Ask Slashdot: Open Patent Licenses?

felipe13 writes "We are working on a new piece of code that will be protected under a GPL license, this is fine for the code itself, but what about our 'innovations'? Are there any 'Open Patent License' models similar to the GPL or Creative Commons? We have Google patenting the highlight of search occurrences, Facebook protecting the word 'Book,' and Apple registering body movements. This is becoming ridiculous to a point. Now the patent trolls are making a killing as well. Does the open source community has a good way to protect its innovations and inventions? There are some initiatives to buy patents and release them to the public or at least place them is a protected area, but where would my very small company register a new way to include titles in a private message? Where could Drupal patent the use of 'hooks' to let developers interact with the core of the application? (If they invented this, I am not really sure.) I don't want to wake up in 10 years and discover that X huge company patented my innovation and that now I actually have to pay them for it." There's OpenPatents.org, there's the Open Source Hardware and Design Alliance, there's CERN's newly-updated Open Hardware license, and there are domain-specific patent sharing organizations like the Open Patent Alliance; what else is out there?

Comments

Patent it

[sakdoctor]

GPL is based on copyright which is automatic.
Patents are not automatic, so you need to patent your invention, and worry about (free) licensing afterwards.
If you wanted something more BSD-like, publish without patenting.

Re:Patent it

[poetmatt]

I don't get why open innovation network wasn't mentioned? That's pretty much defensive patent licensing for free, and clearly defined.

Re:Patent it

[Ruke]

Copyrights are expensive, need to be filed, and don't really do much until you sue someone for violating your patent. An "Open" patent doesn't make much sense; who owns the protected innovation? Everyone? Who can be sued for using it? No one? People who use it in non-GPL'd code? Who is going to initiate the lawsuit?

Prior art, not prior patent, is what kills a later, competing patent. However, both cases are handled exactly the same way: through expensive lawyers. In this case, even when you're in the right, the big corporation will still win. They can afford better lawyers to defend their patents for longer than you can afford to sue them.

Re:Patent it

[rtfa-troll]

First not only am I not your lawyer, I'm not a lawyer at all, so this isn't legal advice; if you have questions like the above you probably want to go talk to the FSF or SFLC who are quite likely to be willing to arrange cheap legal advice.

What sakdoctor said is probably true that we need people in the free software community who actively patent. Having patented it however, the licensing not be like the GPL. You also don't want to get involved in people like the Open Patent Alliance. They seem to be a standard industry "patent pool" and as such their members will be looking for money from WIMAX producers, which of course basically rules out FOSS where it is impossible to control the number of copies out there. The most important thing to understand is that where in copyrights, copying is relatively cheaply and efficiently proven and so people respect them by default, the standard approach to patents is to ignore them.

In licensing, basically a level of aggression is needed otherwise your patent won't achieve anything and won't be worth the money. Basically, you have to be willing to sue companies that are willing to get involved in suing open source companies. That includes any company which has ever owned a patent involved in suing open source companies. If you don't do this then people like Microsoft will just sell their patents on to patent trolls whilst keeping a license for themselves and making sure the trolls know which targets to prioritize in order to get their next fix of patents. Probably the best way to do that is to transfer the patents to a well funded and very aggressive non practicing entity together with an agreement that a) they can't sell the patents on except with a similar agreement and b) they will not sue companies which agree never to sue FOSS software companies or to sell on patents except with a similar agreement.

The one organisation that I know of which comes even close is the Open Invention Network but they have completely failed to sue Microsoft during their recent patent attacks and as such I would say they seem to be pretty ineffective.

Finally, there's a moral question here. Some people believe that if you don't believe in a law you shouldn't benefit from it and so you shouldn't sue over patents. I wouldn't want to get money from patent lawsuit, but I strongly have the opposite belief. As long as patents only cause pain for FOSS people, the corporates and the politicians they buy will continue to support them. The correct way to get rid of the patent system is to maximise the pain for companies like Microsoft who are abusing the system themselves.

Document, document, document

[bennomatic]

The thing that kills patents is prior art. So if you're releasing something to the public domain, make sure to document every step of the way, everything that is potentially patentable. If you invented it first, then with proper documentation, nobody else can patent it...

...at least until the new "whoever files first" rules go into effect.

"Files first" does not invalidate prior art.

[perpenso]

...at least until the new "whoever files first" rules go into effect.

I do not think that is the case. I believe that under both "files first" and "invents first" prior art can cause the patent application to fail. I think the individual hurt by the move to "files first" is the inventor trying to work in secret, trying to postpone filing for the patent in order to maximize the number of years on the market during patent protection and minimize the number of years under development during patent protection.

Re:Prior art

[kimvette]

If you publish an innovation from which you do not intend to profit, and which you would not mind others replication or even would like others to replicate, simply do not file a patent for it. Should anyone ever attempt to do so, your work will qualify as prior art.

  your work will^H^H^H^Hshould qualify as prior art

I fixed that for you. In this day and age where a patent clerk's "research" consists of searching for his rubber stamp to approve the app and grant the patent (usually on stuff a simple google query would show a googleplex worth of results for), you have to wait to duke it out in court and go broke defending yourself against a patent which should never have made it TO the application process, let alone be granted patent protection.

Statutory Invention Registration

[pavon]

Alternately, if you wish your invention to be in the public domain, you can file for Statutory Invention Registration. This will cause the filing to be in the USPTO database and thus more likely to be found during a prior art search for future patent applications than if it was just published in a journal.

Re:Statutory Invention Registration

[Dachannien]

What the parent poster says is correct. Also, SIRs are cheaper than filing a regular patent application.

The pending patent reform bill would eliminate SIRs, however, probably in part because publication is automatic at 18 months. All published patent applications go into the searchable database that examiners use.

If you file a patent application (not a provisional, mind you), it will publish 18 months later as long as you satisfy various formal requirements at filing (payment of fees, properly executed declaration, and a few other things). There's no obligation to pursue the application to get an actual patent, and if you're not planning to get the patent, then with a little research you can probably handle the initial filing pro se, leaving the USPTO fees as the only expense.