5 ms·
Question, So do you happen to know if the troll needs to prove that the company violated a claim, or does the company need to prove that it DID NOT violate the
by streetcat1 7y ago
Question,
So do you happen to know if the troll needs to prove that the company violated a claim, or does the company need to prove that it DID NOT violate the claim?
- chx 7y agoThis will be fought on an entirely different level: the patent is bogus in the first place. There's nothing patentable about it. That's how you hunt trolls. Merely proving you didn't violate their patent is not helping the next guy (and actually, you might have violated the patent, who knows with these frivolous things). Killing their patent does.
- anonsivalley652 7y agoThere ought to be a super PAC / legal collective in US whose primary goal is to basically hack the system by using it against itself for the net effect of mostly eliminate the patent system with the chief aims of: 0. making it harder to get a patent (not adding arbitrary bureaucracy, but ensuring examiners maintain high and fair standards) 1. making it harder/shorten the time to keep a patent EFF, ACLU, etc. might nibble around the edges on this issue, but there's no one going after the politicians with lobbyists, a bucket of money and a deep bench of patent attorneys funded by something like a Kickstarter/IndieGoGo and/or subscription model to the tune of a Bernie Sanders-equivalent funding level.
- rayiner 7y agoThere is an enormous, well funded, Silicon Valley lobbying effort directed at weakening the patent system. The last Patent Office director was head of Patent Strategy for Google for almost a decade. However, almost all the other industries, from automotive to pharmaceuticals to aerospace, and even some of the more traditional players in Silicon Valley, are on the other side of the issue.
- uep 7y agoNo matter how detrimental software patents are to the big players, they are far more detrimental to the small players. If it can keep the smaller players from being real competitors, it's in the best interest of the big companies to just pay the patent tax. They stand to lose far more with a lower barrier to entry to their markets. Software is kind of unique (and even moreso now with the prevalence of cloud providers) with its otherwise low barrier to entry; as capital expenses are extremely low compared to other industries.
- bdowling 7y agoThe burden of proof is on the plaintiff (patentee). The defendant can win by showing either invalidity or non-infringement. Theoretically, there is a presumption of validity of a patent because applications are carefully reviewed before issue. However, in practice that presumption is usually challenged early in litigation. So, the plaintiff ends up having to prove validity of the patent and infringement by the defendant.
- streetcat1 7y agoThanks for the answer. So based on Alice, etc, where a lot of software patents got invalidated. If a patent got issued pre-Alice, can a defendant invoke Alice as a defense? Is there a way to do that without going to trial?
- rayiner 7y agoAlice is just an interpretation of the pre-existing Section 101 requirement for patent eligibility, so it can be invoked against any patent. Section 101 and Alice can be invoked in a motion to dismiss early in litigation.
- streetcat1 7y agoThanks for the response. So to file a motion to dismiss, is there a USPTO process that one can follow, or there must be a real trial? My goal is to find the fastest (and the cheapest) way to starve the troll.
- rayiner 7y agoYou file the motion to dismiss in the district court where the troll sues you. If you win on the motion to dismiss, the case never goes to trial.
- streetcat1 7y agoThanks. You sound very knowledgable. So if I think about it correctly, there are three ways to dismiss it. 1) Invalidate the patent on the basis of novelty (101). 2) Invalidate the patent of the basis of the prior art. 3) Invalidate the claim on the basis of non-infringement. [ Am I missing something? ] Which one do you think is the best strategy (with the most chance of success? Or based on your experience, what will likely work.