prevent a patent application in the future. It is a bit of a fuzzy target and we readily admit that, but that doesn’t mean it is worthless. Even if a single defensive publication that you wrote avoids a patent lawsuit in the future because a patent was never granted, it’s a win, especially because it is very cheap compared to trying to invalidate an existing patent.

      * In Europe there already is a system in place that allows you to submit prior art when there is a patent application (“pre-grant submission”). In the US a similar system (but free and allowing anonymous submissions of prior art) will be launched in September. We will gladly assist with this as well.

      * Patent applications are indeed very powerful. Come talk to us if you want to file a patent.


  2. We should try to involve more people from our community for this kind of tasks, I would like to avoid to use precious developer time with this kind of stuff. When they ask how they can contribute to KDE a possible answer might be “write defensive pubblications”. We should start a defensive pubblication project as first thing anyway if we want to get people involved.

Comments are closed.