I posted the following piece at my other blog (at the Linux foundation site). If you haven't checked out the home page of the Foundation before, you should check it out, as it consolidates quite a bit of news of interest to the Community, including the blog entries of many of the staff.
One of the enduring soap operas this year has involved the ongoing patent infringement threats by Microsoft against “Linux, OpenOffice, email, and other open source software.” According to Microsoft, 235 of its (unnamed) patents are being infringed, and it should be entitled to be paid for this use of its intellectual property. Steve Ballmer believes that Microsoft owes it to its stockholders to file patents to protect its innovations, and then to assert these intellectual property rights in this way, and at this time.
Of course, Linux is based on Unix, which has been around for decades, as have many flavors of Unix created by IBM, HP, Sun and others. Curiously, Microsoft is not now, and never has, alleged that those systems infringe upon these same patents. Apparently, its stockholders do not expect it to assert patents against Unix vendors or users - just open source vendors.
Equally curiously, if not more so, is the fact that until a few years ago Microsoft filed only a small number of patents annually, in comparison to the thousands of patent applications per year that it is filing now. Apparently, its stockholders did not expect the fruits of Microsoft’s innovation to be protected and asserted in this way until now.
It is true that many other major vendors file many patent applications each year. But increasingly, IT vendors secure such patents purely for defensive, and not offensive purposes. As a result, patents are today being secured by inventors not to protect their inventions from infringement, but to secure the practical right to infringe upon the inventions of other inventors. If one vendor objects that another vendor’s products are violating its patents, the second vendor simply points to the patents in its own portfolio that are inevitably being infringed by the first vendor’s products. The result? At most a cross license, often with little or no money actually changing hands.
Linux, of course, includes enormous innovation – and more creativity is contributed every day. But the community does not file patents on its innovation. After all – why would it? FOSS innovation is intended to be used by all, and for free. And patents by their nature exist to prevent others from using innovation, or to charge for the privilege of doing so.
This leaves open source software, and particularly community software (as compared to FOSS that is supported by major vendors), more vulnerable to those that oppose it, or are threatened by it.
All of which goes to show that the whole rationale for the existence of software patents has turned upside down, and inside out. Patents are secured not to protect inventors, but to protect infringers – and to rob FOSS users of the ability to benefit from innovation that is intended by its inventors to be freely available to all.
It’s a sad situation, and one that will soon be at least alleviated by legislative reform. But will that be enough?
My own opinion is that software patents not only cause harm, but serve no useful purpose in fostering innovation at all. That’s been my experience over almost thirty years of representing software companies, as I’ve written about in greater detail here.
Maybe, someday, the US will adopt the view of the EU that software is not entitled to patent protection, and can be adequately protected (as it can) by trade secrets and copyright protection.
For my money, that would be a happy day.
For further blog entries on Intellectual Property Rights issues, click here
subscribe to the free Consortium Standards Bulletin
Correction: last line I think you meant "copyright protection"
Right you are, of course. Thanks for the catch.
It’s also possible that the EU might adopt the US approach, and decide that ‘software’ is deserving of patentability.
This guy Bill Thompson thinks that if software patents came to Europe, as a type of ‘industrial property’, then they wouldn’t apply to the sort of software that you or I could run on our desktop computers.
I think that that’s exactly the sort of software it would be lucrative to sell patent licences for, though. So I’m concerned.
Just as a minor nitpick – I think the SCO saga has comprehensively proven that Linux is most definitely *not* based on Unix. It may be a POSIX compliant, or ‘unix-like’, but it is not derived from the Unix code base.
"Based" may not be the best word, but the sense of what I’m trying to convey is important to the point that I’m making, which is that operating system kernels need to do certain things in order to be useful. Being POSIX compliant also constrains software to a certain extent. And using UNIX as a developmental reference point herds the sheep even more. The result is that presumably both UNIX and Linux execute certain operations in similar ways, though not using the same code to do so.
The second important point is that the SCO suit was about alleged copyright infringement – literal code reuse – rather than patent infringement – deliberate or inadvertent similarity in design alleged to violate an invention (it also alleged trade secrets misappropropriation). So while the SCO litigation was very important to debunk SCO’s contentions that it’s code had been copied, it’s not instructive as to whether Linux (or UNIX, or implementations of POSIX) would violate patents owned by Microsoft, IBM, HP, or anyone else.
The fact that Microsoft has never alleged that any flavor of UNIX – including SCOS’s – infringes any of its patents is therefore significant to the extent that the Linux architecture executes any operations in a manner similar to UNIX.
Wholeheartedly agree – you and I know what you meant – I just didn’t want the ‘less investigative’ journalists quoting ‘Andy Upgrove says: "Linux is based on Unix"’…
"Of course, Linux is based on Unix, which has been around for decades, as have many flavors of Unix created by IBM, HP, Sun and others. Curiously, Microsoft is not now, and never has, alleged that those systems infringe upon these same patents. Apparently, its stockholders do not expect it to assert patents against Unix vendors or users – just open source vendors."
Sadly, it seems this blog may becoming a place where serious discussion is no longer available.
You know perfectly well that implementations may differ between those systems such that one implementation might infringe a patent and another not. That the source code for AIX and HP-UX, and until relatively recently Solaris, is not publicly available might also have a little something to do with it as well.
You don’t like software patents, that’s fine. A case can certainly be made against them, so make that case. But hyperbole like what I quoted above makes me think this blog is becoming the Faux… err… Fox News of open source. Matt Asay is already filling that role quite nicely.
"But increasingly, IT vendors secure such patents purely for defensive, and not offensive purposes."
Increasingly? 99.99% of software patent filings are for defensive purposes.
"the second vendor simply points to the patents in its own portfolio that are inevitably being infringed by the first vendor’s products."
Nah, they just say "We can allege infringement, do you really want to take the chance we’ll actually find something?"
"Linux, of course, includes enormous innovation"
Depends on your definition of innovation.
Btw, as I’ve noted in other places, in the Sun/NetApp dispute at least NetApp was specific about what was infringed (unlike those in the great Northwest). And I find it interesting that Sun claims that several of NetApp’s patent infringement claims are "spurious" (http://www.sun.com/lawsuit/zfs/), yet when I ask WHY they are spurious, I’ve never seen an answer. Things that make you go hmmm…
If we were talking about 1 or 2 patents, it would be fair to call me on that. But Microsoft’s count is 235. And you know as well as I do that whenever a suit is filed between major vendors with big patent portfolios that there are always patent counterclaims. That’s where the phrase "patent thicket" comes from. So I feel pretty comfortable that yes, out of all those patents some would be violated by other OSs as well. As you know, for example, Apple settled with Acacia before Acacia sued Novell. Case in point.
Are software patents filed offensively or defensively? Don’t ask me, ask IBM, HP, etc. I personally know their senior standards and OS strategists and legal counsel. This comes from them, not me. And they’re not "Fox News" – they’re the source of the story itself.
For what it’s worth, I think IBM files for patents (of all kinds) as a sort-of ‘business optimisation’.
In words that may be easier to understand, "IBM files for patents because the commercial laws of the countries that IBM does business in say that patents exist"
Now, maybe you can get someone official from IBM, from HP, from whichever other large patent-holding businesses you can find, to add remarks as to whether that’s accurate or not. How should I know ? But as an engineer I would be interested in the answer from all comers.
Also, do patents grow the business ? And what do patents do for the supply of future employees, those who are in school or university at the moment ?
I think there are two answers to this question.
The more sophisticated way that some companies with large software patent portfolios phrase it is: "it preserves our freedom of movement," by which they mean that if they want to put out a new product or enter a new area and someone already there has a patent that would block them, well, they have one (or more), too. So at most they can enter into a cross license that allows them to move forward despite the patents owned by the companies that are there already.
The second is that many engineering-based companies that have been around for a long time have a very "patent intense" internal culture. Your status inside the company, and your bonuses, are heavily impacted by your prowess as an engineer, as measured by the number of patents with your name on them. As a result, the company ends up with thousands and thousands of patents, many of which it doesn’t really "know" it owns anymore in any productive way. Other people are infringing them, and they don’t even know it. Another example of a system that’s run to a certain extent just on autopilot, and through tradition more than strategy.
Well, that is OK until someone files a lawsuit.
Then you get an ‘innovation crunch’, like the ‘credit crunch’ that is causing collateral damage and house repossessions in the mortgage market (and certain problems with Citibank, Wachovia, Norther Rock, and others).
So how do ‘we’ collectively … large corporations, universities, and hobbyists … get out of it ?
Ah, there’s the rub. The bills in Congress could help some; some recent court decisions help some. The only decisive ways would be to either amend the law to expressly state that software inventions are not entitled to benefit from patents, or for courts to go back to the conclusion of not so very long ago that software code cannot create a patentable invention at all.
Crawling through patents is a huge waste of engineering effort; and the possibility of various kinds of liability is preventing major corporations from allowing their engineers to teach (e.g.) Linux and OpenOffice.org in local schools; correspondingly POSIX standards and ISO26300 are not spreading as fast as they should (or are being suppressed) with obvious consequences for skills in engineering and science. The only major corporation which has an interest in teaching Microsoft Windows and Microsoft Office is Microsoft, so we leave that to them.
It’s likely that all the claims in some pending Apple litigation will turn out to be baseless; or the patents revoked as ‘obvious’, ‘prior art’, ‘unpatentable material’, or ‘owner did not have a working implementation on the filing date’. However it is likely that the litigants will not easily accept the obvious, and so it is going to take a lot of resources to do. Will it be resources from lawyers ? Engineers ? Clerks ? University students ? Librarians ?
If Apple felt like hiring me,I could probably dig up the evidence to undermine things personally. However, as things stand, they are unlikely to hire me and I have better, more productive things to do with my time. But I still think litigating over the patents is a losing strategy; I hope ‘Burst’ don’t see a dime of Apple’s money, and I fully expect that ‘Burst’ will be asked to pay their own and Apple’s expenses over the matter.