Also, patent trolls suck hairy donkey balls

August 10, 2011 12:31 PM

Drew Curtis' FARK.com has settled a lawsuit with a patent troll. The popular "not news" site was sued by "Gooseberry Natural Resources LLC" which held a ridiculous broad patent (6,370,535) that it claimed covered the basic concept of generating a press release online. Other sites targeted included Reddit, Digg, Slashdot, TechCrunch & Others. In the case of Fark, the suit was settled for $0. Curtis writes, "I paraphrased our best one-time settlement offer as "how about jack sh*t and go f*ck yourself."
posted by furiousxgeorge (79 comments total)

TAL: When patents attack.

The US Patent system is an abomination in its current form. In 2000 or so, someone patened *toast*. Patent is called something like "method to refresh bread".
posted by mcstayinskool at 12:36 PM on August 10, 2011 [3 favorites]

Huh. I had forgotten about Fark.
posted by everichon at 12:37 PM on August 10, 2011

That TAL episode was super frustrating to listen too... in a good way of course, but Christ, what assholes!
posted by JBennett at 12:41 PM on August 10, 2011

Huh. I had forgotten about Fark.

My feelings exactly; I left there when Drew decided to revamp the web site to push subscription dollars his way, making it impossible to use. Must have been, what, 6-7 years now?
posted by Old'n'Busted at 12:43 PM on August 10, 2011 [7 favorites]

Huh. I had forgotten about Fark.

Fark is the perfect online compliment to Metafilter. One is peanut butter, one is jelly. I won't get into which is which, but you really need both.
posted by squalor at 12:44 PM on August 10, 2011 [12 favorites]

Huh. I had forgotten about Fark.

I try to check in for the end of the year headline contests. There are times when the subbys over there are just effing geniuses.
posted by eriko at 12:46 PM on August 10, 2011 [1 favorite]

I've actually been spending a little more time on Fark than I used to lately, though I had completely missed this somehow. Wow. Hope the instigator hiding behind shell corporations gets outed as a result of all this. Actually, since he got the NDA stricken, does that mean Drew can name them himself?
posted by Gator at 12:46 PM on August 10, 2011 [1 favorite]

I don't think that you have to name interested parties until you actually sue.
posted by a robot made out of meat at 12:52 PM on August 10, 2011

Huh. I had forgotten about Fark.

So had I, actually. I used to enjoy the site quite a bit, but it got to the point where I couldn't even keep up with the amount of daily posts, never mind actually *reading* any of the comment threads. Especially not if I wanted to keep up my MeFi habit, which I far prefer.

a robot made out of meat: I'm confused, didn't they actually sue? All of the language seems to be "settled a lawsuit" and "the suit was dropped" and "sites were sued by" - were there no actual suits filed?
posted by antifuse at 12:55 PM on August 10, 2011

Fark is the perfect online compliment to Metafilter. One is peanut butter, one is jelly. I won't get into which is which, but you really need both.

It's more like one is peanut butter, and one is crap. You need both to make a good peanut butter and crap sandwich, but most people just prefer not to eat that.
posted by FatherDagon at 12:56 PM on August 10, 2011 [20 favorites]

Come on, Metafilter isn't THAT bad.
posted by furiousxgeorge at 12:57 PM on August 10, 2011 [33 favorites]

In 2000 or so, someone patented *toast*. Patent is called something like "method to refresh bread".

No, someone patented heating a bread product in an oven having a heating element heated to between 2500ºF and 4500ºF for 3-90 seconds. In tests this resulted in a bread product temperature of about 130ºF.

This is not at all like toast. The elements in a toaster are approximately 1500-1700ºF, and the toast reaches a surface temperature of about 310ºF in order to caramelize.

Anyway, the patent in question expired in 2008 for failure to pay the maintenance fees. Clearly it was of no particular economic significance, just like the great majority of patents.

were there no actual suits filed?

There was a suit filed in the Central District of California. The complaint is on the second link, in a vile docstoc Flash thing.
posted by jedicus at 12:58 PM on August 10, 2011 [3 favorites]

It's more like one is peanut butter, and one is crap. You need both to make a good peanut butter and crap sandwich, but most people just prefer not to eat that.

It feels that way to me sometimes as well, but I keep coming back to Metafilter anyway.
posted by squalor at 12:59 PM on August 10, 2011 [2 favorites]

Neotribalism derails post about a funny legal victory.
posted by Horselover Phattie at 1:02 PM on August 10, 2011 [22 favorites]

I read the RSS feed from Fark - it's a quick and dirty way to keep up with what's going on in the world. Anything substantial will be there, and lots of stuff that is interesting but otherwise useless.

And some of the headlines are fracking genius.
posted by COD at 1:03 PM on August 10, 2011

MetaFilter is the crap in that sammich!
posted by Bunny Ultramod at 1:04 PM on August 10, 2011 [1 favorite]

I had dumped Fark about the same time I started reading reddit and hacker news. I got sick of all the political topics getting rolled by a few right-wing trolls.
posted by SirOmega at 1:05 PM on August 10, 2011 [1 favorite]

"patent trolls suck hairy donkey balls"

Not very original.
posted by longsleeves at 1:06 PM on August 10, 2011

Huh. I had forgotten about Fark

I've tried to. But every so often I remember that my Fark account number is significantly lower than my metafilter user number and I can feel my face burning with the shame.
posted by dersins at 1:06 PM on August 10, 2011 [1 favorite]

(FARK quote above.)
posted by longsleeves at 1:07 PM on August 10, 2011

Patent trolls are the scum of the earth. My company is up against one right now. The case is absolutely without merit, yet we're spending tens of thousands to fight it.

I don't understand how the Supreme Court hasn't been able to curb this kind of behavior yet (I mean, I understand, but it's frustrating.) Trolls are severely damaging honest businesses every day.
posted by naju at 1:07 PM on August 10, 2011 [5 favorites]

Kudos to Drew and his lawyers for fighting the good fight. But man, eight months of legal wrangling still sucks.
posted by RobotVoodooPower at 1:08 PM on August 10, 2011 [1 favorite]

All "neotribalism" aside (dude, really? For expressing a preference for this site over another?), it's a bit ironic that Drew Curtis is being held up as some sort of hero for fighting someone trying to take credit for what they didn't deserve, given the context of this.
posted by Halloween Jack at 1:21 PM on August 10, 2011

boobies
posted by infinitywaltz at 1:29 PM on August 10, 2011 [2 favorites]

Next up, Ebaumsworld v. United States.
posted by benzenedream at 1:30 PM on August 10, 2011

FRIST
posted by eoden at 1:31 PM on August 10, 2011 [2 favorites]

But every so often I remember that my Fark account number is significantly lower than my metafilter user number and I can feel my face burning with the shame.

True dat. I log on and comment every couple of years using my super-low Fark user number, just to make a few heads asplode.

If it counts, I was lurking here the whole time I was posting there.

It doesn't count, does it?

*sigh*
posted by Capt. Renault at 1:31 PM on August 10, 2011

Actually, it was through Fark that I first learned of Metafilter, even though I haven't looked at Fark in years and years. So Fark deserves the credit for me being here.

Or blame, depending on what you think of my comments.
posted by Rangeboy at 1:34 PM on August 10, 2011 [2 favorites]

Meanwhile, I just saw one of Obama's economic advisers on The Daily Show talking about how we need to reduce the difficulty and time investment in getting patents. Which is only going to lead to more patents, leading to more of this bullshit. Or to put it in Fark terms:

FAIL
posted by formless at 1:35 PM on August 10, 2011 [1 favorite]

Low Metafilter user numbers don't count for that much, because public signups were closed for so long. Why no I'm not BITTER, why do you ask?

As for Fark, eh.
posted by JHarris at 1:36 PM on August 10, 2011

Low Metafilter user numbers don't count for that much

Dude, c'mon, that's, like, all I've got going for me here.
posted by Dark Messiah at 1:52 PM on August 10, 2011

Abolish software patents. Abolish process patents. Lawyers don't need a welfare system.
posted by mullingitover at 1:52 PM on August 10, 2011 [9 favorites]

TAL: When patents attack.

Recent MeFi FPP on the TAL episode: When Patents Attack.

Another recent and related FPP: Parent reform? LOL.
posted by ericb at 1:58 PM on August 10, 2011 [1 favorite]

A buddy of mine works at IV. When I mentioned that TAL story, he was a bit put off. So to make him feel better, I said "Hey, no worries. Another friend of mine, his company is running guns to mexico for the cartels, and the company I work for seems to be hacking phones. Your company isn't too bad."

He didn't think that was cool either....
posted by brando_calrissian at 2:02 PM on August 10, 2011

If it wasn't for Fark, I wouldn't be here. I'm sure that's disappointing to some, sorry ;)

But on topic: Interesting that a lot of the big names settled. I'm surprised to hear that. Of course, perhaps they're doing that because it's in their plans to troll patents themselves and they don't want precedents set that will limit this 'business' in the future. Hmm.
posted by zomg at 2:04 PM on August 10, 2011

I don't understand how the Supreme Court hasn't been able to curb this kind of behavior yet

It's really Congress that ought to change the system. But this year's patent reforms don't appear to go after the worst misuses of the patent system.
posted by grouse at 2:33 PM on August 10, 2011

MetaFarker? FarkFilter?

No, wait: MetaFarter.

My work is done here.
posted by davejay at 2:33 PM on August 10, 2011 [1 favorite]

Abolish software patents. Abolish process patents.

Don't forget patents on human genes.
posted by grouse at 2:33 PM on August 10, 2011 [2 favorites]

Good on Drew for getting the NDA struck. The part that makes me really grawr is how angry he is. "Imagine someone breaking into your home, then being forced to sit on the couch while their lawyers file motions over how much stuff they can take." It's just hideous how you can spend years on something, build your own technology and community and product, and then randomly one day some lawyer shows up and says "actually I own this idea, and you've been stealing from me". The moment that letter appears you're on the hook for tens of thousands of dollars, if not millions. I know a few people who've been victims of patent trolls and it's always incredibly stressful and ugly.
posted by Nelson at 2:57 PM on August 10, 2011 [2 favorites]

I've been involved with the patent process for several years now. First, I think software (and processes) should be protected by something akin to copyright, and not patents. That should be the long term goal.

I say long term because so much money has been sunk into the current system that there is absolutely no way software patents are going away anytime soon. Also, in most cases, software patents don't overlap as much as that TAL episode would have you believe. An analogy using music would be saying you have a patent on a certain chord progression, when in actuality you have a patent on a particular song using that progression. Another song can use the same progression, but be completely different (in tempo, instrumentation, octave, type of notes, etc).

The big issue is not that software is patentable (though it's not ideal). The big issue is the way the legal system enables patent trolling to occur.
Much abuse of the system could be forestalled by implementing rules such as:

- Have product
Patent owner must have functional product generally available to the same market segment as that of the company being sued (or show cause that actions by the company under suit prevented entry into the market)

- Show harm
Patent owner must provide a reason that their revenue is impacted by actions of the company being sued related to the illegitimate use of the patent

- Have skin in the game
The patent owner's must show that the *product* using the patent IP is a substantial part of their business model.

- Assume good intent
The patent owner must have made a good faith attempt at notifying and working out licensing arrangements with the company being sued, prior to bringing suit.

- Nobody has a time machine
The company being sued is not liable for any damages on infringing product being sold prior to infringement notification.

- Don't punish consumers
The company being sued, at their option, may continue to sell infringing product after being notified of infringement, however any terms ultimately reached will include any of those sales.

- It's not the idea, stupid
Patents, in and of themselves, are worth nothing without a marketed implementation.

There's probably other guidelines, but just migrating to those would be a good start, and probably more feasible than abolishing patents.
posted by forforf at 3:01 PM on August 10, 2011 [10 favorites]

I love a good "I told my lawyers to tell his lawyers to tell him to go fuck himself" story.
posted by louche mustachio at 3:17 PM on August 10, 2011 [2 favorites]

ditto.
nice read furious, tanky-tanky for posting it.
posted by clavdivs at 3:25 PM on August 10, 2011 [1 favorite]

Patent owner must have functional product generally available to the same market segment as that of the company being sued

This is called a 'working requirement.' Many countries have them (e.g. India and Turkey), and they are completely ineffective because there are lots of good reasons to have several broad exceptions that end up swallowing the rule. See this prior comment for a more thorough discussion. Your requirement would prevent universities, individual inventors, and firms from owning a patent and licensing it to others for development unless they themselves entered the market. This is very inefficient.

Patent owner must provide a reason that their revenue is impacted by actions of the company being sued related to the illegitimate use of the patent

This is required for lost profits damages, which are generally higher than a reasonable royalty, which is a floor on damages. So there's already an incentive to show that.

The patent owner's must show that the *product* using the patent IP is a substantial part of their business model.

This is basically the same as your prior point.

The patent owner must have made a good faith attempt at notifying and working out licensing arrangements with the company being sued, prior to bringing suit.

Judges in patent suits push for settlement talks all the time, and the federal courts have mediation programs at the district and appellate levels. However, I would support reversing the MedImmune decision, which made it harder to talk to an alleged infringer about licensing or settling without provoking a declaratory judgment action.

The company being sued is not liable for any damages on infringing product being sold prior to infringement notification.

This already exists. You only get damages for past infringement if a) the infringer is actually aware of the existence of the patent or b) the infringer is constructively aware because you marked your good or service with the patent number. "In the event of failure so to mark, no damages shall be recovered by the patentee in any action for infringement, except on proof that the infringer was notified of the infringement and continued to infringe thereafter, in which event damages may be recovered only for infringement occurring after such notice." 35 USC 287.

The company being sued, at their option, may continue to sell infringing product after being notified of infringement, however any terms ultimately reached will include any of those sales.

This amounts to forbidding preliminary injunctions, which would be a pretty drastic step with severe unintended consequences that I can expound upon at length if you want. Anyway, PIs are not super-common, especially in cases involving non-practicing entities (a set that includes 'patent trolls'). "A plaintiff seeking a preliminary injunction must establish that [it] is likely to succeed on the merits, that [it] is likely to suffer irreparable harm in the absence of preliminary relief, that the balance of equities tips in [its] favor, and that an injunction is in the public interest." Winter v. Natural Res. Def. Council, Inc., 129 S. Ct. 365, 374 (2008).

There's probably other guidelines, but just migrating to those would be a good start, and probably more feasible than abolishing patents.

Your suggestions are generally already in place, and several of them have been found not to work very well. This is a very well researched area, and it might be a good idea to read some of this research and history before making grand pronouncements about how to fix the system.
posted by jedicus at 3:31 PM on August 10, 2011 [3 favorites]

The patent owner must have made a good faith attempt at notifying and working out licensing arrangements with the company being sued, prior to bringing suit.

Oh, and the vast majority of patent issues never come to a suit and ~90% settle before a decision on the merits (i.e. summary judgment or a verdict). Litigation is expensive for everybody, not just defendants, and contrary to some people's belief patent trolls generally make money from negotiated licenses, not litigation. So there's already a very strong incentive to settle rather than sue. So it's not clear what good mandatory settlement talks would do.
posted by jedicus at 3:37 PM on August 10, 2011

In other software patent news Mark Cuban says we should get rid of them.

Any attempt to 'fix' them with various tweaks will just make patent trolls try to figure out ways around the rule changes and it will take years to re-do reform. Regular patents (for hardware, drugs, etc) make more sense bat software patents are just untenable. In particular because all software is in some way 'like' all other software. How do you determine if software is 'different enough' from other software? It's impossible.
posted by delmoi at 4:00 PM on August 10, 2011

Another friend of mine, his company is running guns to mexico for the cartels, and the company I work for seems to be hacking phones. Your company isn't too bad."

Your newscorp buddy actually should be ashamed. People, IMO, need to take some responsibility for working for companies that do bad things.

I say long term because so much money has been sunk into the current system that there is absolutely no way software patents are going away anytime soon.

The fact that money has been struck into them doesn't mean they shouldn't go away. Suckks for the people who wasted their money, but so what? Like I said, any kind of trying to tweak and reconfigure things to be 'just right' only creates a more complex gameboard for patent trolls and means even more money for patent lawyers to figure things out. Abolishing software patents is the only sensible alternative. Making things more complex just means more money for patent lawyers, and less money for software innovators.


---
Also, Fark sucks.
posted by delmoi at 4:09 PM on August 10, 2011

Why am I not surprised that Yahoo settled. Yahoo always fucking settles, but only after spending millions of dollars in legal fees. This is not a snarky comment -- I know this firsthand.
posted by bpm140 at 4:18 PM on August 10, 2011

Hilarity ensues.
posted by Uther Bentrazor at 4:19 PM on August 10, 2011

Reading Fark gives you a 9/10 chance of getting the mystery headline on Good News Week, so there's that.
posted by adamt at 4:26 PM on August 10, 2011

Oh, and the vast majority of patent issues never come to a suit and ~90% settle before a decision on the merits (i.e. summary judgment or a verdict).

Reverse that and force patent disputes to be adjudicated.

All patent disputes must have a decision + Loser pays = Patent trolls die
posted by Revvy at 4:46 PM on August 10, 2011

Abolishing software patents is the only sensible alternative.

First come up with an easy-to-apply definition of 'software patent' that can't be gamed and is neither under nor overinclusive.

Also, that abolition had better grandfather all existing software patents or else the government will have the mother of all takings suits on it's hands. Hundreds of billions in claims, easily.
posted by jedicus at 4:49 PM on August 10, 2011

All patent disputes must have a decision + Loser pays = Patent trolls die

That's a complete nonstarter both in theory and practice. Not that it might not have the intended result but it would be likely be impossible to implement constitutionally and would be extremely expensive and inefficient.
posted by jedicus at 4:53 PM on August 10, 2011

Not that it might not have the intended result but it would be likely be impossible to implement constitutionally and would be extremely expensive and inefficient.

So, you're saying it would or wouldn't be better than what we've got?
posted by Revvy at 4:58 PM on August 10, 2011

We should change the law so that instructions for a computing device are not patentable. We should also get rid of business method patents.
posted by humanfont at 5:32 PM on August 10, 2011

teh b00bies!
posted by rhythim at 5:50 PM on August 10, 2011

Twas Fark that brought me to MetaFilter. I lost interest in Fark after that.

b00bs!
posted by _paegan_ at 6:15 PM on August 10, 2011

jedicus: How informative does the notice have to be? There seem to be letters along the lines of "you're violating a patent that we hold, but to find out which one you'll have to pay and sign our NDA." I've thought that a helpful variant would be eliminating NDAs (eg linux would be able to patch out whatever it is MSFT claims they own, or as a community put together prior art). Is there a takings claim if congress makes it easier to challenge patents, or tightens standards for obviousness and enabling disclosures which apply to future suits?
posted by a robot made out of meat at 6:42 PM on August 10, 2011

"all rise"
[exit judge to chambers]

(Judge laughs into arrhythmia)
posted by clavdivs at 6:48 PM on August 10, 2011

Abolish software patents. Abolish process patents.

Don't forget patents on human genes.

It looked like it might happen after this case last year, but the decision was recently reversed.
posted by homunculus at 6:52 PM on August 10, 2011

jedicus: Oh, and the vast majority of patent issues never come to a suit and ~90% settle before a decision on the merits (i.e. summary judgment or a verdict). Litigation is expensive for everybody, not just defendants, and contrary to some people's belief patent trolls generally make money from negotiated licenses, not litigation. So there's already a very strong incentive to settle rather than sue. So it's not clear what good mandatory settlement talks would do.

A couple years ago during my patent law class that my professor had a slide from AIPLA, showing the relative costs of different sorts of litigation, and patent litigation dwarfing most everything else. My professor went on to explain that the vast majority of those costs stemmed from discovery, not the actual trial.

To that end, I think there are a number of administrative fixes that could help with the current trainwreck of a patent system:

1) Update the USPTO's own requirements for prior art research. Apparently, simple things like performing a simple Google search aren't required, or even used by many of the older patent examiners. The prior art and PHOSITA research by the USPTO ought to be damned near exhaustive. Adjust patent application costs accordingly.

2) A use-it-or-lose-it provision on patents. It could be flexible; for example, upon applying for a patent, the applicant has one year to either put his invention into as broad a market as feasible or license it to someone who will do the same. Allow the applicant to file for a one-year extension upon a showing of a good faith effort to do so. Failure results in the patent entering the public domain.

Don't want someone to just wait you out for two years? Don't patent it; rely on trade secret. Need to patent it because you're trying to beat a competitor who's working on something similar? Too bad, that's patently (hurrrrrr) against the purpose of patent law.

3) Take a cue from defamation law and defamation-proof people, and apply it to patent trolls. Does your company (or its parents, officers, etc.) have a history of suing larger companies in the eastern district of Texas using over-broad patents while having no actual product or market presence itself? Congrats, you are now legally patent-proof; your company, officers, and board members lose the ability to enforce patents for five years, you cannot transfer your current patent holdings or acquire new ones, and all of your current patents are presumed invalid for their duration.

Not to say any of this is a panacea or couldn't use further development, but some sort of administrative remedies are certainly in order. Patent law is a train wreck that, much like copyright, is probably doing more to harm innovation that help it.
posted by Vox Nihili at 7:26 PM on August 10, 2011 [2 favorites]

First come up with an easy-to-apply definition of 'software patent' that can't be gamed and is neither under nor overinclusive.

If it can be distributed as pure data, it's software. Done. How is that even a challenge?
posted by delmoi at 8:22 PM on August 10, 2011

The incentive in the current system is to make patents and nothing else. There are companies that just sell cold call your company after it has been hit with a patent lawsuit offering to sell you patents from their portfolio with which you can countersue the other party. If you never sell the product, but just patent then you can fire away without exposure. They will also look at your quarterly earnings and make a run at you if you look like an easy mark.

We've invented a system where it is profitable for the Intellectual Ventures guys to sit around the dinner table and brainstorm with a lawyer and write the ideas up as patents. No real research, just a pile of BS and clever writing. Just today I was on the toilet and thought it might be possible to develop a multitouch interface for a fly fishing rod. I wrote a claims for method, process and device patent throwing in 300 independent claims. Including photo analysis of environmental conditions to select the optimal lure. I did a google search for the prior art to reference, found none. I shan't be making a fly rod though I will be crushed by the big rod players with their portfolio of 3 million patents on everything related to fishing. See also razors.
posted by humanfont at 8:25 PM on August 10, 2011

A use-it-or-lose-it provision on patents

This was discussed above. How would it stop patent trolls? They're actively looking for people to license their patents to. That's how they make money.

My professor went on to explain that the vast majority of those costs stemmed from discovery, not the actual trial.

That's true of everything, not just patent litigation, and the fix there is to fix discovery costs, not hem in patents.

The prior art and PHOSITA research by the USPTO ought to be damned near exhaustive. Adjust patent application costs accordingly.

I have no inherent problem with this, but bear in mind that the 'winning' prior art is sometimes something that the Patent Office can't find (e.g. a not-available-online college thesis in a library in Germany, to name one famous example).
posted by jedicus at 8:27 PM on August 10, 2011

So, you're saying it would or wouldn't be better than what we've got?

It would be much worse. Most patent litigation involves practicing entities, believe it or not. And the vast majority of those cases settle, usually when both parties realize there's no point in continuing (i.e. not just because one party can't afford to continue fighting). So you'd be frustrating the intent of the parties, litigation costs would skyrocket for everyone, and you'd have hire a bunch more federal judges (who, by the way, get lifetime employment). That would be a full employment act for patent lawyers.
posted by jedicus at 8:29 PM on August 10, 2011

How informative does the notice have to be?

"For purposes of section 287(a), notice must be of "the infringement," not merely notice of the patent's existence or ownership. Actual notice requires the affirmative communication of a specific charge of infringement by a specific accused product or device." Amsted Indus. v. Buckeye Steel Castings Co., 24 F.3d 178, 187 (Fed. Cir. 1994).

Vague informational letters generally do not count as actual notice. They are vague so as to avoid creating declaratory judgment jurisdiction under MedImmune.
posted by jedicus at 8:36 PM on August 10, 2011

If it can be distributed as pure data, it's software. Done. How is that even a challenge?

What is "it?"
posted by jedicus at 8:36 PM on August 10, 2011

jedicus: This was discussed above. How would it stop patent trolls? They're actively looking for people to license their patents to. That's how they make money.

I don't believe it would stop trolls; not entirely at least. But it would help in two ways:

1) Trolls tend to use old patents. Take the '535 patent used in the Fark case, for example. The filing date was 1999, and it was granted in 2002. I'm going to hazard a guess that their enforcement/licensing has been pretty spotty since then. Submarine patents may have been "fixed," but trolls frequently sit on the patent, doing nothing with it until that market segment is ripe enough to sue extort. They just can't perpetually amend it anymore.

Sitting on enforcement rights ought to aggressively used by the courts to strike down patent validity, absent some showing of inability to enforce (e.g. failure of licensing talks, lack of knowledge of infringement). If memory serves, such a court construction does exist, but I can think of few cases of it being enforced.

(Bear in mind, I got my J.D. in 2010, but I'm not a lawyer and never sat for the patent bar; no more access to Westlaw/Lexis to fact check myself, either)

2) It would reduce the patent arms race and bring more patents into the public domain. Google, IBM, Microsoft and the like have massive patent portfolios, where entire companies (and their patents) will be snatched up in the name of acquiring a handful of defensive patents. Since many of the patents cover subject matter that is not part of the company's current business, companies amassing patent portfolios full of things they aren't using would actually cause patents to enter the public domain faster than normal.
posted by Vox Nihili at 4:44 AM on August 11, 2011

I remember back in 2001 when I was graduating from university (holy balls, 10 years ago??) Amazon came to my school for a recruiting event. The recruiters (I can't remember who, exactly was there... But they were pretty high up in the chain - Waterloo is a pretty good computer science school) were bragging quite extensively about what an awesome coup their "One Click" patent was. Made me want to puke. There was another Texas company (whose name I can't remember - ahhh, found em! Trilogy) that recruited QUITE heavily at my school, and they said that they actively encouraged employees to work on patentable algorithms/processes, offering bonuses to employees who managed to create patents for the company. That turned me off right quick. Also, they kinda felt like a cult -your first month at the company, no matter what job you had, was in this weird boot camp exercise thing. *Shudder*

First, I think software (and processes) should be protected by something akin to copyright, and not patents. That should be the long term goal.

Well, to be fair, software *is* protected by copyright. IANAL, rather I am a programmer, and I've just never understood the concept of a software patent. Software patents (at least the few I've bothered reading about whenever one of these stories comes up) always feel like vague hand-wavy "Doing X on a computer" patents of IDEAS, as opposed to patents of actual inventions. Of course, I haven't bothered to actually go and read the actual patents, because YAWNSVILLE, so maybe they're more precise than that... But if they're that precise, how is it that so many people seem to be coming up with the EXACT SAME WAY of doing things independently? And if there are that many people coming up with these ideas independently, doesn't it fail the "non-obvious" requirement?
posted by antifuse at 6:05 AM on August 11, 2011

If memory serves, such a court construction does exist, but I can think of few cases of it being enforced.

It's called laches, and it's an equitable defense. It comes up in some cases, but you are correct that it is not very common with patents. It's considered on a product-by-product basis, so the (typically multi-year) clock restarts every time the alleged infringer substantially changes their product or introduces a new infringing one. But sure, laches is something that could be tweaked a bit.

Since many of the patents cover subject matter that is not part of the company's current business, companies amassing patent portfolios full of things they aren't using would actually cause patents to enter the public domain faster than normal.

In practice, not really. Working requirements always have so many (necessary) broad exceptions that it's almost never a barrier to maintaining a patent. And remember, working requirements mean a significant annual hassle for 'legitimate' businesses, research organizations, and individual inventors.
posted by jedicus at 6:22 AM on August 11, 2011

Say what you will about the Patent office, if it weren't for them, we'd never have had the Monsters of Megaphone tour.
posted by Uther Bentrazor at 6:22 AM on August 11, 2011 [1 favorite]

Talking about Fark on Mefi feels a little like re-connecting with your old heroin buddies at church.
posted by pjenks at 7:01 AM on August 11, 2011 [2 favorites]

jedicus, every time we have a patent discussion on Metafilter you come in and tell us how the current system has to be the way it is and no proposed changes will work. How would you fix things? Or do you think the system as it operates today is fine?
posted by Nelson at 7:26 AM on August 11, 2011

every time we have a patent discussion on Metafilter you come in and tell us how the current system has to be the way it is and no proposed changes will work.

I apologize if it comes across that way. I don't think the current system has to be the way it is. But I think the status quo is less bad than many proposed changes, which are often impossible to implement, politically or economically impractical to implement, wouldn't be an improvement, or wouldn't be a cost-effective improvement.

How would you fix things?

Lots of ways. On the administration side: increase the Patent Office budget substantially, make it easier for people with CS backgrounds to become patent agents and patent attorneys (so that more applications will be written by people who understand the technology), open satellite Patent Office branches in or at least near tech hubs like Silicon Valley, and move to an annual maintenance fee (right now fees are assessed every ~4 years; an annual fee encourages patentees to review the value of a patent more frequently and it's also the global standard), give examiners more time per application, pay examiners much more in order to improve the Patent Office's abysmal retention rate, and require patent assignees to record the real party in interest to reduce the incidence of gaming the system with shell companies.

On the examination side: tighten up the written description and enablement requirements considerably, and move to source code rather than flowcharts for describing software.

On the litigation side: give more deference to district court determinations of claim construction to reduce the incidence of appeals, reduce the standard of proof for invalidity at least when submitting prior art not considered by the Patent Office, make fee shifting more common (i.e. loser pays in more cases), appoint (and confirm) more district court and Federal Circuit judges to decrease the case load and speed case resolutions. There are dozens of district court and two Federal Circuit appointees who have yet to be voted on by the Senate, including several in districts that have a lot of patent litigation. But filling the empty seats isn't enough; the absolute numbers need to go up as well.
posted by jedicus at 7:55 AM on August 11, 2011 [4 favorites]

Thanks jedicus! Sorry to call you out like that, but I figured you'd have an interesting answer.

From my perspective as a software guy, the main problem with patents in my industry is the standard of non-obvious is way, way too weak. There are way too many software patents which are basically "this common technique, but.. on the Internet!" or "this thing we all learned in CS 101, but... on mobile devices!". Better patent review might help that problem, but I think it has to be matched with a higher standard of what is patentable.
posted by Nelson at 8:23 AM on August 11, 2011

There are way too many software patents which are basically "this common technique, but.. on the Internet!" or "this thing we all learned in CS 101, but... on mobile devices!"

Fear not, for the Federal Circuit has heard your cries! There are lots of cases holding exactly that kind of thing obvious, including:

Muniauction, Inc v. Thomson Corp., 532 F.3d 1318 (Fed. Cir. 2008) (holding that conducting previously known methods of bidding through an Internet web browser was obvious because it amounted to no more than applying the use of the Internet to existing electronic processes at a time when doing so was commonplace")

In re Mettke, 570 F. 3d 1356 (Fed. Cir. 2009) (finding it obvious to add Internet access to a prior art kiosk that included a fax-machine)

Western Union Co. v. Moneygram Payment Systems, 626 F. 3d 1361 (Fed. Cir. 2010) (finding it obvious to add the use of internet-based communications, specifically the TCP/IP protocol to an earlier invention).

The district courts have gotten in the game as well. See, e.g., Papyrus Tech. Corp. v. N.Y. Stock Exch., LLC, 653 F.Supp.2d 402, 432 (S.D.N.Y.2009) (finding that adding "a connectionless protocol (TCP/IP) that has been used in electronic communications, such as the Internet, since the 1980s" to other obvious elements of claim "adds nothing new to the field of endeavor"). That case was later affirmed by the Federal Circuit in a nonprecedential per curiam opinion.
posted by jedicus at 9:26 AM on August 11, 2011 [2 favorites]

I shouldn't ask multi-barreled questions. Sorry to be interrogating.
1) Can only the courts retroactively change patent standards (like tightening enabling and obviousness)? Or can congress issue a resolution "we meant X by Y" which applies to already-issued patents (perhaps invalidating them) without it being a taking?

2) Could congress require registration of patent licenses? That seems like it would eliminate some of the bad behavior by giving everyone in an industry a heads-up (to either stop infringing or find arguments) when the first company gets targeted.
posted by a robot made out of meat at 1:50 PM on August 11, 2011

Can only the courts retroactively change patent standards (like tightening enabling and obviousness)? Or can congress issue a resolution "we meant X by Y" which applies to already-issued patents (perhaps invalidating them) without it being a taking?

I'm not an expert on takings, so I'm not sure where the line is there. You should probably ask Richard Epstein. But I think as long as the rule isn't targeted at an identifiable class of patents or patentees then it's probably okay. The Supreme Court's regulatory takings case law is kind of a mess, but simply being harmed by a regulation or law, even more severely than others are harmed, is not enough. "It is, of course, true that the Landmarks Law has a more severe impact on some landowners than on others, but that in itself does not mean that the law effects a taking." Penn Central Transp. Co. v. New York City, 438 US 104, 133 (1978).

Compare that to expressly rendering unenforceable a very narrow category of patents (as was attempted with Data Treasury's patents on electronic check processing) or even a broader but still well-defined category like software patents. In the former case it was widely assumed that it would qualify as a taking. I believe the latter probably would as well.

Could congress require registration of patent licenses?

Probably so, but I don't think it would help. First, lots of licenses are just bog-standard business deals, so there would be a lot of chaff to sort through. Second, it would burden many businesses while only providing a small benefit to a smaller subset. Third, most patent trolls are quick to tout the license agreements that they have made anyway. Intellectual Ventures' press release archive is one such announcement after another, for example.
posted by jedicus at 2:22 PM on August 11, 2011

What is "it?"

Seriously? That's what you're going with? I don't think I can answer that without you telling me what "is" means since otherwise how can I understand the question!?

(It in this case would be the thing being patented, it would be more precise to say that a purely digital/informational representation of something can't violate a patent)
posted by delmoi at 10:04 PM on August 13, 2011

Seriously? That's what you're going with?

It was an honest question. You used 'it' without any apparent antecedent basis. That doesn't make for a very precise definition, and I wouldn't want to make any assumptions.

it would be more precise to say that a purely digital/informational representation of something can't violate a patent

So as long as the patent involves a physical input (e.g. sensor data) or a physical output (e.g. controlling a machine), then that wouldn't fit your definition, correct? In that situation there is an informational representation of a physical thing (e.g. the digitized sensor data or the representation of the machine), but the physical thing itself is also a necessary part of the claimed invention.

So if I understand correctly, your definition would exclude only so-called "pure" software patents (e.g. a generic sorting algorithm that runs on a generic computer). As luck would have it, that's arguably what the machine-or-transformation test does. It's not the only test for patentable subject matter, post-Bilski, but it's a test.
posted by jedicus at 6:10 PM on August 14, 2011