Join 3,367 readers in helping fund MetaFilter (Hide)


How Yahoo Weaponized My Work
March 13, 2012 7:30 PM   Subscribe

A Patent Lie: How Yahoo Weaponized My Work by [MeFi's own] waxpancake, aka Andy Baio.
posted by stavrosthewonderchicken (97 comments total) 8 users marked this as a favorite

 
Mark Cuban is perversely rooting for Yahoo.
posted by mullingitover at 7:39 PM on March 13, 2012 [2 favorites]


I read Yahoo! Crosses The Line earlier today and was saddened that Yahoo had joined the ranks of patent trolls. Will never again use YUI or any other of Yahoo's code libraries.

I am wrting this in outrage at Yahoo! I used to care about that company for some reason. No more. They are dead to me. Dead and gone. I hate them now.

.
posted by Foci for Analysis at 7:40 PM on March 13, 2012 [3 favorites]


Mark Cuban is right in that the patent system is terribly broken. Here's another, equally bleak, story about software patents:
I started writing a paper describing my algorithm using mathematical formulas. Sounds logical, right? But remember, it is forbidden to patent a mathematical formula. I was thus asked to describe everything using boxes and arrow, showing how an input was affecting the final output.

My job became hell on earth. I had to spend my days translating very elegant formulas into dumb boxes. Each time, I received comments like: "make it less software", "hide the fact that we are patenting a formula" or "make it more confusing. You don't want competitor to be able to implement your patent, right?".
posted by Foci for Analysis at 7:50 PM on March 13, 2012 [7 favorites]


Mark Cuban is perversely rooting for Yahoo.

That's fine if you assume it will actually change things - and don't mind Facebook as collateral damage, I suppose.
posted by Artw at 7:55 PM on March 13, 2012


People who get paid to do things are often incredibly ignorant of where the money comes from to pay them. Hence, the Web 1.0 guys who are shocked, shocked I tell you that the developers after them aren't getting paid shit and are all for some free web that denies their sucessors the titanic stock options their predecessors had.
posted by Ironmouth at 8:00 PM on March 13, 2012


Nice piece, too bad about Yahoo. Is there a credible counterargument anywhere to this position, that software patents shouldn't exist (since it's already copyrighted)?
posted by churl at 8:04 PM on March 13, 2012


Will never again use YUI or any other of Yahoo's code libraries.

Don't throw the baby out with the bath water. YUI is BSD licensed, so fork it and move on.
posted by ddbeck at 8:08 PM on March 13, 2012 [6 favorites]


When I was a wee lad I had a conscience and declined to include my name on patents that were filed on my work.

And then I kept getting subpoenaed as defending lawyers tried to get the patent invalidated on the basis of a major contributor being shorted credit.

Now that I'm an old fart I just sign the damn things and let them go off into the world to be used for attack or defense or both. The patent/copyright system is whacked out but unless someone screws the pooch on the level of GIF, the general public will never really care.
posted by Tell Me No Lies at 8:18 PM on March 13, 2012


Ask any programmer — developing software is as creative and unique as writing poetry.

Okay, programmer over here. You are a dangerous hack and should stick to playing around with shareware instead of pretending to be a professional.

You are paid to be a craftsman, not a poet. You should be using the same code over and over as much as humanly possible. You're laying bricks, not writing a sonnet.

Every once in a while there's a tight loop or a tricky interaction where you get to exercise your creativity, but 95% of your production output should be boring and predictable as hell.

Signed,

The developer who just inherited your poetic magnum opus
posted by Tell Me No Lies at 8:30 PM on March 13, 2012 [54 favorites]


Is there a credible counterargument anywhere to this position, that software patents shouldn't exist (since it's already copyrighted)?

Software running on a general purpose computer, which are what software patents basically claim, serve essentially the same function as specialized electronic circuits. Almost no one would argue that electronic circuitry that have been physically wired to perform functions should not be patentable (unless you're not a fan of patents in general), so why shouldn't hardware that's been programmed by software to perform those same functions also be patentable?
posted by gyc at 8:46 PM on March 13, 2012


I started writing a paper describing my algorithm using mathematical formulas. Sounds logical, right? But remember, it is forbidden to patent a mathematical formula. I was thus asked to describe everything using boxes and arrow, showing how an input was affecting the final output.

This is a stupid byproduct of the Patent Office's insistence of the use of flowcharts rather than pseudocode to describe algorithms. It's terrible: it makes the patents long, hard to write, hard to understand, and virtually impossible to search. The Patent Office fought tooth and nail against recognizing computer-implemented inventions, so when the courts finally forced them to do so, the PTO found itself without any technically qualified computer scientists to give them decent guidelines for how such patents should be drafted.

There's nothing in the patent law that requires the use of flowcharts. In fact, I would argue that § 112 requires pseudocode or actual code in a commonly-used language: "The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same." (emphasis added).

Pseudocode or actual code are much more clear, concise, and exact than flowcharts, and people of skill in the art are much more likely to be able to make use of code than a flowchart.
posted by jedicus at 8:48 PM on March 13, 2012 [3 favorites]


Is there a credible counterargument anywhere to this position, that software patents shouldn't exist

IAAPA, but IANYPA. I also developed software for 20 years. I know a thing or two, and have an opinion to go with.
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor...
35 U.S.C. § 101
One can argue about whether we should have patents at all, but I don't think one can credibly argue that software patents are any different from other types of process patents. Software is simply instructions to cause a computer to perform a process, and if the process is new and useful, current law says you can get a patent.

The problems that show up especially clearly with software patents are that USPTO examiners are abysmally bad at examining such applications – many examiners lack software engineering experience – and that the patent applications themselves are often written by lawyers who are either ignorant of how computers work, or actively obfuscating the invention to pull the wool over the examiner's eyes.

The result, often, is a patent that's so hand-wavey as to cover huge tracts of idea-space, which the inventor might not even have envisioned, and which the draftsman certainly didn't describe adequately to allow a person of ordinary skill in the art to make or use the invention.

My pet idea for reforming the system is to require the inventor to deposit a working version of his program. That is, after all, the best description of the invention, and the one that others of skill in the art would most easily understand and learn from. And later on, if the patents claims are asserted against some feature of an alleged infringer's program, if the deposited program could not itself perform the accused function, then the alleged infringer could not have learned how to do it from the patent disclosure, and the suit should fail.

Until then, I'll continue to write meticulously detailed and technically correct patent applications for my software clients, and argue as effectively as I can against examiners who don't see the novelty, or can't understand it.
posted by spacewrench at 8:51 PM on March 13, 2012 [3 favorites]


Okay, programmer over here. You are a dangerous hack

Is it just me, or do more and more Metafilter members these days go out of their way to be obnoxious to other members whose work gets linked here?
posted by stavrosthewonderchicken at 8:51 PM on March 13, 2012 [21 favorites]


Is there a credible counterargument anywhere to this position, that software patents shouldn't exist (since it's already copyrighted)?

Copyrights and patents protect entirely different subject matter. Copyright cannot cover the functional aspects of a program. Patents can only cover the functional aspects.

Why is this distinction useful?

Imagine a company owns a patent on a feature of a program, and the company sells software that implements that feature. The company may choose to license the patent separately from the software so that other companies can implement the same feature. Yet they could not then copy the copyrighted program.

Similarly, the company may choose to license the software so that other companies may create derivative works of the program. But they could not implement the patented feature in a product not derived from the licensed program.
posted by jedicus at 8:52 PM on March 13, 2012


Okay, programmer over here. You are a dangerous hack and should stick to playing around with shareware instead of pretending to be a professional.

Probably the editor did him no favors by letting the sentence that irked you so badly stand as an individual paragraph, but I think his point was meant to apply to the paragraph that preceded it:

Software is already covered by copyright, making patent protection unnecessary.

Taken together, I think he's saying that a given body of code (a "program") is distinguishable from other bodies of code written by others, even when the bodies of code do very similar things, and because of that, we shouldn't be bringing patents into the matter when simple copyright gives us the ability to express that "this thing was copied from that thing and could deprive the original creators of their due."

Maybe the whole "code is poetry" or "programmers are painters" thing gets under your skin really badly — as it sometimes does mine — but your reading isn't very generous and your response isn't very courteous.
posted by mph at 8:59 PM on March 13, 2012 [10 favorites]


Ask any programmer — developing software is as creative and unique as writing poetry.


I'm a programmer, and this programmer says no. No, not it's not.

I'm tired of programmers talking about software like it's metaphysically special. Most software is not solving unique or novel problems, which is precisely why most software patents are legal nonsense.

The vast majority of programs combine a series of very common approaches tweaked to a particular set of needs. Usually, the creative and clever part lies in defining and adapting to those circumstances, finding the precise computational model for a real world problem, or vice-versa, but that customization often happens outside the code, in business processes or marketing tactics.

Sorry, Andy.
posted by deathpanels at 9:01 PM on March 13, 2012 [4 favorites]


I'm a programmer, and this programmer says no. No, not it's not.

That still doesn't make software patents anything but evil.
posted by Mars Saxman at 9:01 PM on March 13, 2012 [1 favorite]


Against my better judgement, I'm going to follow this stupid derail and declare that I've been writing code in iambic pentameter since the late 1970s.

But then again, I am a dangerous hack, so there's that.
posted by stavrosthewonderchicken at 9:03 PM on March 13, 2012 [14 favorites]


Most software is not solving unique or novel problems, which is precisely why most software patents are legal nonsense.

That is not the standard for patentability.
posted by spacewrench at 9:03 PM on March 13, 2012


From the article:

the trouble with software patents: Purposefully vague wording invites broad interpretation.

A broad interpretation means that the prior art is more easily found that invalidates the patent. A broad patent is strong offensively but weak defensively. A narrow patent is weak offensively but strong defensively.

Against my better judgement, I sat in a conference room with my co-founders and a couple of patent attorneys and told them what we’d created. They took notes and created nonsensical documents that I still can’t make sense of. In all, I helped Yahoo file eight patent applications.

The author almost certainly signed a contract giving Yahoo! the right to patent any inventions he created for the company. If he didn't like it he should have refused to sign the contract. There are (and were) plenty of startups with no interest in applying for patents.

Software is already covered by copyright, making patent protection unnecessary.

Except that Yahoo!'s copyright on its own implementation of the invention (if it even has one) would be of no use here because there's no allegation that Facebook infringed Yahoo!'s copyright. Rather, the allegation is that Facebook infringed Yahoo!'s patent on the invention. This argument is like saying that criminal law shouldn't exist because victims can just sue in tort. The two areas of law protect different interests in different ways.

But being first with something generic that would have been invented by someone (like the wheel) — as opposed to something few could have imagined (like the Segway) — is a big difference.

The wheel is actually a remarkably non-obvious invention. No one invented it in pre-Columbian America despite thousands of years of opportunity, for example.

Ask any start-up CEO — execution is everything. As Facebook founder Jesse Eisenberg cum Mark Zuckerberg says in The Social Network, “If you guys were the inventors of Facebook, you’d have invented Facebook.”

This is really an argument for making independent invention a defense to patent infringement, not an argument against the existence of patents.
posted by jedicus at 9:03 PM on March 13, 2012


jedicus, it seems to me that you're assuming the premises when you say "imagine a company owns a patent on a feature of a program." If I don't accept that software patents are coherent, your argument doesn't work.

If one has a set of drawings and precise materials specifications allowing any person competent with the relevant fabrication skills to replicate your mousetrap, then that is patentable. Further, whatever idea or innovation you had that makes your mousetrap better than everyone else's is a non-trivial item.

Software patents essentially read "write a program to simulate a mousetrap."

Stated differently, they remind me of the innumerable clueless idiots who have "an idea" and want to find a programmer to implement their idea for "a cut of the profits." The idea is not the important part of software: writing code that actually implements the idea is the difficult part, the part that requires some protection.
posted by kavasa at 9:03 PM on March 13, 2012 [4 favorites]


You should be using the same code over and over as much as humanly possible.

Whoa. Using the same code over and over doesn't mean writing the same code over and over. Of course code that can be reused is great. But you should never write the same code over and over. If you are, you haven't found the right abstraction yet.

Writing code is creative. When you write reusable code, that's creative. When, as part of a bigger project, you reuse that code, everything aside from your reference to the existing code is creative.
posted by Jpfed at 9:06 PM on March 13, 2012 [2 favorites]


Also, that bit about Geocities: do all veterans of the Web 1.0 Wars have such talent for instant nostalgia? I made my first page on Geocities, too, but I can't blame a company for not wanting to pay the bandwidth bill for decades-old pages full of dead links and dancing baby gifs, all so that every other year we can pull up our old Final Fantasy 7 fan sites on geocities and chuckle about it.
posted by deathpanels at 9:06 PM on March 13, 2012 [1 favorite]


Except that Yahoo!'s copyright on its own implementation of the invention (if it even has one) would be of no use here because there's no allegation that Facebook infringed Yahoo!'s copyright. Rather, the allegation is that Facebook infringed Yahoo!'s patent on the invention. This argument is like saying that criminal law shouldn't exist because victims can just sue in tort. The two areas of law protect different interests in different ways.
No, this is the argument that Yahoo should not have had any grounds to sue in any court. I know I just used one but really, avoid analogies in adversarial discussion. Analogies do not prove things, they are best used to explain your point of view to someone who's having a hard time understanding what you're saying at all.

Google spent 4 billion dollars acquiring Motorola mostly for its patents, as a defense against Apple. Software patents are incoherent and economically destructive. I understand that they're your career right now, but really, they are not good. Patent trolling is a huge problem.
The wheel is actually a remarkably non-obvious invention. No one invented it in pre-Columbian America despite thousands of years of opportunity, for example.
It was invented, it's just that it was only used in toys.
posted by kavasa at 9:09 PM on March 13, 2012


That still doesn't make software patents anything but evil.

This is everything that is wrong about software people. You probably threw that "evil" in there to be cute, but I find it telling that this is part of the dialectic about software.

This is not a moral issue. It appears that Andy Baio is mad that a company acquired his company, thereby acquiring his patents, and then did things with said patents that he did not like. Framing this as an evil activity doesn't really help anyone except the author, who can hit that button any time he wants to drum up sympathy and turn the conversation into a pulp cyberpunk morality tale of the big bad corporation trampling over the noble hacker.
posted by deathpanels at 9:14 PM on March 13, 2012 [1 favorite]


Is it just me, or do more and more Metafilter members these days go out of their way to be obnoxious to other members whose work gets linked here?

It's not just you. We've had a history of snark here, but lately I've been shocked at how viciously mean some commenters have been.

As for Baio, he's idealistic to a fault, but on the other hand he's helped bring us cool stuff like Kickstarter, so there you go.
posted by buriednexttoyou at 9:19 PM on March 13, 2012


The idea is not the important part of software: writing code that actually implements the idea is the difficult part, the part that requires some protection.

That's the job of the requirements of enablement and written description: "The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same."

In theory at least, the patent is supposed to contain information equivalent to (and equivalently useful as) an actual implementation. I would argue that a major reason many software patents are of low quality is because the enablement and written description requirements are not being met.

Besides, do you really want people to be able to claim copyright protection on the functional aspects of software? At least a patent term is only 20 years and there are no statutory damages or criminal liability.

Software patents essentially read "write a program to simulate a mousetrap."

What about a patent that reads "it turns out you can use this nifty equation to program a computer to govern the operation of a rubber curing machine, and if you do you'll get much better results?" How do you define "software patent" in a way that effectively and efficiently separates the good from the bad in a way that isn't easily gamed?

Analogies do not prove things, they are best used to explain your point of view to someone who's having a hard time understanding what you're saying at all.

Well, you said it, not me.

I understand that they're your career right now, but really, they are not good.

They are not my career. My career primarily consists of academic research, of which very little has to do with software patents per se, although I do argue in favor of broad patentable subject matter. I've never drafted or litigated a software patent.

Patent trolling is a huge problem.

Patent trolling is a separate phenomenon from software patents. I don't think Yahoo! is a patent troll here.

Furthermore, most patent trolls acquire their patents through bankrupted companies, which provides a way for creditors to recover some of their losses, which in turn encourages investment in companies that don't produce a lot of tangible assets (e.g. tech startups). They are not necessarily a bad thing, and anyway it would be virtually impossible to separate them from other non-practicing entities like universities, individual inventors, and standards organizations in a legally meaningful way.

It was invented, it's just that it was only used in toys.

The author mentioned the wheel in parallel with the Segway, suggesting that he meant it in the transportation context, as did I.
posted by jedicus at 9:23 PM on March 13, 2012 [2 favorites]


i feel like the internet at large and metafilter in specific likes to take a single sentence or paragraph and argue against that just because they like to argue, even when they agree with most of the entire piece (especially when the piece as a whole doesn't support their kneejerk assumptions against it). maybe it's just more interesting or whatever, but arguing that andy baio thinks he's a unique poetic snowflake and that he thinks he should be held up as an artist because of his programming is so far out of left field. maybe you know nothing of him, but anyone that has read more than one or two things from his considerable backlog of internet publishing surely knows that isn't even close to true.
posted by nadawi at 9:25 PM on March 13, 2012 [8 favorites]


i feel like the internet at large and metafilter in specific likes to take a single sentence or paragraph and argue against that just because they like to argue, even when they agree with most of the entire piece (especially when the piece as a whole doesn't support their kneejerk assumptions against it).

I was going to take this single sentence and argue against that just because I like to argue, but it is late and I'm getting older, so let's just pretend I did.
posted by davejay at 9:29 PM on March 13, 2012 [4 favorites]


This is not a moral issue. It appears that Andy Baio is mad that a company acquired his company, thereby acquiring his patents, and then did things with said patents that he did not like.
No, he very clearly states in the article, and then reiterates in the comments that he didn't file any patents prior to the acquisition. Yahoo! asked him to file patents after and offered "large" bonuses for anyone who did, claiming that the patents would only be used defensively. He must have missed the maxim that sometimes the best defense is a good offense.
posted by b1tr0t at 9:30 PM on March 13, 2012 [2 favorites]


as a defense against Apple
Hah, I see what you did there!
posted by Threeway Handshake at 9:31 PM on March 13, 2012 [1 favorite]


Foci for Analysis: "I read Yahoo! Crosses The Line earlier today and was saddened that Yahoo had joined the ranks of patent trolls."

Perhaps you should rethink your use of the term patent troll, because you have basically made it meaningless there. I do believe they are making use of the patents they believe Facebook to be misappropriating.

Does that mean I think that software and business method patents are a societal good? No, I'd much rather they not exist. But they do, and consequently companies operate in the legal system we have, not the legal system we wish we had. I'll reserve my ire for actual patent trolls who neither invent nor implement, thanks.
posted by wierdo at 9:33 PM on March 13, 2012


I understand that they're your career right now, but really, they are not good.

A similar conflict of interest allegation could be made of any software developer that rails against software patents.

On review: my prior comment was over a page long, which is usually something of a warning sign for me, so I'm going to bow out. I doubt I was contributing much of anything new on the topic anyway.
posted by jedicus at 9:34 PM on March 13, 2012


That's the job of the requirements of enablement and written description: "The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same."
I guess my position is then best stated as: "any patent application that actually satisfies this requirement would also be a functional execution of the idea and thus, eligible for copyright."
Besides, do you really want people to be able to claim copyright protection on the functional aspects of software? At least a patent term is only 20 years and there are no statutory damages or criminal liability.
Well, copyright law (and tort more broadly) needs fixing too, but broadly yes, I think software copyright is a fundamentally sensible idea. If someone wants to steal your idea, that's fine, they just have to write their own implementation of it.
A similar conflict of interest allegation could be made of any software developer that rails against software patents.

On review: my prior comment was over a page long, which is usually something of a warning sign for me, so I'm going to bow out. I doubt I was contributing much of anything new on the topic anyway.
Luckily I'm merely a lay person! But, sad. I was enjoying the discussion, even if I'm not saying anything new either.

And yeah, the wheel thing was minor, I just wanted to mention that they were down there. =) Oh, and used for spinning textiles too, apparently.
posted by kavasa at 9:38 PM on March 13, 2012


A similar conflict of interest allegation could be made of any software developer that rails against software patents.

Are you saying that software patents, on balance, harm software developers? That would be unfortunate.
posted by Jpfed at 9:39 PM on March 13, 2012 [1 favorite]


Hi, Metafilter! I'll just drop in to say that, duh, my point was that programming is a creative endeavor and already covered by copyright, so patents aren't necessary, in the same way that novelists and poets can't patent their work. I wasn't trying to say that coding is poetry. (If you really need a metaphor, it feels more like a mix between architecture and jazz to me.)
posted by waxpancake at 9:51 PM on March 13, 2012 [17 favorites]


Also, that bit about Geocities: do all veterans of the Web 1.0 Wars have such talent for instant nostalgia? I made my first page on Geocities, too, but I can't blame a company for not wanting to pay the bandwidth bill for decades-old pages full of dead links and dancing baby gifs, all so that every other year we can pull up our old Final Fantasy 7 fan sites on geocities and chuckle about it.

Eh, I see your point, but, to speak metaphorically, dancing baby gifs are the raw material of which history is made. In most cases, our understanding of a given time and place in history is closely correlated to the volume and quality of dancing baby gifs that archaeologists have been able to dig up there, plus how reliable and detailed contemporary descriptions of that culture's dancing baby gifs are. It is amazing how quickly human societies forget stuff that was once so obvious that no-one ever even bothered to write it down. And GeoCities really did represent an important cultural stage at an early point in the Internet's history, and one day, humans will have forgotten what that stage was like.

We may not have much interest in it now as "content", as enlightening or entertaining material in and of itself, and no doubt this will remain true a hundred or a thousand years hence. But there will be historians or possibly cyberhistorians just jonesing to analyze the data, to reconstruct the spread of dancing baby gifs, to correlate it with previous work on the First Decline of television, to collaborate with colleagues studying the so-called "beta-stage memes" of popular internet culture's Silver Age (proposed book title: Did the Millennials Believe in their Lolcats?), and the result, despite the asteroidsourced jeers and nanocynicism of much of Cyberpublic 3000, will be that thoughtful humans have tools for understanding themselves just a little better than they did before.

So, yeah, although it does seem unfair to make any one corporation shoulder all the burden of maintaining this material forever, that doesn't make the material itself worthless.

tl;dr None of the pictures of my parents as children are going to win any photography awards, but I'm glad we have them.
posted by No-sword at 10:03 PM on March 13, 2012 [6 favorites]


Is there anyone here who is a currently working dev who is pro software patents? How about anyone who isn't a lawyer?
posted by markr at 10:43 PM on March 13, 2012


How about anyone who isn't a lawyer?

Why not ask a jury? Their perspectives on things might explain a bit why the public hasn't demanded meaningful patent reform.
posted by RobotVoodooPower at 10:49 PM on March 13, 2012


> This is not a moral issue

Yes, it is. Patents are wielded like a mafia protection racket.

Do you deny that those who buy patent portfolios for no reason but sue others are anything but a parasitical cancer on the industry?

I'm not saying that there aren't software inventions that would merit patent protection, but patents are currently little more than an arms race of bullshit, where whoever has the most bullshit wins. We would be better off ditching them entirely.
posted by 0xdeadc0de at 10:50 PM on March 13, 2012 [2 favorites]


Why not ask a jury? Their perspectives on things might explain a bit why the public hasn't demanded meaningful patent reform.

If those opinions were widely held wouldn't there be no reason to go jury shopping in East Texas?
posted by markr at 10:57 PM on March 13, 2012


I'm a programmer, and I'm also a writer (albeit not a professional one, yet).

I don't think poetry is quite the right analogy by any means, but I think long-form writing may be. It's a creative enterprise with a lot of flexibility and nobody there to tell you that you're doing any individual part "wrong". However, lots of people can see your final output and react by ignoring, critiquing, suggesting better features, etc. Throughout the process you need to be good with your words, and need to do a lot of research so that you don't end up writing something stupid. Sometimes you write for yourself, and sometimes you write on spec. Errors or mistaken assumptions in your work can undo your entire work's quality.

Programming tends to have some additional constraints. Your work is validated not just by a human, but also by a computer. The computer could refuse to honor your words, but any word will be printed on page or displayed on an ebook screen. Writing is also rarely truly collaborative, but with programming you often work on bigger projects.

Still the analogy holds, more often than it may seem.
posted by haykinson at 12:01 AM on March 14, 2012 [1 favorite]


I'm grateful to Andy for having written this article. The patent system is a disaster and Yahoo is a dangerously thrashing zombie. It is bad for innovation.

For another take, see Fred Wilson (a VC).
posted by Nelson at 1:06 AM on March 14, 2012


Dear Mr. Baio,

You may not remember, but during your time working for Yahoo, you received, at regular intervals, something called a "paycheck". That "paycheck" allowed you to buy nice things like, for instance, food. Yahoo could afford to give you that "paycheck" because some people, called "shareholders", poured money in it, expecting Yahoo to earn them, through your work, something called "revenue", that is, even more money. This is called "capitalism".

You acknowledge yourself that, for those patents, Yahoo not only gave you a regular "paycheck", but even some extra "incentives", something which Yahoo could hardly justify to its shareholders if it wasn't supported by an expectation of extra "revenue". As it happens, Yahoo hasn't been particularly successful turning your work into "revenue". However, it keeps trying, as well it should.

Now, if you want to work exclusively for the good of Mankind, you are free to join a non-profit. And if you want to write poetry, there's nothing to stop you either. But be aware that neither of those occupations, while perhaps more agreeable to your "hacker ethic", is generally paid as well as you were paid during your time at Yahoo. And this is for a reason.

Apart from that, as a mechanical engineer by training, I also resent the sentiment that programmers are very special snowflakes whose ethereal, poetic work, which apparently must involve rainbows and unicorns, is somehow superior to grubby mechanical or electrical engineering or chemistry, and therefore should not be fenced in by something as pedestrian as a patent.

As it happens, the people working in those other branches of technology can also be extremely creative and devise extremely elegant, beautiful solutions. I personally find Concorde rather more poetic than all your lines of code. Thing is, engineers, chemists, pharmacists, etc. have learnt to live with patents and they have hardly limited their creativity. On the contrary, they have spurred them on, both by helping reward such innovation, and by giving an incentive to find alternatives to existing patented solutions.

Technology, including software engineering, is about finding solutions to practical problems. There is nothing inferior to poetry in that.
posted by Skeptic at 1:58 AM on March 14, 2012 [2 favorites]


You may not remember, but during your time working for Yahoo, you received, at regular intervals, something called a "paycheck".

He also, apparently, received assurances that any patents based on work he did before he started receiving anything at all from yahoo would be used only defensively.

Then there's this bit:
The scary part is that even the most innocuous patent can be used to crush another’s creativity. One of the patents I co-invented is so abstract, it could not only cover Facebook’s News Feed, but virtually any activity feed. It puts into very sharp focus the trouble with software patents: Purposefully vague wording invites broad interpretation.

In their complaint, Yahoo alleges that Facebook’s News Feed violates “Dynamic page generator,” a patent filed in 1997 by their former CTO related to the launch of My Yahoo, one of the first personalized websites. Every web application, from Twitter to Pinterest, could be said to violate this patent. This is chaos.
Most of the people defending software patents seem to be defending some idealised version of the patent system where applications are checked properly for inventiveness before they're granted and it isn't easy to use spurious, legally indefensible patents to extort money from people who are actually producing useful things. Unfortunately, that isn't the system that we have.

As it happens, there most likely is an end-point for patents as we know them: the time when Chinese state-owned enterprises start using them to gouge substantial amounts of money from US and European technology companies.
posted by A Thousand Baited Hooks at 2:28 AM on March 14, 2012 [2 favorites]


As a new business development person and someone who has been involved in creating companies for profit and for the purpose of paying people, I would submit patents are a critical part of that process. Its all well and good to lament a yahoo using those patents to make money by filing lawsuits, but when you are raising capital to fund your little startup, those same patents are a critical part of the value picture that given investors confidence that they should put their money in your startup.

Without the patents, its that much harder for them to believe that you will be able to create a for-profit enterprise that offers a unique (read, not easy to copy) service that people will pay you for. Over time other elements of the business - its partnerships, brand, position in the marketplace, etc. become more important as a way of maintaining competitive advantage, but early on patents can play a critical role.
posted by zia at 2:29 AM on March 14, 2012


He also, apparently, received assurances that any patents based on work he did before he started receiving anything at all from yahoo would be used only defensively.


So he says. And?

He seems to forget that, as inventor, he must have signed something called an "assignment", giving up the entire ownership of his invention to his company, that is, allowing Yahoo to do whatever it chose to do with the patent. Under his employment contract, he probably didn't have the choice anyway, but that's my point: he worked for a for-profit corporation, drew paychecks and bonuses from them, so he is being either very naïve or very hypocritical pretending to be shocked that this for-profit company now tries to extract, well, profit from his past work.

Because that it is what this is all about: profit, money, dough, dosh, gravy, cash, loot. Yahoo isn't going after Facebook because it is creative or poetic. Yahoo is going after Facebook because Facebook is earning dollops of money, whereas Yahoo isn't and wants its cut. Yahoo certainly doesn't want to put Facebook out of business: it just wants part of its profits. All this talk about "restricting the creativity of developers" is high-faluting nonsense. Patents don't restrict their "creativity", it only restricts their ability to make money out of (rightfully or unrightfully) patented solutions. Yahoo's patents didn't "weaponize" Baios work: they just help Yahoo get more money out of it.

Now, can the patent system be improved to avoid as much as possible that patents be granted for not novel or obvious solutions in the software field? Certainly, and software developers could be the first to help by properly documenting their ideas, so that patent examiners have something with which to compare the applications. This would have the additional advantage of promoting the flow of that knowledge and helping us mortals better understand the poetry of their work.
posted by Skeptic at 3:10 AM on March 14, 2012


I detest facebook with great passion, but yahoo clearly doesn't deserve any "cut". Yahoo's customers, developers, and investors won't necessarily see anything from this spat with facebook either, only Yahoo and Facebook's lawyers really benefit, well Yahoo's execs might receive "strategy" kudos from "muppet" investors. Patents are now nothing but rent seeking by the legal profession. There is certainly no "flow of knowledge" aided by such trivial, opaque, and useless documents.
posted by jeffburdges at 4:21 AM on March 14, 2012


yahoo clearly doesn't deserve any "cut"

Why? Even Baios' letter says it is his work which is being "weaponized". Work Yahoo certainly paid him for.

Yahoo's customers, developers, and investors won't necessarily see anything from this spat with facebook either, only Yahoo and Facebook's lawyers really benefit, well Yahoo's execs might receive "strategy" kudos from "muppet" investors.

I don't see why Yahoo's customers should. However, Yahoo's current employees (developers included) may well save their jobs, and its investors get some money. See also: Motorola.

Patents are now nothing but rent seeking by the legal profession.

Both patent prosecution and litigation imply a lot of work for patent agents and attorneys. You may consider it justified or not, but it is hardly "rent seeking".
posted by Skeptic at 4:38 AM on March 14, 2012


I wonder if a reasonable compromise would be to shorten the length of software patents as compared to other patents. It seems to me that software development occurs at such a rapid pace that a twenty year patent seems rather excessive.
posted by jamincan at 5:03 AM on March 14, 2012 [1 favorite]


I’ll never file a software patent again, and I urge you to do the same.

And get reamed like you have never seen by those who have filed the patent that now covers the loving creation you have made.
posted by Old'n'Busted at 5:15 AM on March 14, 2012


I think there is a distinction to be made between software patents and patents in the fields of engineering and chemistry beyond the characterization of software developers as 'special snowflakes.' Look at just the examples from this article: there is a patent on activity feeds, a patent on page generation. You say you find the Concorde poetic - would it have existed if the concept of flight, of seats, of windows, of supersonic travel, were patented? Not implementations of those things, mind you, but the concepts themselves. Any flying device. Anything traveling faster than sound. Any device for sitting on.
posted by Nothing at 5:19 AM on March 14, 2012 [1 favorite]


How patent monopolies work in reality, outside of fairytale land
posted by jeffburdges at 5:36 AM on March 14, 2012


Look at just the examples from this article: there is a patent on activity feeds, a patent on page generation.

Without reading the patents in question, and more specifically the patent claims, you can't judge that. It is a general rule of patent law that you cannot patent a result to be achieved, only the means to achieve it, whether it is a device, a product, a process or a method. This applies in all fields, even when there are computers involved. Now, it can't be excluded that patents that don't fulfil this requirement are granted. Patent examiners are fallible humans, and invalid patents are granted. But one of the things that most infuriates me about reporting on patents in the media is that the scope of the patents is always exaggerated: witness for instance the infamous "patent on toast" cited by NPR as an example of the patent system gone wild. It actually covers a very specific industrial method for making toast at extremely high temperatures.
posted by Skeptic at 5:49 AM on March 14, 2012


It is a general rule of patent law that you cannot patent a result to be achieved, only the means to achieve it

Thank you for bringing this up. There might be a clear distinction between means and results in other fields, but the distinction between results and means is really blurry in software development. Any top-down software, for example, is a hierarchy rooted at a desired result, fulfilled by means that are also results, themselves fulfilled by other means, etc.
posted by Jpfed at 6:11 AM on March 14, 2012


How patent monopolies work in reality, outside of fairytale land

Of, let's see the factual inaccuracies (or plain lies) in that article:

To just apply for a patent monopoly in Europe, the typical cost is 50,000 euros.

Actually, not. The cost to apply is usually about one order of magnitude lower. What the article is quoting is the average total cost of filing, getting granted and maintaining a patent. A cost which can be spread over 20 years. Also, this is particularly high in Europe mainly because of translation costs into a variety of national languages. The cost is significantly lower elsewhere.

Your only recourse is to sue SinghCorp for damages in a court of law. You may win, you may lose; it’s a flip of a coin in your favor, at best. This is where you realize that a patent monopoly lawsuit costs four million dollars on average for both parties.

You'll notice that he's suddenly changed from euros to dollars. Why? Well, because litigating a patent is nowhere near as costly in Europe (between one and two orders of magnitude cheaper, in this case). It's far costlier in the US, and this is probably where he's getting this other figure (which seems quite high anyway). And in the US, unlike in Europe, if you have a good case (patent litigation is risky, but by no means a "coin toss") a law firm may accept to litigate your case on a contingency basis.

So, to make his case, Falkvinge combines the high (average!) cost of obtaining and maintaining a patent in Europe with the high (average!) cost of litigating it in the US, and further muddies the waters by confusing the cost of applying for a patent and its total cost.

One thing which he says is true: patents are generally not for "lone inventors in their basements". But, as it happens, most big innovation does not come from "lone inventors in their basements".
posted by Skeptic at 6:17 AM on March 14, 2012


(Also, it is pretty rich of Falkvinge to point at the high cost of obtaining and litigating patents, when his mate Engström, at the JURI comittee of the European Parliament, has been tabling amendment after amendment to the proposed Unitary Patent Regulation which are single-mindedly aimed at making patent litigation in the EU more protracted, expensive, and uncertain).
posted by Skeptic at 6:47 AM on March 14, 2012


Yet, Falkvinge point remains completely valid :

Patents favor the entrenched players by producing an extreme financial drag upon technological progress.

There isn't any good reason software companies should need to hire any patent lawyers, that's just money down the toilet spent protecting themselves form other rent seeking patent lawyers.

Innovation comes from wherever people are doing new things, which includes basements when discussing software. Isn't there a lovely quote by some google big shot about only being afraid of two guys in a garage?

There are obviously products which require a critical mass to bring to market of course, but patents always favor the larger or more litigious company.

I'm particularly curious about how often patent bullying gets employed during acquisition negotiations too because that clearly consolidates power.
posted by jeffburdges at 6:48 AM on March 14, 2012


It is a general rule of patent law that you cannot patent a result to be achieved, only the means to achieve it

I know of a compnay that basically exists because it leverages a patent on looking up a phone number in a table of numbers and figuring out what town it is from, so long as this happens on a phone. Would the "look it up in a table of numbers" bit be the precious means of acheivment?

Sorry dude, I know you are highly invested in patents, and it's good that you turn up when they are mentioned, but the fact remains that software patents are fucking garbage and the main result of them going away would be the world would be a better place at the expense of companies that sit on patents like that.
posted by Artw at 7:00 AM on March 14, 2012 [3 favorites]


You may not remember, but during your time working for Yahoo, you received, at regular intervals, something called a "paycheck". That "paycheck" allowed you to buy nice things like, for instance, food. Yahoo could afford to give you that "paycheck" because some people, called "shareholders", poured money in it

Very good. Nice work being a dick. Now try to show where the patents helped pay him. Because Yahoo! paid decent money to acquire the company before the patents. And they paid him during his employment before and after the patents. But they're only busting out the patents now as the last gasp of a dying firm trying to do anything to stay alive. Whether they're trying to stay alive out of respect for their shareholders or for their own self-interest is a decision I'll leave to your amazing moral compass, but this is not what the patent system is for.

Patents are to reward innovators, not to preserve dinosaurs and this is a good example of why I don't think software lends itself to patenting. Even when ideas are non-obvious (and you'd be hard-pressed to show me how these are, though they're not quite as offensive as Amazon's "One Click" bullshit), everything in coding relies on 6 zillion non-obvious ideas that came before and we've got 6 zillion problems to solve, so if you're seriously going to down tools to patent something, I think you're part of the problem, not part of the solution. The truth, to my mind, is any idea regardless of its "obviousness" has already been thought of and implemented somewhere. It just hasn't been patented because people had actual work still left to do.
posted by yerfatma at 7:07 AM on March 14, 2012


This is all reminiscent of the pre-IPO Google days, when Yahoo was trying to "weaponize" Overture. It...didn't work.
posted by obscurator at 7:12 AM on March 14, 2012


(pp 392-3)
posted by obscurator at 7:14 AM on March 14, 2012


Without reading the patents in question, and more specifically the patent claims, you can't judge that. It is a general rule of patent law that you cannot patent a result to be achieved, only the means to achieve it, whether it is a device, a product, a process or a method.

I've read the claims of a couple of the patents that were granted (they're linked from the article) and while they are couched in terms of technical processes the claims essentially cover methods of customer service performed using a computer. The claims are so broad, and deal with the technical aspects of implementation in such a general way, that there is no meaningful distinction between "result" and "means".

Or, to put it another way, if you explained to a programmer the result to be achieved by the method described in one of the claims it would probably be impossible for the programmer to achieve the same result without using the methods set out in the claim.

They aren't at all like the claims for the "bread refreshing method". To be comparable with one of the claims from the yahoo patent, the claim for the toast patent would have to be something like:
1. A method of refreshing bread products, comprising:

a) placing a bread product in a device that refreshes the bread product, and
b) using the device to refresh the bread product.

2. The method of claim 1 including the step of selecting said bread products from rolls, muffin, buns and bagels.
That's not how patents are supposed to work, but it is how software/business process patents do actually work.
posted by A Thousand Baited Hooks at 7:19 AM on March 14, 2012 [4 favorites]


To be comparable with one of the claims from the yahoo patent, the claim for the toast patent would have to be something like:

One has to bear in mind the specification, not just the claims. If the specification went into great detail describing exactly what a "device that refreshes the bread product" was and how it worked, then the claims wouldn't be so broad anymore because that term would actually be defined narrowly. While the claims are the key part of a patent, they must be read in light of the specification, prosecution history, and the understanding of a person of ordinary skill at the time of filing. Consequently, some software patents may appear much broader than they really are.

I know of a compnay that basically exists because it leverages a patent on looking up a phone number in a table of numbers and figuring out what town it is from, so long as this happens on a phone. Would the "look it up in a table of numbers" bit be the precious means of acheivment?

You'll be happy to know that the Federal Circuit and the district courts have been invalidating lots of patents lately (i.e. the past 4 years or so) for being "a known process except on the internet" or "a known process except now on a computer." Putting something on the internet or doing it via a computer is no longer inherently sufficient to avoid an obviousness challenge. I happen to agree with this change, which was too long in coming.

Automation of a process can still be nonobvious in some cases, of course. For example, it may be difficult to automate, requiring new and nonobvious algorithms. Or it could produce a nonobvious or unexpectedly useful result, perhaps via emergent behavior.
posted by jedicus at 7:46 AM on March 14, 2012 [1 favorite]


Algorithms should not be patentable because they are mathematics. There are popular general purpose programming languages like Haskell and OCaML where your program consist of nothing but equations.

"I am told that the courts are trying to make a distinction between mathematical algorithms and nonmathematical algorithms. To a computer scientist, this makes no sense, because every algorithm is as mathematical as anything could be. An algorithm is an abstract concept unrelated to physical laws of the universe." -Donald Knuth's Letter to the Patent Office

There is also the problem that all the algorithms for which software patent lawsuits are brought are basically trivial, but that's a separate issue.
posted by jeffburdges at 8:08 AM on March 14, 2012


jeffburdges Algorithms and abstract concepts are not patentable, and I'm not aware of any court trying to make a distinction between mathematical algorithms and "nonmathematical algorithms".

However, the new and non-obvious use of an algorithm for a specific purpose may well be patentable. Take the case proposed by jedicus: automation of a process by using a new and nonobvious algorithm. What would be protected by the patent would not be the algorithm as such, but its specific application to the automation of that process.

You see, a mathematical function isn't patentable either, but a new and nonobvious cam surface or wing profile following a specific mathematical function may also be patentable.
posted by Skeptic at 8:32 AM on March 14, 2012


Would the typewritten works of a thousand monkeys also be patentable?
posted by yerfatma at 8:43 AM on March 14, 2012


Patents favor the entrenched players by producing an extreme financial drag upon technological progress.

Nobody is forced to file a patent application for his inventions. If you want to prevent that others patent your own developments the remedy is simple: publish your ideas.

There isn't any good reason software companies should need to hire any patent lawyers, that's just money down the toilet spent protecting themselves form other rent seeking patent lawyers.

There can be very good reasons for software companies to get reliable advice on patents. There may even be good reasons, depending on the specifics of the case, for the company to file patent applications itself. But there can also be good reasons not to file a patent application. In the software field in particular, the reasons not to file a patent application (disclosure, difficult enforcement, cost) may quite often outweigh the reasons to file one.

Innovation comes from wherever people are doing new things, which includes basements when discussing software. Isn't there a lovely quote by some google big shot about only being afraid of two guys in a garage?

There are obviously products which require a critical mass to bring to market of course, but patents always favor the larger or more litigious company.


This is wrong. In patent litigation, large, entrenched players have much more to lose than the small guys. There's a reason why patent trolls usually go after the Googles and Microsofts of the world.
posted by Skeptic at 8:51 AM on March 14, 2012


Don't throw the baby out with the bath water. YUI is BSD licensed, so fork it and move on.

Are there any arguments for actually starting a new project with YUI (rather than, say, jQuery) in 2012? My impression (as someone who has worked on large YUI- and jQuery-based web projects) is that YUI is 2007-vintage technology which hasn't kept up.
posted by acb at 9:10 AM on March 14, 2012


Both patent prosecution and litigation imply a lot of work for patent agents and attorneys. You may consider it justified or not, but it is hardly "rent seeking".

Actually, it is classic rent seeking. What new product or service is created by all that work from the agents and attorneys? If I were to spend time and money lobbying government to pay me money, it's still rent seeking no matter how much money and time I sink into it - nothing of value is created.
posted by dragoon at 9:13 AM on March 14, 2012 [1 favorite]


I agree with jedicus that there are some applications of algorithms that can be patentable. The problem is that today's patent system fails at both ends for software. Companies are encouraged to patent everything. The patents themselves are written in a way to obscure the real invention that is being patented: software patents do not, in general, give anything of value to the state-of-the-art. At the other end, it is extremely difficult to figure out if a piece of software you wrote violates any patents.

In theory at least, the patent is supposed to contain information equivalent to (and equivalently useful as) an actual implementation. I would argue that a major reason many software patents are of low quality is because the enablement and written description requirements are not being met.

This leads me to ask: If these requirements are not met, can the patent be invalidated? Has a software patent been invalidated because recreation based on the text of the patent is too difficult?
posted by demiurge at 9:24 AM on March 14, 2012


What new product or service is created by all that work from the agents and attorneys?

Patent prosecution produces an identifiable property right around which deals can be structured without fear of Arrow's information paradox. It also produces a public disclosure of the claimed invention. If you don't think that the creation of legal rights and obligations is a valuable product or service, then you might consider how well society would function without, say, contracts or property deeds.

Patent litigation enforces that property right. Litigation doesn't need to produce a new product or service, since the purpose is to compensate for damage done. If patent litigation is rent seeking then so are other kinds of tort claims and breach of contract claims.

Now, one can argue whether a patent should be granted as a matter of law or policy. And one can argue whether the time, money, and effort that go into patent prosecution and litigation are economically efficient. But patents do often produce new wealth in the form of investments and deals that either would not otherwise exist or would be less efficient.
posted by jedicus at 9:26 AM on March 14, 2012


If these requirements are not met, can the patent be invalidated?

Most certainly. "enablement", as it is called in the US and "sufficiency of disclosure", as is called in Europe, is one of the cornerstones of patent law.

Has a software patent been invalidated because recreation based on the text of the patent is too difficult?

Quite possibly. It's a very common and often successful ground for challenging a patent in other technical fields, in particular in chemistry and biotech.
posted by Skeptic at 9:30 AM on March 14, 2012


If these requirements are not met, can the patent be invalidated?

This is true, but it's sort of tautological to say so.

Has a software patent been invalidated because recreation based on the text of the patent is too difficult?

It's hard to say how common this is, since it would typically come up at the PTO level and those records are harder to search. But there are some cases where this has happened. As I say, I think it should happen more often so as to signal to the PTO that such patents shouldn't be granted in the first place.

For example, in Ormco Corp. v. Align Technology, Inc., 2004 WL 5453218 (C.D.Cal. Aug. 20, 2004), the plaintiff owned patents claiming computer-aided design of orthodontics. In fact it was more like computer-automated design. The defendant alleged invalidity on the basis that the "patents are not enabled because neither Ormco nor anyone else has been able to create software that automatically calculates final tooth positions." The court held that the defendant proved lack of enablement by clear and convincing evidence, which is the standard in patent cases ("Align offers strong evidence Ormco never created a software system that successfully determined final tooth positions").

The Federal Circuit upheld the district court's decision in Ormco Corp. v. Align Technology, Inc., 498 F.3d 1307 (Fed. Cir. 2007). The court held that "If an inventor attempts but fails to enable his invention in a commercial product that purports to be an embodiment of the patented invention, that is strong evidence that the patent specification lacks enablement. Substantial doubt concerning the enablement of the invention was cast by the inventors in this case. The district court so concluded, and we have no reason to disagree....because clear and convincing evidence establishes that a person of ordinary skill in the art did not and could not accomplish automatic computer determination of teeth finish positions based upon the Ormco patents' specification."

So in that case the argument was that the patent was not enabling because even the defendant never managed to actually practice the claimed invention despite claiming that its product embodied the invention.
posted by jedicus at 9:44 AM on March 14, 2012


> Patents don't restrict their "creativity", it only restricts their ability to make money out of (rightfully or unrightfully) patented solutions.

That's an odd assertion to make. The rationale for the existence of patents is to promote the progress of useful arts, that is, encourage creativity by allowing it to be profitable for the creator before his imitators.

And if patents are an inducement, patents that are broad, vague and ridiculous bullshit have the opposite effect, discouraging people from entering the field for fear of litigation.

And if you say creativity will follow anyway, then why do we need patents?
posted by 0xdeadc0de at 9:46 AM on March 14, 2012


Isn't the classic patent troll modus operandi to harass small companies to build up their patent's case record, and maybe collect investors, before going after the deeper pockets? I've witnessed various small companies pay off trolls with asinine patents. I said "or more litigious" for a reason.

Any piece of software is actually the "equations" describing it, quite transparently so if it's written in Haskell. A physical surface isn't the equations describing it, but neither is it the 3d printer instructions that produce it.
posted by jeffburdges at 9:50 AM on March 14, 2012


There is quite simply no social value produced by patents or associated litigation. There aren't any technology companies who'd pack up and go home if we shut down the patent system entirely, they'd simply lay off some lawyers. I've never heard anyone who actually "innovates" find any value in reading patent applications. I'd love more disclosure myself of course, heck let's restrict software copyrights to open source software.
posted by jeffburdges at 9:55 AM on March 14, 2012


Artw: "I know of a compnay that basically exists because it leverages a patent on looking up a phone number in a table of numbers and figuring out what town it is from, so long as this happens on a phone. Would the "look it up in a table of numbers" bit be the precious means of acheivment?"

Is the rest of the world too stupid to use a tree instead? Pull the other one.
posted by wierdo at 10:02 AM on March 14, 2012


Any piece of software is actually the "equations" describing it, quite transparently so if it's written in Haskell.

This why if I think about the Navier-Stokes equations hard enough I can cause simulated water to appear on a computer screen. Or if I ponder Newtonian mechanics for a while I can conjure up a physics simulation.

Since I can't actually do that, it stands to reason that there is something different about an equation standing by itself and a computer program that uses that equation to produce a useful result. That something different is called "utility," and it's a requirement for patentability.

There aren't any technology companies who'd pack up and go home if we shut down the patent system entirely, they'd simply lay off some lawyers.

That's not true at all. Lots of drugs companies are completely dependent on patents, for example. Otherwise the investment structure for drug discovery simply doesn't work. I realize that there are other ways to fund drug discovery, and that a lot of drugs are found through those methods, but the point still stands.

The same is true, though to a lesser extent, for the chemical, petroleum, and machinery industries. See, e.g, Edwin Mansfield, Patents and Innovation: An Empirical Study, 32 Mgmt. Science 173 (1986) (reporting that significant numbers of inventions in these industries would either not have been developed or would not have been commercially produced if patent protection could not have been obtained).

And of course it's not necessary that patents be so essential that companies would "pack up and go home" if patents didn't exist. It's enough that they produce more value than they cost.

I've never heard anyone who actually "innovates" find any value in reading patent applications.

First, patents are valuable for reasons apart from disclosure. Second, the fact that something can be patented encourages other forms of disclosure as well, such as publication in journals and open discussion of how the invention works rather than treating it as a trade secret. The patent itself is just a minimal guaranteed disclosure.

But there are, in fact, researchers who read patents. See, e.g, Ralf Blossey, Read patents, not just papers, 1 Nature Materials 199 (2002).
posted by jedicus at 10:12 AM on March 14, 2012


Is the rest of the world too stupid to use a tree instead? Pull the other one.

Well, since you ask:

Customer premises telephone equipment and methods are provided for identifying a calling party's telephone number, household name, and city and/or state location, to a called party. The equipment and method includes one or more directories containing city and state location information corresponding to a complete listing of all area codes an local exchanges throughout a selected calling region or country. A receiver receives an incoming telephone number of an incoming telephone call and a comparator compares the received incoming telephone number with the directory or directories containing the city and state location information to identify a city and state location associated with the incoming telephone number. The equipment includes a display for displaying the incoming telephone call number, the household name, and/or the city and state location associated with the incoming call.
posted by Artw at 10:19 AM on March 14, 2012


Well, since you ask:

This is not legal advice, but looking at the claims it appears that the patent doesn't rely on using any particular method to look up the area code and exchange.

This is also not really an example of a "known process except on a computer." As far as I know, people were not in the habit of taking unknown phone numbers and manually figuring out where they came from based on an area code & exchange directory. I mean, people might happen to know a few area codes or exchanges that were common for them, but I don't think there were commonly used directories. It's just too much work for a small benefit, especially in the days before cheap or free long distance, which greatly increased the number of area codes and exchanges that a typical person would see.

Perhaps most importantly, doing it manually can't be done fast enough to produce a result before the caller either gives up or is directed to an answering machine. An automated system is essentially instantaneous, which means the callee can decide whether or not to answer the call based on the result. So there's a big win there.

But despite these advantages, nobody thought it up before 1997. Caller ID dates back to the late 1960s / early 1970s. Computers became small and powerful enough to perform this task by at least the early 1990s if not before, so why didn't anyone think of it in the intervening years? Even before computers were really capable of it the writing was on the wall that they would quickly become so. A person could have predicted in the late 1980s that within a few years there would be small, cheap, powerful computers that could perform this task. But they didn't. This suggests that maybe it wasn't obvious.
posted by jedicus at 10:34 AM on March 14, 2012


You are paid to be a craftsman, not a poet. You should be using the same code over and over as much as humanly possible.
This is absolutely true. It's much, much better to write a few hundred lines leveraging existing code than it is to rewrite all of it using thousands of lines of similar-but-subtlely-different code.

But it completely misses the point. That's WHY the few hundred lines you do have to write aren't going to be just bricklaying! Because, after going through all the resources out there that you had to choose from, after reusing as much as you possibly could, all the tasks that were left over for you to write from scratch are precisely those tasks where nobody else did it right yet, where you're stuck having to be creative and unique instead.

Well, except for test suites. That's invaluable, unavoidable tedious bricklaying.

Oh, and "dangerous hack"? You're aren't paid to be a talk radio host either.
posted by roystgnr at 10:36 AM on March 14, 2012


This why if I think about the Navier-Stokes equations hard enough I can cause simulated water to appear on a computer screen. Or if I ponder Newtonian mechanics for a while I can conjure up a physics simulation.

Since I can't actually do that, it stands to reason that there is something different about an equation standing by itself and a computer program that uses that equation to produce a useful result. That something different is called "utility," and it's a requirement for patentability.
I'll be damned. A quick Google search shows that straightforward formulations and algorithms for solving turbulent Navier-Stokes are now getting published as patents, not just in journals. It's a shame nobody realized that decades ago; my older colleagues would all be millionaires.
posted by roystgnr at 10:45 AM on March 14, 2012


Isn't the classic patent troll modus operandi to harass small companies to build up their patent's case record, and maybe collect investors, before going after the deeper pockets? I've witnessed various small companies pay off trolls with asinine patents. I said "or more litigious" for a reason.

I read that and noticed it. In fact, I don't think there's a single "classic troll MO". Trolls generally rely on their prey settling, rather than litigating, regardless of the quality of the patent, because litigation is so expensive, especially in the US. This is why they are much less of an issue elsewhere.

A troll may harass small companies to get "easy money", counting on them being less ready than the big guys to risk litigation. However, as a stepping stone on their path to bigger game this is risky, because if one of the small companies decides to fight back, the patent may well be invalidated, losing the opportunity to use it against more lucrative targets.

My advice to small companies harassed by a troll would be: get some professional advice and, if the patent is really rickety, prepare a somewhat decent case for its invalidity. If the troll perceives that you are ready for a fight, it will go away and search for softer targets, well before any actual litigation: after all, litigation would be just as expensive for them as for you. Its financial backers certainly won't see an advantage in raiding their funds for an assault on "small fry".

This is certainly the case in Europe, where defending yourself is easier, and courts can (and in certain countries like France often do) award damages against frivolous litigants.
posted by Skeptic at 10:59 AM on March 14, 2012


One has to bear in mind the specification, not just the claims. If the specification went into great detail describing exactly what a "device that refreshes the bread product" was and how it worked, then the claims wouldn't be so broad anymore because that term would actually be defined narrowly. While the claims are the key part of a patent, they must be read in light of the specification, prosecution history, and the understanding of a person of ordinary skill at the time of filing. Consequently, some software patents may appear much broader than they really are.

Except that the specifications for Baio's patents don't go into any detail; all they do is restate the claims in various ways that are just as vague and general as the claims themselves and give a few expressly non-limiting examples of ways the claims could be implemented. An ordinarily skilled programmer in the relevant field in 2006 would, in order to achieve the results achieved by the patented methods, have produced something covered by the claims and the specifications. Any non-obviousness is in the commercialisation of the method, not any of the technical aspects of the method.

To quote the patent for "Providing event information to third party event applications":
In the foregoing specification, embodiments of the invention have been described with reference to numerous specific details that may vary from implementation to implementation. Thus, the sole and exclusive indicator of what is in the invention, and is intended to be the invention, is the set of claims that issue from this application [...] The specification and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense.
Now, it may well be that a court would read down the claims in some way, or that the patent might not be enforceable. The problem is that it was granted at all (and in 2010, on the basis of a 2006 filing!).

Again, there seem to be two entirely separate conversations going on here: one about the way patents are ideally supposed to work (and possibly do work, in some industries) and one about the way software patents actually work in practice.
posted by A Thousand Baited Hooks at 1:55 PM on March 14, 2012 [1 favorite]


no-sword

your amusing comment led me to this fascinating google search:
did the * believe in their *
posted by idiopath at 2:28 PM on March 14, 2012


An ordinarily skilled programmer in the relevant field in 2006 would, in order to achieve the results achieved by the patented methods, have produced something covered by the claims and the specifications. Any non-obviousness is in the commercialisation of the method, not any of the technical aspects of the method.

In that case the patent should be easy to invalidate. This can be an expensive process, but that's why I'm in favor of greater use of fee-shifting in patent litigation.

Similarly, I think that the application fee for a patent should be very, very high (e.g. $50,000 or so). For smaller entities some of the fee could be refunded if the patent actually issues. The "patent everything we think we can sneak past the PTO" approach so often favored by large corporations doesn't really help anybody, as it tends to gum up both the PTO and the courts with a lot of weak patents. Higher fees would give companies an incentive to be more selective about what they patent, but it will take an order of magnitude or more increase in fees to make that happen.

Again, there seem to be two entirely separate conversations going on here: one about the way patents are ideally supposed to work (and possibly do work, in some industries) and one about the way software patents actually work in practice.

There are many instances in which software patents work just fine. People in the software world hear about problems with software patents because that's what their news sources cover, not because they're really all that much worse than other kinds (although they are probably somewhat worse). People in the biotech world hear about problematic biotech patents, people in the financial world hear about problematic financial method patents, etc.

There's also nothing inherently different about software that needs special treatment. The same reforms that would best improve the quality of software patents would improve the quality of other kinds of patents, and vice versa. Having a bunch of domain-specific rules would turn patent law into copyright law: a tangled mess of special provisions slipped in by special interests. Keeping patent law unified makes it much harder for any particular industry to control the law, as we saw recently with different industries (e.g. IT and pharma) taking opposing views with regard to many potential reforms, most of which ultimately did not become law.
posted by jedicus at 2:33 PM on March 14, 2012


There are many instances in which software patents work just fine.

Has anyone done any investigation of this, or is it just an assertion of faith?
posted by Artw at 2:35 PM on March 14, 2012


ATBH, the bit you mention is completely standard boilerplate. Despite it, in the US, the judge will still have to "construct" the claims in a Markman hearing, and as jedicus points out, his interpretation may be bound by the specification, boilerplate notwithstanding.
posted by Skeptic at 2:38 PM on March 14, 2012


Has anyone done any investigation of this, or is it just an assertion of faith?

Well it depends on what you mean by "work just fine." I take it to mean that a software patent was granted on a new, useful, and nonobvious technology and used as part of an investment or deal that resulted in a useful product or service. I don't consider a later lawsuit for infringement of that patent to invalidate the claim that the patent "worked just fine," since a right without a remedy is pretty useless.

For example, Stanford University owns a patent on the PageRank algorithm. Stanford is a non-practicing entity that signed an exclusive license agreement with Google. As far as I know, the PageRank algorithm is considered new, useful, and nonobvious, and Google's search functionality is definitely useful. The same can be said of AdWords and numerous other patented Google technologies.

Suppose those patents didn't exist? Well, it's hard to say whether Stanford would have been as keen to fund the development of the technology in the first place or whether Google's investors would have been as keen to give them money. Or whether Google would have been as successful, since its competitors could have copied some of its key technologies, which after all had been published in journals.

Another example would be MPEG-LA, which has brought together numerous software patents into patent pools for, among other things, Blu-Ray video. Again, as far as I know at least some of these patents are considered to cover new, useful, and nonobvious technologies, and the existence of the patents allowed a deal to be structured whereby various companies were able to cooperate on the creation of a standard, which has resulted in a useful product.

Suppose those patents didn't exist? The standard setting process would be much more difficult. Each company wants a share of the revenue in exchange for the technology it brings to the table. Without patents it's much easier for a company to learn the details of everyone else's technology and then create compatible video discs without paying a royalty (i.e. Arrow's information paradox). Getting around this with contracts, trade secrets, and nondisclosure agreements is messy. And once the secret is out, it's out, which is a major disadvantage of trade secrets.

So, those are a couple of examples, in my opinion, of software patents that have worked out.
posted by jedicus at 2:58 PM on March 14, 2012


jedicus: "Suppose those patents didn't exist? Well, it's hard to say whether Stanford would have been as keen to fund the development of the technology in the first place or whether Google's investors would have been as keen to give them money. Or whether Google would have been as successful, since its competitors could have copied some of its key technologies, which after all had been published in journals."

Does Stanford exist to profit, or to advance human knowledge? If it's the latter, I don't see why they'd have a problem funding the development of an obviously useful technology. As for Google's success, the vast majority of us really have no stake in that, and I don't think we'd all be worse off if there was a level playing field among search engines in terms of search algorithms. They'd just be forced to compete in other ways.
posted by mullingitover at 3:48 PM on March 14, 2012


Does Stanford exist to profit, or to advance human knowledge? If it's the latter, I don't see why they'd have a problem funding the development of an obviously useful technology.

Funding it with what money? The licensing revenue from companies that commercialize Stanford technologies is used to fund new research.

As for Google's success, the vast majority of us really have no stake in that,

That doesn't change the fact that it was very valuable for Google, its thousands of employees, its millions of shareholders, and the thousands of people who depend on Google's services for their jobs (e.g. app developers, to name one example).

and I don't think we'd all be worse off if there was a level playing field among search engines in terms of search algorithms

Do you not remember how absolutely terrible web search was before Google? Altavista, Webcrawler, Hotbot, etc? Effective web search is an enormous productivity boon for lots of workers and an enormous utility boon for everybody else who uses the web.
posted by jedicus at 4:25 PM on March 14, 2012


jedicus: "Funding it with what money? The licensing revenue from companies that commercialize Stanford technologies is used to fund new research."

The majority of academic research is still federally funded. But then the universities get to keep the patents (and subsequent revenue). Nice little arrangement. Do you really believe the universities would simply stop doing research if they didn't get free patents on the backend?

jedicus: "Do you not remember how absolutely terrible web search was before Google? Altavista, Webcrawler, Hotbot, etc? Effective web search is an enormous productivity boon for lots of workers and an enormous utility boon for everybody else who uses the web."

I wholeheartedly agree, but wouldn't we be better off if they all used PageRank and not just Google? Who's to say that the end result wouldn't have been greater than Google had accomplished with its monopoly?
posted by mullingitover at 5:26 PM on March 14, 2012 [1 favorite]


But that is just the key point. If software patents were only granted for inventions that met a strong threshold of innovation and uniqueness then there would be vastly less derision.

It's the patents for the innovation of "single click purchase", the shopping cart, or an automatic upgrade that are worse than trying to patent the idea of the wheel. And then take advantage of the burden of the legal system to allow extortionistic company to prey on actual innovators.
posted by sammyo at 6:45 PM on March 14, 2012


stavrosthewonderchicken: Against my better judgement, I'm going to follow this stupid derail and declare that I've been writing code in iambic pentameter since the late 1970s.

But then again, I am a dangerous hack, so there's that.


Surely you mean, "But then again, I am a dang'rous hack." Also: the Shakespeare programming language.
posted by whir at 12:41 PM on March 16, 2012 [1 favorite]


Patent Troll Targets Samsung and RIM With Emoticon Button Patent   lol
posted by jeffburdges at 2:10 PM on March 19, 2012


Woz fears patent wars stifling startups
posted by jeffburdges at 1:30 AM on April 10, 2012


« Older Having heard way to many similar stories, Dr. Kate...  |  The Encyclopedia Britannica is... Newer »


This thread has been archived and is closed to new comments