Summary: OpenMediaNow? is the container for this effort.
Short term goals:
Medium term:
Long-term goals:
Present: Ed, Benjamin, Rob, Chaals, Tim
Topic: What did we want?
CMN: Basic need is to show that this stuf is clean in public-record
RS: Lawyers can argue this endlessly.
BO: Why did SCO end up being a clear victory?
RS: Because the code they claimed was stolen from someone else - i.e. no case to answer. Otherwise, it might have been successful
CMN: What is out there already?
RS: A problem is that if you knowingly violate a patent you get punished more, so people are afraid to research it.
TT: There is pressure to make that go away...
RS: Yes, there is talk...
...we start researching patents and it sets us up for danger.
TT: as a developer, it is hard for me to say "I know about this and that". Supposedly it makes sense according to someone to specifically document everything.
RS: Trick is documenting your own work, because that turns into prior art.
TT: There is documentation of specific stuff for CELT
CMN: There is value in documenting what you did to show that you made a real effort to avoid knowing infringement
TT: One of the things Nokia was upset about was that there is a detailed spec, so anyone with an open application can file an amendment.
CMN: Yeah. Should be illegal, but it is possible.
TT: There is a first-to-invent law in the US, but in the rest of the world it is first-to-file. In disputes in the US it becomes very expensive - although in principle you can get an application invalidated. You just need to have a lot of money to play the game.
... Question is how to make the situation so clear that people can't sue.
... On2 stated that they would not assert patent claims against VP3?. We don't necessarily know what those are.
CMN: First step is to get a list of things that are 20 years old, and cannot have a valid patent anymore. Second thing is to find out what patents On2 has that we can claim are therefore clean technologies.
TT: Difficulty is that you cannot search by company.
RS: We are going to work with Groklaw and start collecting this stuff to make a body of material people can mine. Mozilla and Google have done it but won't release results.
TT: I am trying to work on this
RS: Seems like the best hope is someone accidentally leaving an envelope full of papers on a bench....
RESOLUTION: Short term goal - harass Mozilla, Google, Xiph, On2, BBC (and anyone else we think might know something) to release any existing research they have.
RS: There is an issue with Android that it is supposed to be patent-free. It turns out that a lot of device creators already have a license (or don't care)
BO: ...until they get raided in some European trade fair.
EH: Is there prior art on container formats? The only one we could find related to Qt was RTP hinting.
RS: The patents are very specific to pieces of the technology that are required to implement effectively.
CMN: The issue to prove clarity is that we need to go through the functionalities one by one and show that they are each clean.
RS: One problem is that people think it will cost a huge fortune, so they don't start.
BO: The question is whether you go through patent by patent, and see if they cover anything, or whether you are better to go through the spec piece by piece and compare it.
RS: We think you need to research everything for prior art, and at that point you can work out how to do non-infringing implementation in order to get the patent-holders to negotiate.
RS: There are problems getting Moz/Google to have anything to do with this.
... Best way to duplicate their research is to throw lots of bodies at it.
BO: Current mindset is that you can't do anything about patents and should just pretend they don't exist. Might be useful to try and combat that mindset.
RS: You need support to do stuff that is legal, and to be someone that can't be hurt by having patent rights revoked.
RESOLUTION: Long-term goal - convince Open Source developers that publicly documenting stuff is a good anti-patent strategy.
CMN: Applying for patents is a very good way to establish something as prior art. Another strategy is to publish something in a journal.
TT: IBM has a journal basically for that purpose
EH: Would check-in logs work for that?
TT: There are issues with it - there is some push for it, but it isn't established yet.
EH: So is there an established list of MP3? patents?
BO: There is a list of things you pay for to make an MP3? player
TT: There is no guarantee that the list is complete, not that they are genuinely required.
RS: There are some of them that are expired, or about to expire, and some that can probably be invalidated, but there are no changes to the license term.
EH: How hard is it to change the intent of the implementation?
RS: Not very hard, costs maybe 20k$ - if you are working around other stuff then the costs can go up pretty fast.
CMN: Problem is that you need stuff good enough to convince corporate lawyers that it is reliable
RS: Which is why we want to work with FSF, EFF, ...
RS: We started some analysis with the EFF. They dropped it but have picked it up again. We are looking for funding to hire someone to move it forward and to give Groklaw money for infrastructure. It is something around 50k$. We want to give 10k$ to Groklaw to get started, and they will start collecting data.
EH: Isn't there some work that could be done by people with other skillsets - Open Source world.
RS: This is what Groklaw is good at - getting the research done with the necessary level of quality.
RESOLUTION: Get 10k$ for Groklaw, to get started, 50k$ to pay for work-time put into it, as short- and medium-term goals.
TT: I know what is in Theora, so I can provide engineering expertise.
RS: I like the idea of getting enough data together to do clean-room implementations. Tim, would it help you if we have a collection of papers and ask you to look at them?
TT: Absolutely.
EH: Next step is to sweep through patent stuff with keywords...
TT: Yes, but that is often very difficult in patent applications, which can be defined very obscurely to make that hard.
RS: There are areas that are probably worth focusing on for clean-room work
EH: Look at files used across multiple codecs.
RESOLUTION: Look for and document expired patents in this area. Short term goal.
RESOLUTION: Short-term goal - look for people who will put their names to this project.
RESOLUTION: OpenMediaNow? is to blame^W^Wthe holder of the balls^W^W^W^W^W the container for this effort.
RESOLUTION: Get people to tell the story of why people are paying for this stuff.