OSI Board reponds to FCO Questionaire concerning CPTN Transaction
Towards the end of March, we received a message from the German Bundeskartellamt (Federal Cartel Office or FCO) advising us that the CPTN transaction had been re-notified to them. That means that the consortium seeking to acquire Novell’s patent portfolio – Microsoft, Apple, EMC and Oracle – had once again asked for permission to proceed.
OSI Concerns Heeded
Notably, the terms of the transaction seem to have been significantly changed, apparently in response to concerns like the ones OSI expressed at the start of the year. OSI is very pleased that the FCO has been clear about the transaction with CPTN and congratulates them on continuing to consider the overall health of the evolving software market and not just the concerns of the existing dominant players.
Here is a non-expert summary of the differences (summarised with permission from the FCO):
- CPTN will now only exist for long enough to distribute the shares equally among the participants in the transaction (no more than three months), and thus will not form a new long-term patent troll itself.
- All parties to the transaction will retain a license to the full Novell patent portfolio, thus immunising themselves from patent actions with the shares they do not hold.
- Microsoft will sell its 25% share of the patents on to Attachmate and retain only a license to the portfolio.
- EMC will ensure that the 31 patents it has determined relate to virtualisation are not among the 25% share it acquires.
- All patents will still be subject to all existing licenses, covenants not to sue and similar restrictions.
OSI Still Concerned
The FCO went on to ask OSI for its views on the revised transaction. The OSI Board responded to their questionnaire as follows:
Response from the Open Source Initiative to FCO Questionnaire
April 4, 2011
The Open Source Initiative (OSI) would like to thank the Bundeskartellamt (Federal Competition Office) for their continued investigations related to file B7 – 11/11 (the proposed CPTN-Transaction). When the OSI first learned of this proposed transaction, we were alarmed that four companies with dominant market positions and a mixed attitude towards open source software could redeploy what the open source community had considered to be a friendly asset–Novell’s patent portfolio–into a weapon against open source software. We are delighted that you have made clear that the FCO cannot allow a transaction that would create or strengthen a dominant position on markets in which such investors are active, and we are happy to provide the additional information you have requested about the proposed restructuring of this transaction. Our responses follow your questions.
1. In which way could in your view the CPTN-Transaction have effects on your activities?
Open source is a development method for software that harnesses the power of distributed peer review and transparency of process. The promise of open source is better quality, higher reliability, long-term stability, more flexibility, better cost control (and frequently lower cost), and an end to predatory vendor lock-in. The Open Source Initiative (OSI) is a non-profit corporation with global scope formed to educate about and advocate for the benefits of open source and to build bridges among different constituencies in the open source community. One of our most important activities is as a standards body, maintaining the Open Source Definition for the good of the community. The Open Source Initiative Approved License trademark and program creates a nexus of trust around which developers, users, corporations and governments can organize open source cooperation.
Our first concern about the CPTN-Transaction, and the perspective we take in our response today, is whether and how the CPTN-Transaction stands as an existential threat to open source software, and hence to the OSI itself. For all the strengths of the open source software model, which has encouraged the creation of more than one billion source lines of code with an economic value estimated at more than $150B USD ( http://www.opensource.org/files/OSICPTNPosition.pdf ), its premise of liberal access, distribution, modification, and redistribution of software source code is largely incompatible with per-unit patent royalty models. Some companies, such as IBM, have resolved this by granting explicit licenses to specific patents when practiced in software covered by an OSI-approved license. Some companies, such as Red Hat, have resolved this by effectively promising to not assert any patent claims against any use of software it issues (and provided that the recipient does not initial patent litigation against the company). Some consortia, such as the Open Innovation Network (OIN), have pooled patents to create a mutual defense for open source projects against patent harassment. Some OSI-approved licenses, such as the GNU General Public License and the Apache License, require that any contribution of software also convey with it any necessary claims of any patents that might read on the software or the contribution. All of this is to say that the open source community does not close its eyes to the existence of patents, but that we actively and positively promote the idea that being a member of the community means not asserting patents against open source. As an independent company, Novell understood this and never once asserted a single patent against any open source software developer, project, proprietorship or company.
If Novell’s 882 issued patents and patent applications (“sold patents”) are converted from a defensive asset that helps protect the integrity of the open source community into an offensive weapon that can be used against the open source community, the activities of the OSI will be greatly affected. Instead of promoting the benefits of one billion source lines of code, we will be forced to defend the legitimacy of that asset. Instead of promoting trust in community-developed software, we will have to defend the very idea that individuals can safely develop software without the protection of expensive and extensive legal departments and corporate war-chests. As we explained in our first response, the open source community has thus far largely avoided the damage of strategic patent assertions. If the CPTN-Transaction allows one of the “sold patents” to be asserted against open source software, and such assertion leads to diminished competition, then the OSI and the open source community will suffer not only that loss, but also the loss of trust of what it means to rely on open source software for any number of applications. The OSI has no resources to defend against or to mitigate the damage of even a single patent assertion against open source.
2. How do you evaluate the modifications to the transaction according to which Microsoft will not acquire any patents but just licenses for the Novell patents?
This is a significant modification, one which makes Microsoft’s participation in this transaction largely unobjectionable. That does not make the CPTN-Transaction itself unobjectionable, but by merely taking licenses to the patents without taking ownership of the patents, Microsoft does not itself become a greater threat to the open source community than they already are.
We do not have the resources to investigate how much influence Microsoft has over Attachmate’s strategy, nor do we have the resources to develop the knowledge as to whether Attachmate has any inclination to use the Microsoft-selected patents strategically against open source software. Nor is there any way to know whether Attachmate intends to dispose of the patents after acquiring them from Microsoft. This would not be a concern if Microsoft took the step of taking a license not only for itself, but also for the open source community, before handing the patents to Attachmate. Had they done this, the OSI would not only not object to Microsoft’s participation in the CPTN-Transaction, but would have held it up as a model that each of the other three companies should take.
3. On the basis that EMC will not acquire the above-mentioned 31 patents relevant for virtualization – would you still expect any negative impact of the transaction in this field?
This is an extremely difficult question to answer with confidence for a number of reasons.
First, the OSI does not know the details of the “sold patents” and does not have the financial, technical, and/or legal resources to determine that the above-mentioned 31 patents are the only ones that could create or strengthen EMC’s dominant position in Virtualization.
Second, In the US v. Microsoft case it was determined that Microsoft was able to use monopoly power in one domain to achieve monopoly power in a second domain. We are concerned that the outright acquisition of more than 100 patents could effectively create or strengthen EMC’s dominant position in Virtualization, even if they do not contain the 31 patents EMC claims are related to Virtualization. Patents unrelated to EMC’s own technology might still be used to weaken an open source competitor’s position.
Third, if EMC were to use one of the once-friendly Novell patents strategically against an open source project in order to weaken overall trust in open source software, then even an unrelated patent could weaken competition from open source virtualization. If EMC were to make an enforceable promise not to assert any of its patents against software covered by an OSI-approved license, we would then have no such concerns. As it stands, EMC has removed one specific objection from our original complaint, but has left the majority of our concerns unaddressed.
4. If according to you the transaction raises concerns, please explain in detail, which of the CPTN-Investors would acquire or strengthen its dominant position on which product and geographic market.
The OSI remains greatly concerned that the CPTN-Transaction will allow market-dominant companies to diminish competition from open source software by asserting patents that were once held in friendly hands (Novell). After reading the FCO’s summary of the proposed changes to the CPTN-Transaction we remain highly alarmed about Oracle and Apple’s positions. (Our position on Microsoft and EMC have been expressed already.)
Having acquired Sun Microsystems, Oracle has a dominant position in Middleware (with Java) and strong positions in Operating Systems and Virtualization (with Solaris). In contrast with Microsoft’s position to hold only licenses but not patents, Oracle is clearly taking more than a defensive position. By holding patents, Oracle has both the right of action to assert once-friendly patents against the open source community, but they also have the right to grant patent licenses as well. Unfortunately, they have stated no intention to grant any patent licenses under terms compatible with the Open Source Definition ( http://opensource.org/docs/osd ), so we must focus our response on how they might acquire or strengthen their dominant position when it comes to Middleware, Operating Systems, Virtualization, and even Mobile applications and platforms.
Java is not only the core of the Middleware market, but it is the basis for Google’s Android platform, one of the fastest-growing mobile platforms and a great success story for open source software. The OSI is not alone in believing that one of the key reasons that Java became the universal platform for Middleware was due to the patent non-aggression pact written into the Java Community Process’s (JCP) Java Specification Participation Agreement (JSPA)
in 2002. In 2006, Sun further enhanced the value of Java by relicensing it under the GNU General Public License (GPL), an OSI-approved license. In 2010, Oracle launched a multi-pronged challenge against Google, asserting both copyright infringement and patent infringement, based on Oracle’s ownership of the Java code base and Java-related patents. Oracle has also refused the Apache Software Foundation (ASF), a JCP and JCP Executive Committee (EC) member, access to the TCKs as agreed between Sun and the ASF. What is noteworthy is that before Oracle acquired Sun, Oracle was on the record as supporting the ASF and in strongly urging Sun to adhere to its contractual obligations as specified in the JSPA. After the acquisition, Oracle’s position changed, and they continued to disregard the commitment made by Sun before them.
The OSI is specifically concerned that Oracle may use the patents it selects to disrupt competitors — including open source competitors — in Operating Systems, Middleware, Virtualization, and Mobile applications and platforms. While the Linux kernel itself may be somewhat protected by OIN from direct attack via patent aggression, the kernel itself has no value to users except as a platform upon which to run applications. The OSI does not have the financial, technical, or legal resources to research in detail which “sold patents” Oracle may assert against which components of a Linux distribution to make Linux a less-viable competitor to Solaris, or how they might assert such patents so as to minimize the viability of the open source community as developers and purveyors of Linux vs. their own Linux development and distribution. We do know, however, that Oracle has been legally aggressive with regard to Java. Oracle has already out-flanked the Java community and is challenging Java-based projects and platforms. Without further investigation or limitations to their behavior, it would be difficult if not impossible to repair the damage to the Linux ecosystem if Oracle took once-friendly patents and turned them against the Linux community.
Oracle is using the technological fact that Virtualization and Middleware are key components to enabling mobile technology. Thus Oracle’s prior actions suggest that Oracle may be planning to create a dominant position in Mobile at Google’s expense. The OSI has no financial interest in Google itself, but the Android platform is an index case of the success of open source software development, and to see Oracle take Android down based on legal assertions of Middleware and Virtualization patents would significantly weaken our efforts in promoting open source to mobile developers.
Finally, though we don’t know which patents relate to Virtualization or whether Oracle will get some or all of them, Oracle could use once-friendly Novell patents to shut down Virtualization innovation in Linux and attempt to shift the market to their proprietary platform, Solaris.
In summary, the newly structured CPTN-Transaction does little to allay our fears about what Oracle may do, and given that Oracle has not taken the opportunity to adopt the more purely defensive position chosen by Microsoft, it actually heightens our fears that Oracle is planning to do precisely what they should not, namely to create or strengthen a dominant position in the market as a result of this transaction.
The OSI has many similar concerns with respect to Apple, especially since Steve Jobs has personally threatened to use software patents to “go after Theora and other ‘open source’ codecs now”. (See http://www.freerepublic.com/focus/f-chat/2505640/posts ) The OSI does not have the financial, technical, nor legal resources to specifically investigate how Apple might use its share of the 882 “sold patents”, but Apple clearly has a dominant position in Mobile applications and platforms. The Mobile market is critically dependent upon Operating Systems, Middleware, Virtualization and virtual-machine technologies (in addition to networking technologies that Novell helped to pioneer back in the days when PCs were not network nor Internet-enabled). Specifically, it seems plausible that Apple’s most credible competitor in the mobile market, Android, would be vulnerable to challenge by the patents involved in the CPTN-transaction. The OSI is very concerned that Apple could use once-friendly Novell patents to make it difficult or impossible to create competitive Mobile platforms or mobile applications developed as open source.
Both Apple and Oracle could address these concerns by stipulating fully-paid, world-wide, royalty free licenses to any software covered by an OSI-approved license. But absent such a promise, it would be far too easy for either of them to harm not only a specific open source project (such as Open Manager, a file browser optimized for mobile) or an open platform (such as Android), but the larger open source ecosystem, whose goals of eliminating vendor lock-in are at odds with the de facto result of monopoly: perfect vendor lock-in.
The concerns expressed above are independent of geographic region, but they are certainly relevant in the US and EU (not to mention Latin America, India, Eastern Europe, and the ASEAN countries). The concerns also extend beyond mere commercial applications: the Linux kernel today runs more than 90% of the Top 500 supercomputers ( http://www.opensource.org/node/492 ), and is thus a crucial component of cutting edge research. Federal, State, and Local Governments looking to provide more services at lower cost have created more than 350 policies concerning open source software evaluation and procurement ( http://www.opensource.org/node/549 ), an increase of more than 100 policies in the past two years alone. Many e-Government initiatives depend on Middleware at the application layer, Operating Systems and Virtualization for deployment, and Mobile applications and platforms to deliver services to citizens, and the OSI sees great progress in open source software to provide the necessary innovation in research and the public sector. We fear that several of the CPTN-Investors want to stop this progress and that the CPTN-Transaction provides both direct and indirect means to do so by asserting once-friendly patents against open source software.
5. Are there any further remarks or aspects for the competitive assessment of the CPTN-Transaction which you would like to bring to our attention?
The over-arching problem with the CPTN-Transaction is that no matter who ends up with what patents, with so many thousands of applications and so many billions of lines of open source software, any of these once-friendly patents can be used to weaken the competitive challenge that open source software offers in the domains of Operating Systems, Middleware, Virtualization, and Mobile applications and platforms. The fact that we cannot know who will be the first aggressor should not be a reason to permit the transaction to go through as currently constructed. No-one can know what agreements have been made between the companies involved in the CPTN-transaction regarding targets for the patents involved, and having any one of them avoid ownership of patents relevant to strengthening their own dominant position does nothing to prevent one of the other participants acting as their proxy. The solution with respect to the open source community is for CPTN to grant, in addition to its four members, a fully-paid, royalty-free, world-wide license for all patents when practiced by software covered by an open source license. Until all four parties have clearly moved to a purely defensive position, the threat of these once-friendly patents becoming a weapon against competition from open source remains.
Thank you for your consideration of our concerns. Please feel free to contact us for any clarifications or for any further information we can provide.
Michael Tiemann
President, Open Source Initiative