451 CAOS Theory 
A blog for the enterprise open source community
On standards and discrimination
Matthew Aslett, July 4, 2008 @ 6:39 am ETThe Business Software Alliance has got itself in something of a flap over the European Commission’s European Interoperability Framework for pan-European eGovernment Services and its stance on patents in standards, according to this report in The Register.
The report quotes BSA’s European software policy director Benoit Müller as saying:
“Our concern is that they [IDABC] will still insist on standards that have no intellectual property [IP] rights and they will not revise the definition of an open standard as contained in EIF version one. They define open standards inconsistent with the common understanding of the term in what we believe is a dogmatic approach. It fails to recognise that almost all standards that help interoperability and that governments should indeed use to promote the very objectives of the EIF do have intellectual property.”
The BSA’s position is pretty much identical to that adopted by the Association for Competitive Technology, which you can read more about here.
However, according to the Register report, both the BSA and ACT are misinterpreting the forthcoming draft EIF. It quotes Karel De Vriendt at the IDABC as saying:
“We stressed the need for the evaluation of standards along many dimensions… one of them being openness and the possibility to implement the standard in open source software (which is of course only important if you value open source software). Standards with IP rights regimes that prohibit the implementation of these standards in open source software should, if more open alternative standards exist, not be used as the basis to achieve interoperability if your organisation wants to be able to benefit from open source software.”
In other words, if you value open source software (which the EC does, as it is seen as a means by which the competitiveness of the European Union’s ICT industry can be improved) it would be best not to adopt standards that discriminate against open source software. Sounds pretty reasonable to me.
But then it depends on your definition of discrimination, I suppose. As Müller told The Register:
“When open source is the best solution, the most cost effective and most suited solution to a given government needs then that is of course what a government should procure. But we’re against preference policies. Some attention should be paid to open source as long as it does not lead to discrimination.”
Comments (7) Categories: Software
It’s not a matter of if you value open source or not, it’s a matter of if you value the best standard or not (best of course being in the eye of beholder). What you’re saying is that you’d prefer an unencumbered, but inferior standard to a superior standard encumbered with associated RAND patents. In some cases, that may be the right approach, in others it may not.
In other words, if you value open source software (which the EC does, as it is seen as a means by which the competitiveness of the European Union’s ICT industry can be improved) it would be best not to adopt standards that discriminate against open source software. Sounds pretty reasonable to me
That does SOUND reasonable. However, it is a complete red herring. It is all in how you define what is open source compatible. Any standard, even those fully encumbered with royalty-bearing patents, is actually open source compatible! Any claim otherwise ignores the reality of the marketplace today.
Matthew, as you well know, enterprise versions of Linux from Red Hat et al almost ALWAYS ship with proprietary code alongside the open source code to enable them to integrate certain protocols and standards. Even the pre-XP OLPC had proprietary components in it. The GPL allows this and every other open source license allows it.
Unfortunately, however, it is the Free Software ideologues that are driving this debate. They want to redefine open source compatible to mean that the standard must be able to implemented in GPL code. Based on their religious convictions about Free Software, they refuse to use proprietary or non-GPL code even when it is allowed by the GPL license.
So, we need to stop confusing the concepts of “Open Source Compatible” with what is “Compatible with the Free Software Religion.”
Mark Blafkin
Association for Competitive Technology
Actually what I am saying is that, because it values open source software, the EC prefers an unencumbered standard to a standard encumbered with associated RAND patents, as is its right.
Of course, if Mark Blafkin is right and there is no problem with encumbered patents and open source, then the statement is even simpler: the EC prefers an unencumbered standard to a standard encumbered with associated RAND patents, as is its right.
Matthew
The Commission certainly has the right to choose whatever policy it thinks is right in the end. It is also our right and the right of our European members to advise them that we believe the policy will not serve the interests of government, citizens, or industry. The Free Software community has been lobbying on this issue for a very long time and it important the Commission here from alternative voices about the potential effects of this policy.
Finally, if the final result of the policy is as balanced as the EC claims in Register article, I don’t think we’ll have a problem with it. However, that was not the impression given during the presentation last week or from the first draft of the report.
I don’t think we need worry about the EC not hearing alternative views to the free software community when it comes to lobbying.
Dear Mr. Mark Blafkin,
You statement ignores the reason why the EU is such a proponent of open standard in their given definition. They want government applications and electronic documents to be as available as possible to all citizens. I think this is a premise that even your organization could approve off.
However, this statement implies that the implementation of these standards should not be encumbered by patents. Let’s take take a look at the wording of the EIF report: “The intellectual property – i.e. patents possibly present – of (parts of) the standard is made irrevocably available on a royalty-free basis.” I see two reasons for this part of the report:
- people should not be forced to have to pay for any software using the standard. If this would be the case, there would be a barrier for a sizable percentage of citizens to work with the standard, and thus the government that uses that standard.
- people should have a free choice in the software they choose to access the documents from the government. However, if the standard would not be open in the sense given above, there would be an artificial limitation on the implementation of that standard. This would force out a certain amount of software makers, most notably the Open Source software makers of course, but also small software creators who would have to pay royalty fees for using the standard. This, in it’s own, would also limit the amount of choice the citizens have in software programs for accessing the government documents.
Do understand that such a policy by the EU would only affect any standard adopted by the governments. For any standards or formats adopted by other organizations (e.g. for-profit businesses), there would be no restriction on the formats they choose to use among themselves.
As to your statement that “enterprise versions of Linux [...] almost ALWAYS ship with proprietary code”. Enterprise versions are expensive. Very expensive for normal citizens. Therefore, you should not look at the enterprise versions of Linux but the completely open versions that could be had for free or a small fee. In the fear for being sued over usage of patents, several of the common versions (OpenSUSE and Fedora come to mind) are limited in their standard installation with respect to, e.g., multimedia software (e.g. playing or creating MP3 files). You have to go to external repositories to install these multimedia components and even then, it’s unclear whether this is legal or not.
This is a good example of why the statement given by the EIF on patents is so important. By indicating right from the start that standards used by the government should not carry this uncertainty, anybody in the industry (both open source and proprietary software developers) can implement this standard. This, in it’s turn, greatly promotes a free and competitive software market. And I think this is something your organization can only applaud.
Eric,
Thanks for your thoughtful comment. We can all agree that governments should strive to be as accessible as possible (although there are legitimate arguments about what types of government information should be made freely available given legitimate privacy concerns, but that is a separate matter). Let me try to address your comments one by one:
First let me note that this policy is NOT just about file formats. That is a common misconception and also one of the reasons why this policy could have lots of unintended consequences.
On the royalty-free requirement: Your basic premise is that people shouldn’t be forced to pay for software to engage with government. It is a noble goal, but it ignores the fact that citizens are always forced to pay in order to interact with governments. Even if the software is all “Free as in Beer,” they still need to pay for the computer to use it on and the internet service to communicate with. If you suggest that the citizens can use public internet terminals, then there is an awfully good chance those terminals will have the necessary software on them. In fact, SAAS providers have also proven willing to integrate standards that are royalty bearing into free services, so that is a possibility too. In the end, it is unlikely that any RAND license fees on open standards would cost enough to create a financial burden any greater than those we already impose on them to use the Internet for communications.
On the issue of creating an impediment to software developers, I would have to disagree as well. Small firms implement RAND-based standards (and license proprietary standards) all the time. It is rarely a significant poses a significant cost to the end consumer. On the open source side, it is very easy for open source companies to implement proprietary as well as RAND open standards as discussed before. You’re right that enterprise server versions of the Linux are very expensive, but Linspire provides a great desktop distro that includes many proprietary and nonGPL programs to ensure interop at a very small price. In reality, the cost of the RAND license to the end consumer will likely be far less than the cost the Internet connection that is required to use it.
To summarize, accepting RAND open standards (as standards bodies and governments have accepted forever) create no real impediment to competition in the software sector. They are “reasonable and non discriminatory” fees. They also provide no bigger hurdle to citizens than other related costs and in case of things like DVDs, more citizens have the equipment to play proprietary DVDs than they do nonproprietary Ogg formatted files.
Here is our position on this to clear things up. Government should explore the use of royalty free open standards whenever available. However, it should not be a litmus test or the only factor in the decision. If a standard has been agreed up on by industry players and has been adopted worldwide by companies and citizens (say the DVD codec), should the European government refuse to use it and try to create an alternative format??? That doesn’t seem like reasonable policy or a useful way to spend limited government resources to me. Am I missing something?