About a year ago -- on May 18, 2010 to be precise -- Microsoft sued Salesforce.com "for infringement of nine Microsoft patents by their CRM product." Apparently Salesforce.com had refused to enter into a royalty-bearing license agreement. About five weeks later -- on June 24, 2010 -- Salesforce.com countersued Microsoft over five patents.
The assignment history of those patents shows that three of the patents asserted by Salesforce.com had been assigned to it on May 3, 2010 by the Open Invention Network (OIN), a patent pool company that claims to protect Linux.
On the same day, the OIN had also given Salesforce.com a fourth patent, but for whatever reason Salesforce.com decided not to assert that one against Microsoft.
The dispute was settled another six weeks after Salesforce.com's countersuit: on August 4, 2010 Microsoft made this announcement, according to which "Microsoft indicated that it is being compensated by Salesforce.com based on the strength of Microsoft's leading patent portfolio in the areas of operating systems, cloud services and customer relationship management software." In other words, Salesforce.com came out on the losing end regardless of OIN's patent gift.
The OIN usually likes to brag about how it thwarted Microsoft's allegedly Linux-hostile plans and helped TomTom (although that's highly doubtful, given that the outcome was like in the Salesforce.com case and Microsoft ended up receiving royalties). But very interestingly, the OIN never talked about its transfer of patents to Salesforce.com. This raises important and serious issues concerning not only the questionable effectiveness and unquestionable opacity of the OIN but also this defensive-patents-on-demand kind of service, which other patent pools such as publicly-traded RPX Corp. offer as well. With so much patent litigation going on, there will be more lawsuits down the road in which patents checked out from pools will be asserted.
The patents at issue in the Microsoft-Salesforce.com dispute
I looked at the nine patents Microsoft asserted against Salesforce.com. While Salesforce.com runs its server farm on Linux, none of those nine patents appears to me to be specifically Linux-related. There are various Microsoft patents (such as the FAT patents) that definitely read on the Linux kernel, and the aforementioned announcement of the settlement suggests that Salesforce.com is now paying Microsoft royalties for using Linux, but the nine patents Microsoft asserted against Salesforce.com in court appear to cover typical cloud and web service functionality (though only detailed claim charts could provide absolute certainty concerning the actual infringement patterns).
Therefore, the OIN's involvement with this dispute was in a gray area, which could be part of the reason why the OIN never talked about it. The OIN claims to protect Linux (according to its own arbitrary and changing "Linux System" definition). Salesforce.com's cloud software running on top of Linux was never part of that definition. Nevertheless the OIN apparently interpreted Microsoft's patent suit against Salesforce.com as a Linux issue. Since the OIN gave Salesforce.com those four patents even before Microsoft's suit, it's possible that the OIN and Salesforce merely expected Microsoft to sue over Linux-related patents.
These are the assignment records of the four patents the OIN transferred to Salesforce.com in the midst of the dispute (but prior to the first formal lawsuit):
U.S. Patent No. 6,813,633 on a "dynamic multi-level cache manager": initially (2002) assigned to a company named Foedero Technologies in the Canadian province of Ontario; later (2004) assigned to another Ontario entity named Avokia Inc.; in 2009, acquired by the OIN, and transferred to Salesforce.com on May 3, 2010 (recorded on June 17, 2010)
U.S. Patent No. 6,918,059 on a "method and system for handling errors in a distributed computer system": initially (2000) assigned to the world's largest record label, the Universal Music Group, whose name and address was incorrect in the original record and corrected in 2005; acquired by the OIN in 2010 and sold to Salesforce.com on May 3, 2010 (recorded on June 17, 2010)
U.S. Patent No. 7,024,454 on "work sharing and communicating in a web site system": initially (2001) assigned to a company named Practicefirst.com, LLC, which may have been a different entity than the current owner of that Internet address; acquired by the OIN in 2008; sold to Salesforce.com on May 3, 2010 (recorded on June 17, 2010)
U.S. Patent No. 7,251,745 on "transparent TCP connection failover": this patent was initially (2003) assigned to Eternal Systems, Inc. in San Jose, California; reassigned in 2005 to a company named Availigent, Inc. in the same city; acquired by the OIN in 2008; sold to Salesforce.com on May 3, 2010 (recorded on June 17, 2010); this is the ex-OIN patent that Salesforce.com did not assert against Microsoft
The two non-OIN patents asserted by Salesforce.com against Microsoft were U.S. Patent No. 7,209,929, an original Salesforce.com patent on a "Java object cache server for databases", and U.S. Patent No. 7,305,454 on an "apparatus and methods for provisioning services", which was filed by a company in San Francisco and changed hands a few times (mostly within the same building) before being acquired by Salesfore.com in 2007.
Checked-out patents are better than none -- but not the strongest weapon
If a company doesn't have enough patents of its own to assert, it may try its luck with patents checked out from pools. Apparently Salesforce.com had only one internal patent that it thought it could assert against Microsoft, plus one it acquired a few year before, so getting patents from OIN was necessary to put together a set of five patents, which looked reasonably solid just based on the number.
In principle, a patent can be transferred and asserted on the same day. It's not against the law, but is it effective?
I'm sure that every litigator will prefer to assert patents that can be portrayed to a judge and a jury as rights related to legitimate internal inventions. It's much better in psychological terms to ask the courts for help if someone else (allegedly) infringes on one's own intellectual property.
If the dispute between Microsoft and Salesforce.com had gone to trial, Microsoft's lawyers could (and probably would) have pointed out to the judge and the jury what the history of those patents was. In that case, Salesforce.com could have told the story of how it needed to acquire those patents for defensive purposes. But no matter how much the concept of mutually assured destruction (or mutually assured damage) is a reality, it's just not the story you want to tell a judge and a jury when it comes to why you ask for an injunction and a damage award. In that scenario you want to be as close as possible to the lone inventor whose rights someone else blatantly disregards.
Besides the psychological aspect of this, the legal position of a patent holder can't be better than if the right holder actually practices the invention (such as by selling products on which the patent reads). Of course, if one acquires a patent and practices the invention, then the fact that the patent wasn't "homegrown" makes only a psychological difference. But it's not easy to find and acquire patents covering inventions one actually practices.
The Salesforce.com patent transfer raises serious questions about the OIN
I previously pointed out that the OIN apparently transferred four of its patents to Salesforce.com even before Microsoft had formally sued, and I can't see a clear Linux context in Microsoft's lawsuit (only in the announcement of the settlement). Interestingly, Salesforce.com was not even an OIN member or licensee at the time of that transfer. It isn't on the list of OIN licensees even now that I write these lines (more than a year after the transfer). Unless the OIN did a clandestine agreement with Salesforce.com, this means that Salesforce.com got support from the OIN (even though that support apparently didn't have much impact) but would still be free to assert its own patents against any current or future OIN members. That seems very unbalanced and questionable. By contrast, TomTom became an OIN member during its dispute with Microsoft.
It's generally disconcerting that the OIN secretly sells some of its patents to third parties. The OIN portrays its patent purchases as a way to ensure certain patents don't fall in the hands of "trolls" or strategic enemies of Linux. But if the OIN sells patents to non-members and non-licensees like Salesforce.com, it's possible (unless the OIN took precautionary measures it never talked about) that those entities might sell those patents on to other parties who could use them against OIN members, at least in non-Linux contexts. In that case, the OIN would end up contributing -- indirectly -- to the rampant troll problem.
I believe the OIN owes its licensees -- some of whom have supported it for years now -- an explanation as to the criteria by which the OIN lets others check out some of its patents for the purpose of countersuits and counterclaims. It's a safe assumption that Microsoft's settlement with Salesforce.com (since the announcement suggests it's a mutual license to the companies' entire portfolios) precludes any future assertion of those patents against Microsoft. So the OIN believes certain patents are useful for countersuits and counterclaims, it would have to use them wisely. It should usually reward the loyalty of long-term supporters, but in this case it came to the aid of an entity that never supported the OIN in any official way. (Maybe Salesforce.com would have joined the OIN if those patents had been any good.)
All patent pools have the problem that a patent can only be checked out by one licensee/customer at the same time. By giving those patents to Salesforce.com, the OIN made it impossible for any of its supporters to have access to those patents if they had needed them at the same time.
And there's the question of openness. If an entity calls itself the Open Invention Network and claims to be a guardian of open source, it should be reasonably transparent. If it transfers patents, it should tell its licensees that it did so, and specifiy why and on which terms such a transfer took place. The Salesforce.com patent deal went almost completely unnoticed. The only mentioning of that deal that I was able to find on the Internet was on a Japenese CNET blog.
The OIN is preparing a major reform, and I hope "OIN 2.0" will then be more forthcoming about its dealings -- including its defeats.
If you'd like to be updated on the smartphone patent disputes and other intellectual property matters I cover, please subscribe to my RSS feed (in the right-hand column) and/or follow me on Twitter @FOSSpatents.
Share with other professionals via LinkedIn: