Software Patents Need To Change

[Thorin]

So Uniloc is suing Mojang AB for patent violation (and previously sued Microsoft and lost/won/lost/won - gotta love appeals - and has sued 73 companies altogether according to Wikipedia).

The problem I see here is that a physical product could be approximated without duplicating, thus allowing competitors to make similar products without violating the patent.  However, to check if software is a duplicate instead of an approximate, the source code has to be looked at.  Given that the case is trial by jury, I'm willing to bet not a single juror will be technical enough to compare source code to see if the techniques are copied or merely approximated.

What's real interesting is that Markus Persson (Notch) has emphatically stated he will throw as much money as needed and/or possible at fighting what many see as a patent troll to try and fight them.

Oh, and I read through the patent application and it starts off extremely vague and general, then in the details keeps talking about a "smart card".  So I'm not sure if their system really has been duplicated by Mojang, since there's no smart card with Minecraft.

It'll be interesting to see how this one ends up in three or four years.

Anyway, I'd love to see software patents get dropped to a short term (5 years? 8 years?), and I'd love to see patent lawsuits have some kind of penalty for the plaintiff if they lose, to deter frivolous lawsuits.  I also wouldn't mind seeing a requirement that a working piece of software (or at least a library) must be named as the patented invention, rather than just an idea.

[Tom]

So Uniloc is suing Mojang AB for patent violation (and previously sued Microsoft and lost/won/lost/won - gotta love appeals - and has sued 73 companies altogether according to Wikipedia).

The problem I see here is that a physical product could be approximated without duplicating, thus allowing competitors to make similar products without violating the patent.  However, to check if software is a duplicate instead of an approximate, the source code has to be looked at.  Given that the case is trial by jury, I'm willing to bet not a single juror will be technical enough to compare source code to see if the techniques are copied or merely approximated.

You might be surprised then. I had been following the Oracle vs. Google trial the entire way through via Groklaw. The jury had all kinds of code, and technical details to run through. They did in fact rule on google's side (of no infringement). The only count they sided against google was due to a somewhat misleading direction from the judge, so they never even really discussed that one, and sided for oracle, because it seemed like the judge said that google was guilty on that one.

I was impressed by that whole trial to be honest. But maybe it had something to do with the games the Oracle lawyers were playing. Same group that worked for SCO. So you can expect the hyjinx that they got up to. To begin with I thought the judge was being a bit too Oracle friendly, but it turned out that thats somewhat normal for a judge to do for the plaintiff. They give them a bit more leeway that you might expect, even with silly claims. It seems to be done to reduce the likelyhood of the plaintiff succeeding on getting a mistrial declared.

The judge was very impressive. He even learned how to code during the trial. He pretty much just laughed down Oracle's argument that the 7 line rangeCheck method was some how not "de-minimis'.

Judges are also starting to get wise to the whole pattent trolling, and software pattent messes. Several highish profile judges have started getting rather vocal about the whole thing. The only judges that haven't really gotten wise, are most of the 9th circuit (or might be the federal circuit, I can't quite remember). I say most, because recently one of the 9th circuit judges /roasted/ the rest about a recent ruling. She pretty much said "We shouldn't have done that, it is completely counter to what the SCOTUS has directed, several times.".

[Thorin]

Well, read the patent that Uniloc is suing Mojang over and you'll see that there is no source code even mentioned in it.

Now, judges learning to code and jurors understanding technical details of code is indeed impressive.  I hope it continues, but I don't expect it to.  Judges catching on to the ridiculousness of patent trolling lawsuits are good, too.

The problem with the patent under question is that it is merely an _idea_, not an actual product, that Uniloc claims Mojang is using.  Ideas should not be patentable, only verifiably unique _actual useful objects_.  Microsoft was successfully sued under a different patent, which _did_ describe an actual invention, but this patent only describes an idea.

One of the real questions that must be addressed at some point, though, is how close of an approximation should someone be allowed to make to something that's been patented?  If I invent a new electric roof shingling machine, should others be allowed to make similar electric roof shingling machines with some different component to them?  It happens all the time (look at intermittent wipers, for instance, every carmaker has those these days), but does that require a royalty payment or can you claim it's different enough that you don't have to pay?

[Tom]

Well, read the patent that Uniloc is suing Mojang over and you'll see that there is no source code even mentioned in it.

Now, judges learning to code and jurors understanding technical details of code is indeed impressive.  I hope it continues, but I don't expect it to.  Judges catching on to the ridiculousness of patent trolling lawsuits are good, too.

I completely agree.

The problem with the patent under question is that it is merely an _idea_, not an actual product, that Uniloc claims Mojang is using.  Ideas should not be patentable, only verifiably unique _actual useful objects_.  Microsoft was successfully sued under a different patent, which _did_ describe an actual invention, but this patent only describes an idea.

Yeah. Recent SCOTUS decisions have made is /much/ harder for patents that are just ideas to survive. They've made the patent rules regarding ideas much broader.

One of the real questions that must be addressed at some point, though, is how close of an approximation should someone be allowed to make to something that's been patented?  If I invent a new electric roof shingling machine, should others be allowed to make similar electric roof shingling machines with some different component to them?  It happens all the time (look at intermittent wipers, for instance, every carmaker has those these days), but does that require a royalty payment or can you claim it's different enough that you don't have to pay?

I'm hoping that the recent SCOTUS decisions have been included in the USPTO's own rules now. So generic idea software/computer patents will be rejected more often. Should reduce their load a lot I think, if they can just look at the patent application and go "Nope, that's just an idea, gtfo. revise your scope and reapply".

[Darren Dirt]

So Uniloc is suing Mojang AB for patent violation (and previously sued Microsoft and lost/won/lost/won - gotta love appeals - and has sued 73 companies altogether according to Wikipedia).

http://en.wikipedia.org/wiki/Patent_troll much?

The problem with the patent under question is that it is merely an _idea_, not an actual product, that Uniloc claims Mojang is using.  Ideas should not be patentable, only verifiably unique _actual useful objects_.  Microsoft was successfully sued under a different patent, which _did_ describe an actual invention, but this patent only describes an idea.

Wikipedia article on Uniloc:

On July 20, 2012, Uniloc filed a lawsuit against Mojang, citing the Minecraft Pocket Edition, incorrectly called "Mindcraft" within the lawsuit documents, as an infringement upon patents that give Uniloc exclusive rights to license checks on Android cellular phones.

ridic ... I mean how many ways can a developer come up with to accomplish the task of license checking and/or unlocking software, etc. and considering some of the patents were filed in the mid- to late-1990s :yeesh:

[Thorin]

Well, but read the patent first before you start saying "how many ways...".  I think that the approach invented at the time was indeed novel, and there was even a physical product created for it (CDs shipped with magazines).  But the patent is quite broad, covers a lot of ideas, and doesn't name any specific executable or library as its actual embodiment.  And that's what I don't like about that particular patent.  But there is the distinct possibility that the idea of server authentication started with them, at least in the way they claim Mojang is doing it.  If so, it's highly likely that Notch simply didn't know that process was patented (as I didn't know before reading about this lawsuit).

Still, the patent needs to be much reduced in scope, in my mind, to only software that has been produced.

[Darren Dirt]

So... looks like AMAZON wants to out-absurd Samsung, Apple, and other patent trolling high-water marks.

http://www.engadget.com/2014/05/09/amazon-studio-arrangement-patent/
http://www.wired.co.uk/news/archive/2014-05/09/amazon-photography-patent


Yes that's right -- if you read it -- and the XKCD-like drawing confirms that -- this is basically a patent covering photography in front of a clean white background.

[Mr. Analog]

I'm sure the backlog rubber stampers at the Patent Office will keep their eyes out for this kind of nonsense oh wait, they won't

Thanks Obama

THOBAMA