Microsoft, Linux, and Patents

One of my readers asked if I’d speculate on why Microsoft won’t disclose the 237 of its patents it thinks Linux violates.  Given news this week of Amdocs agreeing to license Microsoft’s Linux related patents (via a patent cross-licensing deal) I figured the timing was good.  I apologize in advance for how long this is going to be!

Microsoft invests billions of dollars a year in R&D and then discovers that others have stolen their work and are profiting from it.  Now historically for any company this is a pretty straightforward thing, you go to the “thief” (who may in fact not even realize they’ve stolen your work) and ask them to stop using it and/or pay you for the right to use it.  And if they refuse you take them to court.  As companies got bigger and more complex they realized that it’s actually rather hard not to unintentionally infringe on patents held by others and so they started doing patent cross-licensing deals.  In those deals the value of the two patent portfolios is assessed and the company with the weaker portfolio trades licenses plus cash for the licenses to the stronger portfolio.  This drives companies to build larger patent portfolios so less cash is involved in any transaction.  The larger portfolios also serve as a defensive measure since nearly anyone who has a product offering of their own, and comes after you for patent infringement, will find that they are probably infringing on one of your patents too.  So one either ignores the possible mutual infringements or takes the cross-licensing route.  In general this reduced patent infringement fights between product companies to an acceptable level.

To put this in perspective let me present a straightforward example.  With all the aggressive evolution in database systems the last thirty years, and the sometimes vicious competition amongst the database players, why is it that IBM, Oracle, and Microsoft aren’t in court constantly battling over patents?  It is inconceivable to me that all three’s database products don’t infringe on at least one patent of the other two.  I have no idea what the state of cross-licensing between these three is, but I can tell you that in earlier times when cross-licensing was uncommon that no one focused on going after the competitors for patent infringement.  Oh there could have been cases.  For example, at both DEC and then Microsoft we patented inventions that gave us a significant boost in the TPC benchmarks.  If a competitor had turned around and implemented that same technique in order to beat us in the benchmarks then we would have cried foul.  But otherwise the reality was that patent infringements rarely have a material effect on the business.  And I think that’s where todays rather litigious environment differs.

Over the last 15 years we’ve seen two forces emerge that upset the delicate “mutually assured destruction” patent peace between companies.  One is the rise in the “Patent Troll” industry.  The other is the emergence of the anti-Intellectual Property community.  Let’s deal with patent trolls first.

While Patent Troll is a pejorative thrown around a lot these days, usually by a thief trying to make the actual owner of the intellectual property look like the bad guy, the proper definition is someone who acquires patents solely for the purpose of collecting licensing revenues.  They have no product of their own, so the “mutually assured destruction” cross-licensing model doesn’t apply.  They can be ethical or unethical.  The ones that are ethical acquire stronger patents and seek reasonable (relative to the actual value of the patent to the infringer) royalties.  The unethical ones acquire and assert (often) BS patents in the hope that potential infringers will find it cheaper to pay royalties than the legal fees (and bad PR, etc.) of fighting.  Either way, Patent Trolls are widely despised.  Having personally fought one of the unethical ones, I tend to despise them as well.  Yet as an individual inventor if I patented something how would I make money off my invention?  The easiest way would probably be to give Nathan Myhrvold at Intellectual Ventures (an attempt at an ethical approach to patent trolling) a call and see if they wanted to acquire the patent.  So Patent Trolls can serve a role in encouraging innovation, even though it often seems like they are just getting in the way.

The other force at work is the anti-Intellectual Property community lead by the Free Software Foundation (FSF) and it’s GNU General Public License (GPL).  The problem here is actually two-fold.  There are those who really disdain patents and other forms of IP protection and those who support the idea of Intellectual Property for their own work but are willing to look the other way on theft of other’s property as long as it provides them sufficient economic gain.  The question here may be who is worse?  I would argue that it is the latter crowd.  I may consider the FSF and it’s followers wrong (and I’m being kind) but at least they are intellectually honest.  Those who turn a blind eye to the theft of another’s Intellectual Property while benefiting from it economically are pond scum.

A brief editorial break here.  Many of you are going to realize you are in the category I have just labeled “pond scum”.  Yes, I know that is harsh.  And that (most of) you aren’t really pond scum.  And you certainly don’t consider yourself pond scum.  But I am trying to make a point here, and it’s a point best made with a large hammer.

So we get to what I believe is the crux of Microsoft’s patent strategy with regard to Linux and it’s Android variant, go after the pond scum.  Now this is certainly more than an emotional reaction to what is happening, it is actually rather hard to figure out how to effectively go after the original authors of the infringing code.  Let’s say you take Linus Torvalds and the Linux Foundation to court for an infringement and you win.  And then what?  The code is already out in the wild.  Winning against Linus wouldn’t stop Red Hat from continuing to ship the existing code.  Let’s say you went after Red Hat and they removed the offending code.  CentOS is already out in the wild and it wouldn’t change availability of that.  In fact you’d have to go after all the Linux distributions, and then any site hosting a distribution, and even if you win it wouldn’t stop Amdocs et al from shipping servers with an offending Linux Kernel.  So you’d have to go after the Amdocs of the world.  Or the end-users  who are “receiving stolen property” and will continue to use the infringing product for their own economic gain.  In other words, you’ll have to go after the pond scum even after you’ve spent years fighting those who stole the property in the first place.  And since the pond scum actually have more to lose economically, why not simply go after them in the first place?  So that’s what Microsoft has been doing, with both Android and (more quietly until now) Linux itself.

Why go after anyone?  Well for a true Patent Troll the reason would be to make money of licensing the IP.  For Microsoft that is just a secondary (and largely immaterial) benefit.  The primary reason is that infringing on Microsoft’s IP, and then giving it away for free, is no different from if you engaged directly in software piracy by stealing copies of Windows.  There is nothing wrong with free software itself, but if you steal Microsoft’s intellectual property and give it away for free that is unfair competition.  And so Microsoft (and Apple, just to make it clear this isn’t purely a Microsoft viewpoint) is fighting back against that unfair competition by demanding compensation for use of its intellectual property.

So now we get to the crux of the question, why won’t Microsoft simply publicly disclose the 237 patents it believes Linux violates.  Well, a better question might be “why should it”?  What is the economic or strategic benefit for Microsoft to do so?  Ok, people don’t like FUD (Fear, Uncertainty, and Doubt) and lack of public disclosure seems like a FUD-based strategy.  Microsoft getting into a very public battle over which of those 237 patents Linux really infringes on won’t help Microsoft one little bit.  It will keep the press, bloggers, industry analysts, other pundits, lawyers, developers, and a gaggle of others busy for months or years as they argue over which 237 claimed infringements are real, which can be removed without causing any serious problems, etc.  But it won’t do a damn bit of good for Microsoft.  In fact, I can only see downside for them.  Not releasing the list gets them labeled for using FUD.  Releasing the list will get them labeled even more strongly for FUD.

Let’s  think about those alleged 237 patents.  How much due diligence has Microsoft done in identifying the infringements?  Obviously it did enough work to believe that Linux does infringe on those patents, but has it done all the work necessary to prove (in a court of law for example) that Linux infringes?  Almost certainly not.  What it probably knows is that if there are 237 apparent infringements that their must be anywhere from 25-100 very likely infringements.  And it probably has done the work on less than a dozen patents to have definitive infringements that they’d likely win on in a court challenge.  Why? Because that’s all they need.  Releasing the entire list would simply cause most people to focus on the perhaps 200 that either aren’t actual infringements, or that have already been removed from the code, or that could easily be removed from the code.  The press (et al) would just bury the fact that there are still dozens of real infringements.   Why not do the work to validate all 237 and just identify the real patents at issue?  Cost primarily.  I would guess it costs them in the neighborhood of $500K on average to create a definitive case for each patent.  Where is the cost/benefit analysis that shows it is worth spending $118 Million so they can be comfortable disclosing the actual list of infringing patents?  I can’t see one.

Second, how does releasing the list of 237 patents (or any list at all) help Microsoft’s primary cause?  What Microsoft wants is for a user of Linux to make the Linux vs Windows decision on some basis other than “Linux is Free”.  Releasing the list makes Linux no less free.

Third, think back to any negotiating class you may have taken (or book you may have read).  Information is power, and in any negotiation the party with more information has the advantage.  Releasing the list of patents Microsoft thinks Linux is infringing on reduces Microsoft’s negotiating power.  Why would it want to do that?

The only way to stop the FUD label is to actually seek to enforce the patent rights.  The only way to rationally do that is to go after those who are getting the most economic benefit from infringing on the patents.  With Android that was the phone manufacturers.  With Linux that is large IT shops and those who incorporate Linux into their own offerings such as Amdocs.

A secondary benefit of Microsoft’s approach is that the companies it targets are more likely to settle than fight because Android/Linux is not their primary business.  Samsung makes money selling phones and tablets, just how much money do they want to spend defending Android itself?  And Microsoft (unlike Apple) isn’t trying to stop them from shipping Android-based phones, it just wants to be compensated for its IP.  Amdocs makes CRM systems for the telecommunications industry.  It uses Linux, but Linux is not its business.  Fighting over Linux makes no sense for them (and as a long-term Amdocs shareholder I support management for avoiding this becoming a major legal distraction).  And a large IT shop?  Do they really want to spend a dime on lawyers to defend their use of Linux?  Will their CEO, or particularly their Board of Directors, really want the company distracted from their actual business (like making and selling shoes or cars or operating restaurants) over an IP issue?  No.

These guys don’t all fold because Microsoft calls up one day and demands money.  I’m sure Microsoft has to prove to them that there is actual infringement going on.  That’s where those small number of patents that Microsoft has done the full due diligence on come in.    A targeted company’s lawyers don’t have to believe Microsoft’s claim that Linux infringes on 237 patents, they just have to believe it truly infringes on ONE.  If so, they’ll advise their client to settle.  I’m sure Microsoft is taking the fact that all 237 probably aren’t provable claims into account in pricing discussions, so that it ends up feeling it is being fairly compensated and the infringing party doesn’t feel it makes more economic sense to fight.

So that’s it, Microsoft is simply following a strategy that makes large users of Android and Linux stop thinking of them as free.  And by focusing on large users it also leaves small users, enthusiasts, researchers, etc. that are more likely to be hurt by a direct attack on Linux itself completely alone.  I think it’s the right thing to do, though I don’t think it will change the market dynamics much at this point.

And in pursuing its strategy I can see no benefit to Microsoft of disclosing the 237 patents it believes Linux infringes on.  So that, quite simply, is the reason they haven’t done so.

This entry was posted in Computer and Internet, Microsoft and tagged , , , , . Bookmark the permalink.

76 Responses to Microsoft, Linux, and Patents

  1. Fallon Massey says:

    BTW, this patent activity generally occurs early on, and the dust eventually settles, and everybody goes back to whatever it was that made them money in the first place.

    Sure it’s messy while it’s going on, but the industry will settle down, and the media will cool off.

  2. J Burnett says:

    A general note on patents: Many people are not aware that patent holders are not allowed a choice as to which violators (“thieves” in your article) to approach about the patent. The patent holder must defend the patent in every case. Turning a blind-eye to a tiny company (“thief” ) may well invalidate the patent, particularly since a larger company. I found this nuance of patent law surprising.

    • halberenson says:

      While I recall hearing this decades ago I haven’t heard it mentioned in years. I wonder if the requirement was eliminated in some update to patent law?

      It is clearly a rediculous requirement. Take the case of an individual investor whose only asset is the patent. How would he afford the the legal costs of pursuing an infringer, particularly one small enough that their assets would not cover the patent holders legal costs should he win?

    • S Martin says:

      I think that you’re confusing patents and trademarks. Trademarks must be defended in all cases or they can be lost but patents are for a set period and are valid until they expire. The only exception that i am aware of is if the patent is revoked due to improper issuance in the first place. There is no requirement to defend the patent at all.

  3. I don’t think I follow the pond scum / intellectual dishonesty argument. Is it intellectually dishonest to support copyright but not patents (or perhaps just software patents)?

    • halberenson says:

      It’s intellectually dishonest to steal someone elses property and come up with all kinds of rationalizations for why its ok to do so.

      • Jorge says:

        Are you really claiming that Microsoft is so original that in a world with a population of 13 billion people the only way to that someone would think of the same solution is by stealing?
        If only a 0.0001% of the population thinks of the same solution, it means that there are 1.3 million people that think the same. Substracting Microsoft’s total head count, it will leave a generous 1.2 (and a bit more) million people in the world that can come to the same solution as Microsoft…
        … My fault, I though you had a defendable point, and I wasted my time.

      • When I was in school (70’s), it was generally understood that one could not patent a procedure or algorithm, only hardware. Then the patent office decided to allow patents to cover algorithms if they were associated with a patented piece of hardware. Then the patent examiners threw up their hands and started handing out software patents to anyone who wanted one.

        The word “steal” has a clear moral meaning when physical objects are involved, and it is not possible to steal someone’s car without knowing you have done so. The same is true of music and books, although it gets murkier. The patent office is supposed to ensure that patents are not granted on things that would be obvious to anyone skilled in the art, but they have clearly failed to do so in the case of software. The result is that I can spend 10 minutes thinking about a problem and arrive at a solution which you claim I “stole” from you. Bilge.

        “Property” exists because the laws say so. Laws can change. The idea behind patent law is that the ability to obtain patents will increase the supply of useful inventions. If the reality of the issuance and enforcement of software patents is that anyone who thinks about a problem for 10 minutes will arrive at a solution that it may be illegal for him to deploy, and he has no certain way of knowing whether it is or is not, and further there is a powerful incentive for patent holders to increase that uncertainty, then the system is not working.

        If you wish to be intellectually honest, you need to explain how the granting and enforcement of software patents leads to favorable results for society. Good luck with that.

        • halberenson says:

          The purpose of the post was not to justify the patent system

          • “The purpose of the post was not to justify the patent system”.

            The post appears to be intended to justify Microsoft’s use of the patent system to sow FUD. I, and a number of others, have pointed out that software patents are different from other patents, in ways that are uniquely pernicious. You explicitly admit this in the body of the post, when you say that the “thief” does not know he has stolen anything. But you then continue to argue from the position that someone who patents air has a moral right to prevent others from breathing. The reasonable response to that claim is that it should not be possible to patent air, and such a patent can grant a legal right, but not a moral or ethical one. Any other response is beside the point.

      • @Jorge: Your math leaves something to be desired, as do your pre-arithmetic assumptions. First of all, there are “only” about 7 billion people in the world. But even assuming your incorrect figure of 13 billion, multiplying that by 0.00001% yields 1300. Using the correct population figure yields a result of 700. All of which could be working for Microsoft. Just in Redmond alone!

      • Ashley Yakeley says:

        You are holding Microsoft and the “pond scum” to different standards. You justify Microsoft’s behaviour only by appeal to self-interest, not to any other ethics. But the exact same principle applies to those infringing its patents.

        “So now we get to the crux of the question, why won’t Microsoft simply publicly disclose the 237 patents it believes Linux violates. Well, a better question might be “why should it”?”

        Why won’t the Linux distributors stop distributing software that infringes on Microsoft’s patents? Well, a better question might be “why should they?”.

        • halberenson says:

          Simple, because they are violating the law.

          • Eddie says:

            > Simple, because they are violating the law.

            You agree the law is fundamentally broken. You understand that all businesses are in the business of making money. And then you make a statement that absolutely (in both a legal and moral sense) supports the current broken system. There is a contradiction there.

            Companies are motivated by the terrible brokenness of the software patent world to not even look for violations, due to the large increase in damages they are subject to if the infringement is “willful.” You clearly understand this, but then call at least some companies “pond scum” for doing this. Do you suggest that Microsoft is actively scanning all software patents that relate to its own code base to be sure it is not infringing any of them, either directly or via external company? And that when they do find a conflict, they either license the patent or they reimplement their code? If they are not doing this, are they “pond scum” because this means they are willfully acting in a way that pretty much guarantees they will violate patents?

            You are not holding everyone to the same standard. Your bias is very clear.

            • halberenson says:

              Microsoft isn’t demanding that others tell it which of their patents they think Microsoft is violating, that’s the difference.

              • John Kiely MCSE says:

                Microsoft isn’t demanding that others tell it which of their patents they think Microsoft is violating, that’s the difference.

                In reply to this, and adding to this, Microsoft’s corporate Lawyer’s have such a reputation and Microsoft are such a Giant I can’t imagine anyone making an accusation of Microsoft infringing on their patents…………and then refusing to name them!
                We have to bear in mind that Microsoft are the ones making the accusation!
                Would it be reasonable to believe that investigating all these Patents takes time and Money.
                Because Microsoft are the only company able to invest in the massive expense of researching each and every Patent, perhaps they feel why should they reveal what patents are being infringed after they spent Microsoft money investigating? Maybe they feel it’s now the Linux community’s turn to invest in Patent Investigation? It could be argued that revealing which Patents are in question would be like giving the money they already spent away to the competition? If this is the case the Patent system has now well and truly become unworkable and is definitely broken!
                I know there are some efforts being made currently here in Europe to place all current patents in an electronic Database and have it managed by one central body.
                I wish them luck in trying to sort this out. It’s the right idea but can it be done?

              • Eddie says:

                Oh, come on. If someone alleged that Microsoft violated their patent, you can bet that Microsoft would say, “put up or shut up.” Again, your bias is showing. If someone in the Linux camp (or Apple, or Oracle, or anyone else for that matter) alleged that Microsoft violated some number of patents, Microsoft would — reasonably — ask which ones. If the person making the allegation refused to say which ones, but then went to Microsoft’s customers to extort money out of them, Microsoft would most likely intervene — again, reasonably — to protect their product from causing such third party costs to their customers.

                Extort? Yes, based on the current brokenness of software patents, extort. If patent lawsuits did not cost such ludicrous amounts to fight, you would see a far higher fraction of companies fight the allegation rather than settle. When companies settle, one reason is because the cost of fighting the lawsuit and winning is still significantly higher than the cost of settling. And people have lost patent lawsuits to patents that should never have been awarded, so righting is really a crap shoot even when the patent is logically invalid.

                With Linux, there isn’t a single entity “in charge” that has deep pockets to intervene. With Android there is. Google has the ability to intervene. I find it disappointing that they have not.

                • halberenson says:

                  Sure Microsoft would say it, I still don’t get your point. Microsoft is “putting up”, just not in public.

                  And Microsoft does indeed step in to protect their customers. I lead one such effort. It’s not Microsoft’s fault that Linux has structured itself to make that difficult. Indeed I’m sure Microsoft is exploiting that quite intentionally. But certainly Red Hat is big enough to step in and defend its customers. So are some others in the Linux ecosystem. Why don’t they?

                  BTW, I make no claim to lacking some amount of bias. I spent a long time there and I was part of the leadership. I both criticize and praise from that perspective. Over time my attachment will fade and so will any bias. In the meantime I am fine with people thinking I’m biased.

                  • Eddie says:

                    > BTW, I make no claim to lacking some amount of bias.

                    Fair enough.

                    > I still don’t get your point. Microsoft is “putting up”, just not in public.

                    But they are *complaining* in public. They complain in public but go after it in private, not allowing settlers to disclose anything. This makes it look like a FUD campaign, whether not not it is one. Microsoft has very publicly complained about these patents that are supposedly being violated, yet when companies and individuals very reasonably ask, “Which patents?” there is no answer in public. The answer is essentially, “I’ll tell you in court but you won’t be able to tell anyone else.” This very explicitly says that Microsoft does not want the alleged violations to be removed from Linux, but instead they want to monetize the allegations.

                    Patent licensing agreements don’t tell us very much factually, only one of these is true from the viewpoint of the company paying the licensing fees: 1) the patents are sound, and are being violated, so the company agrees to pay because it’s the right thing to to, or 2) the patents are either invalid or not being violated, but 2a) the cost of fighting and winning is higher than the cost of settling, or 2b) the patent system is so broken that the risk of losing the patent battle is high, so it’s cheaper to pay the licensing than to risk losing a patent battle over an invalid patent. Thus, the fact that companies agree to pay license fees doesn’t tell us anything about the number of quality of actual violations. In the software patent arena, it is that broken.

                    > And Microsoft does indeed step in to protect their customers.

                    They have done this many times, and I admire them for it. It’s the right thing for any company to do. I agree with you that Red Hat, SuSE, Google, IBM, and some of the other companies heavily (or fully) invested in Linux logically have the ability to either step up individually to defend Linux, or do so jointly. I find it disappointing that this has not occurred in the ways I believe it should.

                    And from someone else:
                    > Just to be clear, you are calling people “pond scum” for violating a law which you
                    > refuse to justify?

                    You captured in many fewer words than I something I was trying to get at. That is what I see as a sign of bias on the part of the original author — two groups doing the same thing but THIS side is pond scum and THAT side are heroes who deserve to be compensated by the pond scum. THIS side is stealing ideas (even if the ideas are independently discovered and there is no awareness of violation), but THAT side is only doing what the broken patent system forces them to do. Wildly different standards for the same behavior.

    • Ashley Yakeley says:

      “The purpose of the post was not to justify the patent system”

      “Simple, because they are violating the law.”

      Just to be clear, you are calling people “pond scum” for violating a law which you refuse to justify?

      After all, we can stop all software patent infringement at a stroke of a pen: by disallowing software patents. Surely those who complain of the moral horror of software patent “theft” should support this?

      • Ashley Yakeley says:

        You know, the real clue that the software patent system is broken is the instruction I’ve received from my employers (Microsoft, Cisco, Amazon) never to look at patents. Why? To avoid triple damages for “wilful” infringement. In other words, all three companies are admitting that they are pond scum!

        If they weren’t pond scum, they’d have teams of folks scouring through patent filings and forcing changes on code whenever they found an infringement.

        • halberenson says:

          I had long discussions with lawyers about this, and they provided examples where even a failed search was used to justify treble damages. So their advice, and company’s usualy follow their lawyers’ advice, is to avoid the search. And the lawyers agree the system is broken.

      • halberenson says:

        I also think drug dealers are scum, even though I believe anti-drug laws are rediculous and recreational drug use should be decriminalized. And I think it’s totally nuts that an 18 year old can be trusted with a tank, but can’t buy a beer. And I wonder how we can let it be that if someone kills a USPS delivery person while they are delivering a package the punishment is the death penalty while if they kill a UPS delivery person while they are delivering a package the maximum sentence is life in prison. And I find the way mechanic’s leans by subcontractors and suppiers are handled completely unfair to the homeowner. And…. But my response to any of these is not to ignore, subvert, or advocate subverting the law, it is to support (with varying degrees of intensity) changes in the law.

        Patent law sucks. Live by it until you can change it.

  4. Hal, you make some good points on why it makes sense for Microsoft to not disclose the patents, but because of your incorrect (both legally and ethically) qualification of patent infringement as theft the article is very hard to get through. I only finished reading the article because I greatly respect your insights based on your other blog posts.

    • halberenson says:

      On the legal front you are certainly technically correct, it is a civil rather than criminal matter. On the ethical front it is more of a grey area with it probably being a reasonable term to apply when the party knowingly infringes but not otherwise. However, from the standpoint of the property owner the loss is identical and that’s why I feel free to use it to make my point.

    • halberenson says:

      One other thing, I did put thief in quotes when I introduced its use to help indicate I wasn’t meaning it literally.

  5. Rob says:

    Protecting innovative inventions, even certain advanced software algorithms, may be fine. For example, if I invent a completely novel way of recognizing objects in images that is a giant step forward from all existing methods and not just a simple twist and after years of research, then perhaps it would be OK to get a patent for that. But in general I feel that many current software patents simply just protect nonsense.
    Microsoft sued Tom-Tom with around five patents and one of them described the following ingenious invention: A GPS system which should be designed identically to any other GPS system in all other respects except that it should use an off-the-shelf X86 computer with a general-purpose OS instead of some specialized hardware with a specialized OS. How come you could get a patent for that so called idea, which basically only says “use a general-purpose computer and an OS in a car” And says nothing else such as describing any advanced procedure on how to accomplish this. How can such nonsense be patentable?
    What do you think? Most other patents are like this: “A patent which highlights URLs in e-mails on a mobile phone and allows you to activate them.” Ingenious! I want to congradulate you for your great invention. It is so much unlike what I have already seen before.
    I know that some inventions which might look obvious after the fact might not have been obvious at the time. But a patent on highlighting URLs or a patent on using an X86 computer with Linux in a car? Give me a brake.
    Inventing junk and then asking for money is like operating a protections racket.

    • halberenson says:

      This is a different topic. I am also a skeptic on software patents (and patents in general) because the USPTO (and its peers around the world) regularly allow patents that are not “non-obvious”, frequently not all that novel, and often useless (though of course if they are useless then infringement is not an issue). Governments set the parameters on what is a valid invention and the inventor’s ability to protect that invention. If a patent is “junk” then that is the government’s fault, not the inventor’s.

      Also, while you are knocking things based on their title’s it is not the title’s that matter but rather the detailed claims. And at the time the patent was granted those claims could indeed have been non-obvious, novel, and useful.

  6. Rob says:

    As I said, I agree that some things that might not look obvious today might have been obvious at the time they were designed. However, I still think that most patents simply cover junk.
    However, there is also a second more important issue that I forgot to raise. Companies which have a monopoly power should not be allowed to enforce their patents in specific cases as this could exclude competitors or prevent progress in the market. For example, Microsoft has sued all Android vendors and Tom-Tom (a GPS vendor) asking for $0.25 per unit sold for their FAT32 patents. Firstly, the FAT patents, to my mind, describe things that are obvious, trivial but most importantly non-essential when compared to the whole of the FAT specification. It is like getting a patent for two paragraphs on a specification that runs to hundreds of pages. Regardless how novel these two paragraphs could be, when you should be judged as the implementor of the specification as a whole. Secondly, Microsoft has a monopoly on desktop OSes. If I want to manufacture a USB memory stick, a camera SD card, a phone with onboard storage or in general any storage device that has to interact with a computer I simply *have* to use FAT or FAT32. 93% of users run Windows. They would simply *have* to connect my device to their PCs without a hassle as if they don’t, they would go to my competitors which might include Microsoft (e.g. Windows Phone for Android vendors). Using FAT on my devices is not steeling your invention but is simply trying to interoperate with a product that has attained a monopoly position. Not being able to interoperate with a monopoly product would introduce a lot of friction in my user experience (my users would need to install drivers etc) something which effectively forces me to follow the defacto standard (e.g. FAT) and pay money to the protections racket. Unlike standard-based organizations, there is no control on the rate I would have to pay and a monopolist could demand extremely high prices ($0.25 per unit sold) and there is no competition to help reduce prices. Finally, there is no way for me to argue that the patents in question cover only a tiny part of the implemented specification (e.g. long filenames in the case of FAT) or that e.g. FAT is a tiny part of my own product, so tiny as the amount of money to be demanded to be really high compared to the actual price of my own product, my own invention. For example, say I am manufacturing a USB memory stick and I have to pay $0.25 for the FAT patents. How much does a memory stick cost for example? 1$?). Is 25% of my product FAT related? I would answer categorically not. Again, asking for $0.25 per unit sold is like running a protections racket. There must be regulatory intervention to stop this kind of madness.
    As a second point, patents which are aggressively enforced by a monopolist could prevent progress in the marketplace. If there were more choices in the number of desktop OSes that should be supported by an electronics device, e.g. a mobile phone, then I could go around the different desktop OSes vendors and try and negotiate an agreement whereby I would use a specific vendor’s file system for example, in exchange of they promoting my new electronics device in their OS by allowing a deeper integration of my product providing for a much better user experience. Much like Apple’s products work very well with one another. When I cannot negotiate anything because a particular desktop OS has a monopoly position and so I have no leverage against them, this stifles innovation and progress.

  7. Rob says:

    I forgot to say that if a vendor is using FAT purely in their own devices without any interoperability reasons then sure they should pay Microsoft for the priviledge. It’s the interoperability angle that I am arguing for.

  8. Mike Gale says:

    This issue is badly clouded by the fact that many patents should not have been issued. It should be practical to hold the people who approve and issue the patents accountable, for a crime akin to “criminal incompetence”, if they get it wrong. One aspect of getting it wrong is whether people need to look up the patent. If they just naturally do it that way should it have been patentable in the first place?

    It should not take deep pockets to sort out the incompetents and reverse their work.

    If that were out of the way (to a greater degree), so we are talking about “justified patents”, then the points you make would carry more weight, with me.

  9. John Kiely MCSE says:

    Interesting Microsoft won’t disclose which patents are being infringed? Leaving the Linux community in turmoil. The Linux community has now no opportunity to re-engineer the patent infringements giving Microsoft an eternal stranglehold on the OS? Any future development on Linux will also have this same stranglehold.
    Considering the origin of “Windows” Microsoft really have some cheek!

  10. Tom says:

    This seems pretty inconsistent to me. Microsoft is (apparently) suffering damage because others have ‘stolen’ (by which you usually mean ‘invented independently’) its patented technologies. If that were so, then wouldn’t they be better off if the patented technologies were removed from Linux? Wouldn’t that restore their competitive advantage? But no-one is able to do that, because Microsoft won’t disclose the list of patents allegedly infringed.

    If the dispute is really about the billions of dollars of R&D investment, the ‘theft’ of the valuable technologies that came out of it and the resulting unfair competition, then Microsoft should prefer a situation where those technologies are removed from the competitors’ products, And the only way that can even begin to happen is for Microsoft to disclose what the infringing technologies are.

    That Microsoft refuses to make that disclosure leads to the (fairly inevitable) conclusion that Microsoft prefers the technologies to remain in Linux. Now, why would that be, I wonder? Could it, perhaps, be because its wonderful technology is in fact more-or-less worthless, conferring no competitive advantage? In that case, it’s much better off leaving the patented technology in Linux, and in fact fighting tooth-and-nail to keep it there, because then it can extract license fees from anyone using it. God forbid there should be a free OS that _doesn’t_ infringe Microsoft’s patents.

    Instead, the strategy is to go around to end-users of Linux, pressuring them into licensing agreements for patents that they had no reasonable way of knowing they were infringing (because Microsoft refuses to release that list) and then insisting on NDAs attached to the deal that prevent them from making any contribution back to Linux that removes the infringement (because that would reveal what the infringement is). Anyone with the resources to fight this is left alone, because if the case actually made it to court then the patents at issue would be out in the open and the problem could be dealt with.

    Pond scum, indeed.

    • halberenson says:

      The Linux community is welcome to go through Microsoft’s patents and figure out which ones Linux (in its infinite variations) violates and then work around them.

      • Tom says:

        Rather missing the point. If the harm to Microsoft is unfair competition through use of patented technologies, why don’t they want those technologies removed from Linux?

        I’d love to see someone try a defence of equitable laches. But, of course, Microsoft wouldn’t want to risk that by going after someone who actually has a chance of building a defence. I mean, why haven’t they sued IBM yet? Or Oracle, say? Hmmm? They sure practice those patents in a much bigger way than Amdocs and make much more money out of it.

      • Mikernet says:

        Tom definitely has a point here. I read your whole article and understood what you were getting at, but Tom’s point is much stronger. Remember, Microsoft makes $15 off of each Android device sold. That’s almost as much as it makes off of a Windows Phone license, which is absurd, just work out how much money that is. Android makes WAY much more money for Microsoft than Windows Phone. Imagine if Android actually didn’t have any infringing code, or removed lots of infringing code to make the $15 unjustifiable.

      • Mikernet says:

        My number might be a bit off there, just re-googled the issue and it looks like current estimates are $5 – $15 depending on the manufacturer and device. Not that that really changes my point, but just thought it would be prudent to correct my figures.

  11. shebaw says:

    It’s Torvalds, not Torvald. And I agree with Rob.

  12. Anonymous Coward says:

    IMO, the big software houses, by patenting lots of BS (I’d estimate that of all SW patents in the US maybe 0.1% are non-BS), have deliberately devaluated the notion of a software patent. Which, IMO, is enough of a technical justification not to pay for _any_ software patent. There’s IMO really no other defense for the little guy against the software giants except a law stating that you can’t violate patents by writing software.

    Think of it from another point of view. Patents aren’t something natural. They were introduced as an artificial, completely novel construct in the legislation of Western countries a long time ago, to further technological development – somebody spending time and money on developing something innovative should have some reward for his effort guaranteed. The way software patents are used nowadays completely defeats this purpose. OTOH, in today’s economy, it’s innovate or die. There’s no economical benefit from software patents. Why do we keep a solution around to a problem we no longer have?

    Put yet another way, software patents, in their current incarnation, are a form of legalized monopoly. In effect, the state grants one individual the right to do something nobody else is allowed to do for a long period of time. For what? Monopolies should be illegal, shouldn’t they? It is one thing to build machinery which is better than that of the competition. Building machinery is in itself expensive, operating it is expensive too, and building it cheaper, or making it easier to operate usually means lower costs for the customer. Whereas multiplication of software is free, and having to pay higher royalties for software due to tons of BS patents does not benefit the customer at all.

    Yet another point of view is that SW patents are equivalent to business methods. A software patent describes a way of doing something – which is exactly what a business method is. If business methods are not patentable, how come software patents are allowed?

    • halberenson says:

      I have not argued (in this blog post) the merits of the current system for (Software) Patents, which I tend to actually find significantly flawed. I’ll I’m addressing is Microsoft’s (and others’) behavior within the current system. And yes, Patents are a form of Monopoly. Governments specifically allow many such explicit monopoly systems (e.g., charters for cable companies, utilities, etc.). And no, Monopolies aren’t and should not be illegal. Certain behaviors of Monopolies are illegal, though one can debate the merits of that approach too. The counter-argument to anti-trust laws is that monopolies are unsustaintable because someone eventually figures out an end-run around them. In essence monopolies foster innovation, though not necessarily rapid innovation 🙂

  13. Mark Hayden says:

    Interesting read, but there is a major mistake or misunderstandings in your article that hurts your argument–that RMS and his creations the Free Software Foundation and the GPL are “anti Intellectual Property”. This is factually incorrect. I don’t believe you deliberately lied though; it is a common mistake many people make. Stallman and the FSF are ardent defenders of copyright law as it pertains to software. Indeed without copyright law the FSF would not be able to fully suport and enforce “copyleft” (the concept applied through the terms of the GNU public license). Thus, to say they are anti-IP is a falsehood. What they ARE against are SOFTWARE PATENTS because software is already protected by copyright and because patents protect mere ideas, not the implementation (the code itself). As such, the over-application of patent law unduly stifles innovation (in their opinion–you are free to disagree there as that is the point of debate here).

    In the interests of a healthy debate here let me offer up an alternative opinion…

    The framers of the patent system we have today intended on creating a system of granting what is essentially legislated monopolies for a limited term in order to provide inventors an incentive to innovate. Those framers worked under two important assumptions:

    1) Inventions deserving such a monopoly are “non-obvious” and “novel” in terms of not only the idea itself, but in the amount of work it would take to implement the invention. The original intention of patent law was to offer protection to inventions of a physical nature–that is, a PHYSICAL device or transformation. Software is ephemeral–a computer program is at its heart a complex algorithm. It was only very recently that patent law has been perverted to the point that software can be patented, through the dubious argument that a general purpose computing device becomes a special or “novel” device in combination with software. This is a case of interpreting the letter of the law (dubiously even) in direct opposition to the original spirit of the law (since software patents are almost universally used to stifle innovation rather than foster it).

    2) Inventors pursuing patents for their ideas would be putting forth sincere efforts to implement their ideas and bring them to market. After all, the whole point of fostering innovation is to benefit mankind, and there is no benefit to mankind in an innovation that nobody can enjoy. The rise of NPEs (non-practicing entities, the proper name for patent trolls) was obviously not anticipated way back then. Because of the first assumption above, it was inconceivable that an inventor would put so much effort into an invention just to sit on it in the hopes someone else would miraculously and accidentally start building his invention.

    All the major players in computing violate both of these assumptions. Software-based patents are NEVER novel enough to warrant patent protection, no matter how clever they are, because they are ALGORITHMS–which were not supposed to be patentable! I know that many software-based ideas take a great deal of effort to implement, but the IMPLEMENTATION of such ideas is already VERY well covered by copyright!

    As for “intellectual honesty” I have a couple of opinions here. Firstly there are is no such thing as a “good NPE” in my opinion. You consider Nathan Myhrvold to be a “good witch” in that realm, but you never say WHY he is. What makes an NPE a good NPE vs. a bad one? Maybe he is a nice guy and very smart, but that doesn’t was with me. NPEs are companies that make 100 percent of their income through the use of legislation. They do not take ideas in patents and find people or firms to make them real and put them out into the world (i.e. the “practice” part), rather they skulk about and find people or firms that ALREADY have done that hard work and shake them down for money. NOT COOL and not honest at heart. It is also a reason I thing that as a condition of being granted a patent the applicant must include an “implementation strategy” in the filing–a plan on how they will practice their invention even if it means merely licensing or finding a buyer for the patent. That implementation plan would be limited to a relatively short term (5 years perhaps) and re-sold patents wouldn’t be granted extension without a full re-application. That would effectively eliminate NPEs which shouldn’t have been allowed to exist in the first place.

    Secondly, it is ALWAYS intellectually dishonest to negotiate patent licensing deals that do not permit the licensee to disclose which patents they have licensed. Disclosure is a core component of the patent system–if something is so special then keep it a trade secret the way KFC, Coke and Heinz keep their recipes secret. MSFT is committing intellectual dishonesty here in a serious way. When they or anyone else make public assertions that their patents are being violated, then keep all detail secret and further go on to extort money from licensees who are also not permitted to know and fully disclose what they’ve licensed is not only morally dubious, it also suggests they believe the patents in question are potentially worthless. Basically, MSFT is worried that the risk of those patents being invalidated on legal challenge is to high, or that the ideas described in them are trivial enough that they can be removed or circumvented with little pain the way BSD devs were able to resolve their legal challenges through code refactoring.

    It is bad enough when “good” companies like Google professes to be get dragged into this steaming pile of legal crap. It is doubly offensive when Microsoft, with its history of talking liberties with others’ ideas and suppressing innovation, presents itself as a defender of innovation.

    • Tom says:

      Indeed. The argument offered in the article for NPEs is wholly spurious, too. The idea that patent trolls somehow encourage innovation is ludicrous. If the little guy invents something big and patents it, he has two other, much more sensible options for monetising it: (1) license it out just like anybody else, or (2) sell it to someone who wants to practice it.

      The article tries to suggest that the little guy makes his big invention but is lost amidst the swirling sea of patent law… UNTIL the white knight races up on his charger to save the day! If you can’t see what’s wrong making a white knight out of patent trolls, go ask Desire2Learn, Thompson and Morgan, Multimedia Live, Brown Shoe Co, Travelcenters of America, Gamestop, Rovio, Illusion Labs, Richard Shinerman, Wulven Game Studios, Combay… honestly, the list goes on!

  14. brianbalke says:

    It is certainly distressing that a collaboration constituted in such a way as to prevent the acquisition of patents should be subject to tactics normally applied to a commercial enterprise.
    I wonder how many features Microsoft has adopted from the Linux distribution? Has it done due diligence to assess their worth?
    And has Microsoft shut down its own Linux servers? Or – as the holder of patents – is it the only entity allowed to operate them?

    • halberenson says:

      Many people sue Microsoft for patent infringement, and many win. It isn’t that Microsoft is trying to infringe, it is that most infringements are unintentional. Either because the infringer truly didn’t know that the patent existed, or when they read the (notoriously obscurely written) patent they did ‘t believe what they did infringed on it. This is why we have patent cross-licensing agreements.

      Microsoft goes out of its way to make sure no GPL’d code is looked at or incorporated into its products. But that doesn’t mean some idea that made it into Linux, but was not patented, didn’t also make it into Windows. That’s the price the Linux community pays for not patenting things.

      • Eddie says:

        Don’t forget about when Microsoft started to patent ideas it incorporated from the BlueJ IDE. Fortunately, they did the right thing and withdrew their patent application when it was discovered.

      • Bob C. says:

        Pure BS. First of all, many people can’t afford to hire the bank of lawyers required to sue Microsoft for patent infringement on the one or two patents. Secondly, when I worked for a major corporation who owns a significant number of patents, I was involved in one project where Microsoft claimed patent infringement, only to find they had patented algorithms that came from 1970s Unix code (which had not been patented). I suspect this might be in that same set of 237 patents they refuse to identify – after all, Linux is based on Unix, which was out long before S/W was ever allowed to be patented or copyrighted. In essence, all the early Unix code is considered public domain and not patentable. To say that a piece of code using pointers to locate memory blocks infringes upon their patents of a larger larger implementation because that implementation uses pointers is sheer nonsense and a lot of gall.

        Also, you made earlier reference that the USPTO should be more careful when reviewing patents, while true in concept fails in practice unless they are given the budget to hire another 15,000 agents ala the IRS. If companies like IBM, Oracle, H-P, etc. have no problem revealing their patents which they thing have been infringed upon, why should Microsoft feel otherwise – UNLESS they are non-patentable BS, which is most likely OR they allow Microsoft to extort money from individuals and small “mom and pop” shops that do not have the deep pockets to fight back in court. It isn’t often you have two “giants” fighting it out in court over S/W patents – just cross-license to avoid the “Oracle vs Google” case (an Oracle “spite” move?)

        • halberenson says:

          Does Google do cross-licensing? Do they have much to cross-license outside of search (which they won’t do of course)? Since cross-licensing deals are patents plus cash it might be that Google would have to part with a lot of cash; as much as they wanted in the lawsuit perhaps.

          Microsoft is subject to the same problems as everyone else in the industry. Someone thinks they have something novel and starts the process of writing a patent app. Outside parties review it and determine it is valid. They submit it to the USPTO and the examiner agrees. Meanwhile there is indeed prior-art from the pre-patent days (or just that no one ever patented) but also no one in the chain realizes exists. Most of the people writing the code weren’t even born when Unix (or TOPS-10, or MVS, or…) were created. Heck, some of them weren’t born when Windows itself was created!

  15. Helio says:

    tl;dr: MS won’t disclose their list of Linux patent infringements because that would leave them with less of a stick with which to hit back at the competition.

    Or to put it another way: if MS is confident it has a case for twelve or so patents, and not so sure about the rest, why else go with the higher number – the one that likely wouldn’t stand close scrutiny?

  16. MICHAEL J. REEVES, AA, ASc says:

    This reminds me of Apple’s lawsuit against Microsoft Windows GUI – Look and Feel. As it turned out, Xerox did the original research, development, and proof-of-concept, BUT didn’t think the results were worth a patent. And the results of that suit is public history.
    All of the Operating Systems were first developed at universities using public funds (government funds – ie: taxpayer money) which means that much of the results of the research and development (R&D) actually belongs to the taxpayers and resides in the public domain. All the current and past operating systems were developed as offshoots of this original government funded R&D to meet new technological developments such as mini-computers and then microcomputers. Additionally, new key concepts and features evolved as new technology became available.
    One might say that everything allegedly new is built on the shoulders of everything in the past. In the software industry, this is a completely relevant perspective. Unfortunately, separating what is a new and patentable concept is not so clear cut. The biggest issue is the available tools used to implement these concepts. We all have limited sets of tools to program and implement our ideas. This limits the ability of our software and hardware engineers to a finite set of variations in implementing the end results.
    This is never more apparent than in the development of programming languages when updated standards are approved because new features or fixes (deprecation of older features that are still supported for maintenance of legacy software) are added to make software development easier and more secure against user error or deliberate attacks (Murphy’s Law of Users: Any button/s a user can push that will crash a system will be found and pushed!), or just poor programming technique (programmer’s implementation errors).
    The facts are that Apple OS, Microsoft DOS & Windows, and the Linux Kernel and Open Source software all owe their existence to UNIX, MULTICS, and any other operating system that predates them or the addition of specific features.
    Features added are often the result of a user/programmer saying “I need to do this, and how can I get this done.” And then another user says, “Yeah, I need to be able to do that, too.” Eventually, enough users have said, “I need that, too”, and it goes viral in the current terminology.

    Now don’t misunderstand me here. I firmly believe that if a programmer spends 100 hours or a 1,000 hours, or whatever to develop and implement a product, they should be paid for their efforts and time. If inventors and developers don’t get paid, there can be no progress. I don’t know anyone that can afford to work for free. People have to barter for food and other items necessary for survival. As the paraphrased saying goes, “People migrate to where the money is…” This is fully acknowledged by the Patent laws that seek to protect the work of individuals or companies who are the creative force in improving the Standard of Living and Quality of Life for societies.

    Unfortunately, in today’s environment, it is often difficult to separate fact from fiction in assigning credit for developing a feature or idea. We have Wikipedia where anybody can post and take credit for the work and efforts of others without acknowledging those shoulders they stood upon to see the next stage of development. This is equally troubling when inventors and developers demonstrate their work, but haven’t submitted any documentation for copyright or filed for patent, and find their work has been pilfered by a company or individual that sees value in the product.

    In the recent past, programmers often worked for a company, and after work, they would develop special projects, and when polished enough for distribution, they would self-publish as freeware or shareware. I was and am a supporter this concept. I have supported many an author over the years if I found their product useful. My children benefited greatly from many of the games developed by independent programmers and published by Apogee Software or other software publishing houses.

    This is the historical basis of the Free Software Foundation and social movement. But what isn’t acknowledged by the Foundation and movement is that someone was footing the bill either directly or indirectly.

    Red Hat Software and similar companies have learned how to profit and survive by supplying subscription support for free software distributions. Additionally, they hire and/or support programmers to review and develop code for new features or to fix bugs in existing code. This is a good thing, imho.

    One other significant point about patents and companies that own them: oftentimes companies buy up patents to stifle progress because of a conflict of interest. This has been documented fully in the automotive industry where the patent would, if bought by a competitor, force the company to re-engineer a product at a significant cost of R&D. But by buying up the patent, re-engineering can be delayed until such time as the invention becomes necessary to compete in the open market.

    This last issue is currently being reviewed and addressed by possibly putting a time limit on the implementation of a patent, and if not utilized within this time frame, the patent is declared null-and-void, and enters the public domain and free for anyone to use.

    The only reason that I can see for Microsoft to not divulge the 237 alleged patent infringements is that under the lens of a microscope, it might turn out that there exist prior documentation that would demonstrate that even though Microsoft filed and received the patent, the original work was not done or purchased by Microsoft.

    Hence, it is to Microsoft’s advantage keep silent to avoid public disclosure, and avoid the possibility of the originator suing and providing evidence Microsoft’s unethical and possibly illegal business practices. Hmmmm. This sounds familiar? Am I having a “Deja Vu” moment here?

    In My Humble Opinion,



    • halberenson says:

      Most patents are for small specializations, not for core concepts. And recall that Software Patents did not exist prior to the 1980s and that the case law around them was not solid until the 90s. As a result much foundational work could not be patented.

  17. Eddie says:

    You lost me when you called the FSF not only “Anti-Intellectual Property” but leaders in that area. What is the GPL, but the embodiment of one form of intellectual property? The very term “Intellectual Property” is itself a misleading and confusing term, as it lumps together as if similar the starkly different ideas of copyright, patent, trademark, and other forms of intellectual property. You are not talking about the (vague) notion of IP here. You’re talking specifically about not patents, but software patents.

    If you want to say the FSF (or the GPL v3) is anti-software-patent, that is different. But to say they are anti-IP is ridiculous and you lose your argument right there. I am not the biggest fan of the FSF and I find Richard Stallman arrogant and grating. But you went too far in your argument. Most arguments I come across that conflate all forms of IP as if they are one end up going too extreme.

    The quality of software patents is so terrible, so many of them should never have been granted, that this seriously weakens your argument about “thieves,” which as someone else pointed out, in the software world, are practically always independent inventors of the same, often relatively obvious invention. The intent of the patent system was to give the inventor incentive to make their invention public knowledge, for which they got a legal monopoly on that invention. That is, things they would otherwise have the incentive to keep a trade secret, they would make public instead. If you look at software patents, however, the clear purpose of almost all software patents is numbers or money. Companies want large numbers of (usually weak) patents so they don’t have the lower number of patents when cross-licensing. Very, very little of what is patented in software is worthy of being a trade secret.

    Patent suits are ridiculously expensive to fight. This is one reason that so many companies will settle. Not because it’s the right decision for other reasons, but because it’s significantly less expensive to settle than to fight and win.

    If Microsoft truly wanted their (quote) innovations (unquote) to be removed from Linux and other competing products, they would reveal the list. That they refuse to do so, to me, indicates that they don’t really want those patents to stop being infringed. They want to monetize. To me, that is just a different flavor of patent troll. After losing the (stupid) EOLAS lawsuit, you’d think Microsoft would avoid playing that game, but like most other large companies, they only wish to avoid these lawsuits when it’s not in their favor. They are happy to abuse the patent system when it is to their favor.

    • halberenson says:

      You’re focusing on a pretty minor point in the whole thing. I was trying to be nice in my comment about FSF by acknowledging they had a valid intellectual position, not trying to dive into it.

      • Eddie says:

        If you think I focused on a minor point, then you didn’t really read my whole post, but only the first paragraph or two. I discussed several different things. Also, you suggest the FSF is anti-patent, which is silly. They are anti-software-patent (at least I think they are against all software patents), which is different. Conflating the two notions is not rigorously honest. You yourself acknowledge (in responses) that there are problems with software patents.

        The government needs to do one or more of: Drastically shorten the lifetime of software patents, raise the bar for non-obviousness to cut out most of the junk, rule out software patents altogether (after all, “algorithms” were not supposed to be patentable), make it vastly cheaper and easier to invalidate a software patent, and/or other things to make this market less damaged. Software and business process patents are used far more to prevent competition or monetize ideas than to protect real innovation.

        • halberenson says:

          I have repeatedly posted that I agree that patent reform is needed. But that also wasn’t the point of my post. People must operate as the system exists today, not some imaginary (if desirable) future.

  18. Bob - former Decie says:

    Thanks for writing this post. I was hoping you might have a possible, even if only theoretical, reason why Microsoft is refusing to freely and openly identify the patents they believe Linux violates, other than they can make more money by not disclosing the patents. Unfortunately I didn’t read any such reason.
    As far as suggesting that Linux developers read every single Microsoft patent to try and discover which ones they might be violating, you know better than many people that many companies forbid their employees from researching patents because that would open them up to the treble damages provision of patent law if they were found to be in violation of any patent.

    • halberenson says:

      Sure, it’s purely money plus strategy.

      You’d have to hire outside people to do it, just as Microsoft must have, in order to avoid the pollution/claims of willfull infringement.

  19. Khaja says:

    Excellent post Hal! Your analysis of what Microsoft’s reasoning might be is cogent.

    That said, as you have noted and others have pointed out, the patent system is badly broken. Others have hinted at this but I want to call out clearly that, entirely too often, it puts people and organizations in the position of being legally wrong and morally right or vice versa. This makes it very hard to broadly judge violators of the law as being morally wrong without knowing and considering the specifics of each case.

    Microsoft itself has been found to be in violation of the law in some suits brought against it in the past. They have also been seen as unethical in their conduct in some of these suits. As you point out, it is not because anyone sets out to be unethical or a ‘thief’. It is just the nature of the game. As a big player with too much going on they have to focus on the legal defensiveness of their position. Moral / ethical considerations often take a back seat – with predictable consequences.

    IOW, there are robbers and victims on all sides of this patent debate.

    • halberenson says:

      Thanks Khaja. Oh, and I forgot I promised to do something…will solve that as soon as I dig out of a hole I’m in this afternoon.

      I should just write another post about how broken the system is.

  20. ChadF says:

    Just a thought.. Could Microsoft be charged for slander for not revealing them? I mean if some person/organization made a public statement like “John Doe here raped 237 women, but I’m not saying who and refuse to show any evidence of that accusation”, wouldn’t they be criminally liable? Not that I’m equating patent infringement to the level of actual rape, but in the software world it might make people just as avoiding of out of fear (which could [hypothetically] be M$’s goal, being a huge PR machine).

    Another thought (which relates to those that mentioned about patents must be equally enforced).. If someone takes “something” (whatever parts of linux that could be patentable/or could have been if the original author had chose to be patent it instead) from linux and implement a product that threatens Microsoft (causing M$ to sue them), would that person/organization be able to claim that because M$ didn’t point out the [claimed] patent violations in linux before hand that anyone using it because it was derived from linux that M$ can’t claim [past] damages? This would be different than the whole GIF compression thing where at least the patent involved was publicly known, and they just didn’t choose to enforce it at first. Sounds like a good justification on why patents _should_ be required to be enforced equally (at least in _specific_ notification of such violations, even forgoing legal action), if they are not so.

    And on the assumptions of legitimate infringement(s), maybe Microsoft is scared to make SCO’s mistake.

  21. honestjoe says:

    I like the concept that prevents someone stealing my unique process, my tool, but to me this is copyright.

    IPR is (in my eyes) the right to impend someone else from using a set pattern of copyrighted or non-copyrighted processes or devices to achieve some desired output of end state. It is not a tool per say, because the tool is the implementation. It is the pattern that specifies the tool and its objective.

    …I also want to use other words to express the conceptual difference in Micro and Macro level.
    for me the Macro process is the IPRable part. It represents a composite of parts to create a higher order function that is not be obvious to the skilled manufacturer (developer). [ a key point is invention through insight and expansive cross domain knowledge]
    and the Micro process is the non IPRable part, because its development is driven by common education and reasoning by a skilled manufacturer (developer). [ a key point is prior structured/academic training and limited cross domain knowledge ]

    Now in respect of software, the tool is the binary. The binary is manifest, and thus is physical all be it in a purely electrical representation. The code that instructs the compiler is the Design. (like lines on a blueprint) and the compiler process is the manufacturing process, the common forge.

    I thus conclude that all IPR violations, in respect of low level software are bollocks. Higher order software process are another matter but should be distinguished using the Macro and Micro rules above to test for “obvious to a skilled person in the trade”

    Code is an issue of both design protection and copyright protection. (1st party/3rd party)
    Compilers are specialist processors that convert raw designs into instances and thus are IPRable. (4th party)
    Binary objects are (IMO) only subject to copyright. (1st party)
    Executing Binaries (i.e. a running system) are a business process and thus subject to IPR in whole, but not in part. (2nd Party)
    Macro level software tasks are Business processes and thus imply a higher order of understanding and design thus are IPRable in isolation. (1st party design)
    Micro level software tasks are functional processes developed by “laboures” and thus are not IPRable, because they are of common function. (i.e. by 3rd party design)

    1st party = Microsoft
    2nd party = Microsoft client
    3rd party = Joe developer.
    4th party = Compiler developer

  22. honestjoe says:

    So I would think that the Micorsofts and Apples, being partes 1,2,3 and 4 in most cases can assert rights of design, copy and or IPR at multiple levels on pretty much everyone in the world that uses software. Unless they lived in a cave for the last 40 years and had their own silicon fabrication plant and Denis Ritchie visits in dreams to explain c and assembly. … but even that’s infringement.

  23. Interesting article. Just remember that Microsoft could also be in the ‘pond scum’ category. Remember the mouse patent when it first was released?

  24. Franz Lichal says:

    This reminds me of another company that claimed IP patent violations in the Linux kernel. Just the same, it refused to tell about which code sequences caused the violatins. See Wikipedia’s “Timeline of SCO–Linux controversies” on how US court of justice thinks about those practicies. I’m not afraid that Microsoft will be a Chapter 11 candidate when going on that way, but it will not succeed either. I would appreciate if one company would stand a trial and close this dark chapter of computer industry. I think patent laws should be rewritten as a whole, they simple do not reflect todays reality.

  25. Pierre says:

    Interesting article but I am trying to understand some basics of patents. To be granted a patent are you not required to disclose the details of the invention to the patent office? How else can you file for a patent? If they were never disclosed how can jurisprudence be upheld? So Microsoft has 237 patents, they must be documented somewhere!

    • halberenson says:

      Microsoft has many thousands of patents and of course they are all publicly disclosed. What they haven’t done is say which 237 of those thousands they believe Linux infringes upon. The Linux community (which includes those with very deep pockets like IBM and Intel) could very well fund an effort to review Microsoft’s patent portfolio and figure out which ones it may be infringing on. But why do that when you can get better press by attacking Microsoft?

      • John Kiely MCSE says:

        I love the way you turned that around LOL!In defence of Microsoft they do employ a lot of people and are a very good company to work for, I believe. However they do have a litigious reputation. They can become a very overbearing presence in the courtroom and can sometimes come across as an uncaring and almost hostile MultiNational Corporation. They have in actual fact got very good Corporate Lawyers!
        I personally think that they are more about “the Money” rather than pushing the industry and technology forward. I like Linux, but it has it’s faults as well!
        Now apart from the gargantuan effort required and the money needed, what is stopping some new company coming out with an Operating System which is Multi / Cross Platform and does not infringe on any current patents? Is it even possible to do so? How would you avoid infringing on someone else’s patents? Would constantly checking make the cost prohibitively expensive to even try? If it’s the caset then the computer industry is in real trouble!

        • halberenson says:

          Look, all commercial enterprises are “about the money”. Especially public ones. If they don’t maintain a focus on that then their management usually finds themselves out on the street and replaced by people who will. Facebook is learning this the hard way. Before their IPO everyone was focused on the size of their user base, after the IPO the focus shifted to how to monetize it. It’s all about “the money”. (BTW, I got the Facebook point from an analyst commenting on CNBC).

          It is certainly true that when Bill was Microsoft CEO the product groups worried far less “about the money” that they do today, but then growth was so rapid in the day that Microsoft could afford to absorb many failures. That isn’t so true today. Most product group people joined Microsoft because of the opportunity to change the world, and I think that’s still an important though faded part of Microsoft’s character. The “you build it, we’ll figure out how to sell it” attitude (which is literally what Steve, then running the field, said the first time I brought up the challenges of going against Oracle and IBM in the database world) is indeed gone. It was killed off by the “lost five years” and all the subsequent years playing catch-up. But it’s also about the maturity of both the industry and Microsoft itself.

          • John Kiely MCSE says:

            In Reply to this:
            Look, all commercial enterprises are “about the money”. Especially public ones. If they don’t maintain a focus on that then their management usually finds themselves out on the street and replaced by people who will.

            I would agree with you people go in to business to make money. But don’t you think Microsoft are hurting themselves these days with their pursuit of the almighty dollar? Microsoft have made a lot of money but they come across as this Titanic bully who closes out every opportunity to the competition and I mean any competition! They can appear to some as Mercenary! Now I don’t want to be accused of Microsoft Bashing as they do provide some really good products albeit done their way. But I have to question their abuse of their dominance of the Market? Add to that their relentless grasping of every technology and new idea that comes on the market? It’s like here is a new idea…. have they infringed on any patents “yes” means they go to court tying a new company up in court for decades. “No” and they buy the company!
            I use Microsoft products every day, I support Microsoft Products every day. I am very happy with how they do somethings and unhappy about how they do others, like wise I can say the same about Linux etc.
            But thanks to the current patent restrictions it would be impossible for “any” newcomer to create his own OS. Once he started with a boot manager he would have instantly violated some patent somewhere.
            The whole regime needs to be looked at right across the board. No new ideas/OS’s can ever be created as the system stands, It could appear to some that the whole industry is standing shoulder to shoulder and keeping everyone else out!
            As an example of how things go, support for XP is to be dropped in the near future. Windows 7 is next in line and then we are to expect Windows 8. Now I still can do pretty much everything I need to on XP. I currently see no technical reason to move OS. But Browser standards will change and the Operating systems will need to change to support the new browsers and then the hardware has to change to support the new OS’s. The user is then “forced” to upgrade.
            I am all for improvement but how is going from XP which takes about 3GB on your hard drive to going to W7 which takes 16GB on your hard drive an improvement?
            It’s a huge increase for not much gain? I mean XP will still browse the internet and download your emails, so what makes Windows 7 so special? This last point is exaggerated but maybe you can see my point?
            From time to time I still support and use Windows 98. Despite their being no support for this OS anymore.
            I have nothing against Microsoft I just think that right at this moment in time they have lost the plot!
            By the way I don’t want to be known as someone who criticises Microsoft for no good reason. I have merely made the points above as I currently see them. Overall they do a pretty good job.

  26. DQNOK says:

    There is LOTS of complaining on the internet about the current state of software patents. Indeed, there should be. The USPTO (United States Patent and Trademark Office) issues patents for the most frivolous software ideas. It’s just absolutely ridiculous.

    My first run-in with patents occurred when I was coding a recursion mechanism in a database system that didn’t have a recursion mechanism built in. I considered the problem for a few hours, came up with three possible approaches (completely on my own, with no internet research), rejected two of them and chose the third. About a month later I discovered that one of the methods I had rejected was actually patented! GOOD GRIEF! It was like patenting a method for painting a picket fence: “My patentable idea is: use a paintbrush.”

    But despite all the complaining (mine included), I have yet to read one article/opinion on WHY this is happening. What has run amok at the USPTO that is driving this bizarre behavior? The answer is: (drum roll please…) MONEY!

    In the United States, government organizations are SUPPOSED TO serve the needs of the greater citizenship of the US. To do that, and to prevent all kinds of abuses, including preferential treatment of certain companies, kingdom building, runaway salaries, etc., they are TAXPAYER FUNDED. This permits them a degree of objectivity that cannot be obtained where private funding is involved. It also makes them responsible to the taxpayer. And this is where the USPTO breaks from the standard of being a taxpayer-centric organization, and falls headlong into the corruption that comes from having the authority of the government, but being funded by private concerns.

    Here is a statement made to a congressional subcommittee. You can find it on their web-site: (
    “… the $2.7 billion in user fee collections currently projected for FY 2012 to execute our multi-year operating requirements. This results in an appropriation of $0 budget authority. As a fully user-fee funded agency, the USPTO’s requirements are addressed at no cost to the taxpayer.”

    This “let the user pay for the services” has fostered a very unhealthy atmosphere within the USPTO. To understand why, one must recognize that the employees at the USPTO are just like the rest of us: when possible, they will a) kingdom-build, and b) enrich their own pockets. There is nothing wrong with these behaviors within the free-market at large. It’s the essence of capitalism, and it’s what drives innovation and competitiveness. But within a government organization, where there is no risk of organization failure, and where you can enforce your will on the citizenship, it is deadly.

    Think about how this works, and of how the logic flows from the top management at the USPTO:
    Q: “How can I enrich myself?”
    A: “Build a kingdom. That will insure I’m elevated to a SES (Senior Executive Service) level, and insure my longevity.”

    Q: “How can I build a kingdom under myself?”
    A: “Increase the number of patent applications.”

    Q: “How do I increase the number of patent applications?”
    A: “Make it super-easy to get a patent. The word will get out.”

    Q: “How do I motivate my patent reviewers to always favor the applicant–make it very easy for them?”
    A: “Build a culture that favors patent approval over the requirement for genuine innovation.”

    Q: “How do I alter the culture?”
    A: “Give them financial incentives to approve patents. Promote them if they approve.”

    While the latter two questions aren’t addressed in the statement quoted earlier (obviously), the statement makes it clear (to me) that upper management is intent on kingdom-building.

    Now for some unsubstantiated rumor. I have heard, from what I consider to be a reliable source, that a good percentage of patent reviewers are GS-14’s. If you don’t know what that means, Google it (or just jump here: You will (or should) be stunned at the pay scale. That same source also told me that there IS a culture of approval among patent reviewers, and that promotions follow strong approval numbers. I could not confirm any of this from the USPTO web site, but I believe it. There is really no better explanation for the current patent insanity.

    Patents don’t favor innovation–at least not within the software world. Instead, they favor attorneys, and they stifle innovation. Designers are NOT free to innovate and produce. Instead, they have to constantly worry that the simplest solutions they come up with might actually be patented. So, software patents should really ONLY be granted for TRULY innovative devices. In contrast with that idea, check out another quote from the same statement quoted earlier:
    “… our overriding goal is to focus our resources more effectively on improving overall operations and reducing the time it takes to get a patent”

    OUCH! They actually want to move toward making it easier to get a patent, not more difficult.

    The USPTO needs to be completely reformed. They need to turn ALL patent fees into the general-fund where they cannot be swayed by them, and instead be taxpayer funded. This would help restore them to a citizen-centric view of their jobs, instead of their current big-business-centric view. Then they need to radically alter their culture of approval, and replace it with a culture of reluctance. They should QUIT rewarding reviewers for approving patents and instead reward them for legitimately documenting the historic and current state of the art, and explaining how the patent compares against that state. Then the reviewers work needs to be peer-reviewed, and finally vetted by an unbiased industry or academic panel (or something like this anyway; details, details). If the patent idea is TRULY a quantum leap forward, then fine, grant the patent. But if it is a natural flow of things that most intelligent designers would naturally arrive at, or if it is work that others have published in scholarly journals (or elsewhere) or it is even closely related, then NO, don’t award the patent.

    Until the profit/prestige motive can be divorced from the patent approval process, we will likely not get out of the current abysmal situation we’re in.

  27. Microsoft invests billions of dollars a year in R&D and then discovers that others have stolen their work and are profiting from it.

    Right. That’s why every product Microsoft has ever produced is a copy of someone else’s work.

    Note that I’m not just picking on Microsoft. Apple, Sony, Samsung, RIM, Oracle, etc. are all copyists.

    The other force at work is the anti-Intellectual Property community lead by the Free Software Foundation (FSF) and it’s GNU General Public License (GPL).  The problem here is actually two-fold.  There are those who really disdain patents and other forms of IP protection and those who support the idea of Intellectual Property for their own work but are willing to look the other way on theft of other’s property as long as it provides them sufficient economic gain.

    Horse manure. I’ve released software under the GPL in the past, because it is the STRONGEST form or protection against theft by someone like Microsoft or Apple (not that the specialty stuff I produce would be of interest to them, but th GPL provides the protection I want).

    I’ve also worked in the hardware business, and read hundreds if not thousands of patents. 90% of the patents that I read were garbage. In one case someone had patented something in 2007 that we’d been producing and selling since 1995, and approached us to take out a license.

    The only real proof that your R&D is worth the amount spent, is how good your sales are. If your sales are crap, you waste your money.


  28. halberenson says:

    Folks, I’m going to lock comments on this post. I think all the points have been made, and frankly I’m tired of the topic!

Comments are closed.