Interoperability Specifications Pledge

Print Friendly

This morning IBM announced a broad patent pledge that covers over 150 of the core software interoperability standards. These standards had all been previously given royalty free licenses, but from a variety of sources at a variety of times. This action simplifies and makes more consistent the intellectual property situation around these important infrastructure specifications. This pledge applies to all implementors of the standards on all platforms.

This is part of our continuing evolution and program to help spur open innovation and collaboration, as well as moving to true interoperability via the widespread use of open standards. Previous examples include our Linux kernel pledge in 2004, the “500 Patent Pledge for Open Source” and the Healthcare/Education standards pledge in 2005, and the co-founding of the Open Invention Network.

Personally, I believe such positive, constructive actions regarding intellectual property are the preferred routes to accelerate the shift to better products and services for customers via open architectures.

The press release has more information. The pledge and list of covered standards is now online.

We welcome others to join us in this pledge to not assert their necessary patents for these important standards.

As a complete aside, it’s fascinating to see such a huge, comprehensive list of standards and to realize just how far we’ve come as an industry.

Press and Blogs


  1. Nice move, Bob, congratulations.

    One question, though. Why did you frame this in terms of “necessary claims”? That leaves a developer wondering if they have done the right things in their code to avoid a patent claim against them. It leaves them (potentially) to research IBM”s patent portfolio in order to get security that the approach they have take is unavoidably necessary. And it leaves them fearful (reasonably or not) of action against them if they decide to partner with a company or project of which IBM does not approve (the same way the police can always find a law you broke if they want to nail you).

    When we designed Sun’s equivalent patent pledges, we decided we would not limit the waiver to “essential claims” for just this reason, so that open source developers could implement with confidence. Presumably you decided against this. Can you explain why, please?

  2. Thanks, Simon, appreciate it. To sum up a response: different companies, different attorneys, different patent portfolios, and different ideas of how to scope things to the standards at hand.

  3. Nice move! Nice long list too. I’m particularly pleased to see that it is not locked to an open-source model or platform.

  4. All the legal stuff aside, I hope people find the list itself to be a good resource. It doesn’t just mention the names of the specs, it includes links to them as well. That can be hard to find in an organized manner.

  5. Bob, congratulations. This move by your company is a great step forward ensuring that open standards achieve what they are created for. I hope as you do that others will join IBM in this pledge to not assert their necessary patents for these important standards.

    On a sad note, SAP does not seem to be on the same page as has become evident in its action to file a large number of patents on related to UN/CEFACT’s ebXML Core Components Technical Specification (CCTS). Not only has SAP been the main contributor but also controlled the content and direction of the standard by provided the project leader and editor. Over the past two weeks a number of TMG members have asked SAP to provide clarification on their intentions for filing those patents and what they consider an infringement, only to be told that this is standard business practice. May this latest action by IBM open the eyes of SAP that change has taken place in the Open Standards World.

    For more information on CEFACT’s Patent Crisis go to:

  6. Thanks, Bob. I am genuinely interested since I believe this should become a routine requirement of the standards process for all participants. Paul Krill may think I’m criticising but I’m not – I’m just asking within the framework of the explanation of patent covenants I blogged in May[1] when we issued our covenant on OpenID (and I’m a bit bemused by the call for Sun to imitate IBM imitating IBM!)

    And yes, your impressive list will prove useful. Do you intend to lodge an IP waiver with each standards group (whoch, again, has been our approach)? That way you’ll not need to update the list each time a new version is produced.


  7. Good question, Simon, and we’re looking at how we want to handle that. There seem to be some interesting legal questions about how to handle it most easily given the structure of the entire pledge. Let’s discuss it next time we talk (which I hope is soon).

  8. The trap is: what do they really give something away? While it is very useful for market confidence do they really have patents on these standards? The standards they speak about should be open standards, that is license and patent-free.

    simon: Is there a “standard” legal document. I find at Patent Commons various solutions but I cannot say what is best and you probably cannot copy the wording of pledges from others.

  9. I think the proposition is that it’s important for the future of IBM’s business that people should collaborate, and the collaboration should be celebrated.

    That’s what flipped IBM out of the OS/2 era and into the Linux era; and is flipping IBM out of the Lotus SmartSuite era and into the era.

    It’s maybe a little harsh for the developers, marketers, and sellers of the products which die; but there are plenty of new seedlings growing in the ashes of the past.

  10. Haren Visavadia

    A long of list patents does not impress me (especially from a developer perspective)! I consider IBM is now attempting subvert Open Standards and extends its patent game into it.

    The patent license does not live up to the real spirit and purpose of Open Standards.

  11. Haren, with all due respect, we had already given royalty free licenses to all of these, as we clearly said. This simply makes things easier. Please feel free to justify your statements, but do that, not just assert things.

  12. @Andre: No, there’s no standard for a patent pledge that I know of. I would like to see a group like FSF or OSI assert what terms in a patent pledge create freedom for developers in open source communities.

    I did create a partial list of desirable attributes in my blog posting[1] about patent covenants, but it’s a subject that would clearly benefit from discussion. I’m sure Bob had to do a lot of talking to get his lawyers to do this (we certainly had to at Sun) and having independent external guidance would probably help.


  13. Sure it’s royalty-free license however some strings attached (see Simons Pipps’s first comment).

    You know Software can have bugs, my main concern here is a bug discovered in the Software of a Covered Implementation of the Open Standard that could trigger “Necessary Claims” issue (specifically “compliant implementation”).

    With the following statement in the patent license to me, indicates an offensive patent strategy :

    “However, this covenant will become void, and IBM reserves the right to assert its Necessary Claims against you, if you (or anyone acting in concert with you) assert any Necessary Claims against any Covered Implementations of IBM or of any third party.”

    This puts any developer implementing any of those Open Standards at a patent risk.

    I prefer if the patent license terminates upon asserting a Patent Claim against IBM and/or any third party that implements the said Open Standard. This is to encourage patent peace for any legal or natural person using, implementing and/or developmenting any Open Standard.

    If possible, please consider revising the patent license taking account these issues.

  14. That actually is considered a defensive term, because it stops OTHER people from being offensive and running around suing people. You have MORE general protection on the standard because of this term, not less.

  15. Patents are supposed to accelerate the commercialisation of inventions. They are supposed to be about arranging ‘who gets the money’ from a successfully-commercialised invention.

    Of course, IBM likes it to be IBM.

    But they’re not supposed to be about stopping people (or corporations) from making the inventions or bringing them to market.

    I have no idea whether a patent court in any country would ever ask a developer to hand over to any other party more money than they had made by selling whatever they had developed. But that is the $64,000,000 question.

  16. Bob: You say
    > That actually is considered a defensive term,

    In fact, the patent peace you assert here could be viewed as imbalanced – I explained this in connection with Microsoft’s covenant language too[1]. Since you only grant rights on “neccesary claims” (i.e. a limited set that is at IBM’s discretion) but revoke those rights in the case of any action by another party (even though you’re using the “neccessary claims” language it in fact is a much wider scope in the case of other parties), IBM is gaining broad protection at the expense of discretionary grants. This is another problem with the “neccessary claims” formulation.


  17. Well, how do you deal in the absolutely most general case with use of “unnecessary claims”? That is, things being used out of left field that are not germane but are nevertheless pulled in?

  18. I suggest that you just have to covenant not to assert any patent claims against code which is a bona fides implementation of the standard. After all, that is what you are demanding of others with the non-aggression requirement.

  19. That might be an option, but I also can see different intellectual property attorneys having trouble with making that rigorous.

  20. Hang on !

    It’s the executives and managers who decide what they want (i.e. what they think best for the shareholders, over some appropriate timescale) ; the attorneys just codify the decision on paper. And what’s best for IBM probably isn’t what’s best for Microsoft.

    But yes, deciding to share in a controlled way, and then making the implementation match the decision, is tricky.

    You can’t take back something you have given away, and you mustn’t give away something someone else owns.

  21. About these intellectual property attorneys. I know that Sam P pays their salary; but I believe they are United States attorneys.

    Does that mean they owe allegiance to Sam P and his shareholders, or do they owe allegiance to the flag and the republic for which it stands ?

    Not that I mind which. But I am curious.

  22. Just as we have research employees around the world, we have attorneys around the world as well.

  23. Bob,
    How does IBM feel about OpenID, was there any particular reason for not including it in your lengthy pledge?

    It seems some in the OpenID community are asking the question already.
    “IBM is the latest to make a unilateral non-assertion promise about intellectual property, called the Interoperability Specifications Pledge. … IBM’s list of specifications is by far the longest. Which is good. What is surprising is that it does not list OpenID. Some people are wondering why … intentional? … forgot? … not deemed important enough?”

    I’ve got to thank IBM (and Sun) for taking steps towards making standards open and accessible.

  24. There are a few that arguably should have been on the list that didn’t make it in this round, but don’t ascribe any great devious plot to that. We might have just missed something. Please do let us know what you think we should have included and we’ll see what we can do moving forward. Do remember that this initial list was of standards where we had previously given an explicit royalty free license. In any case, keep those cards and letters coming!

  25. Haren Visavadia

    I have noticed you’ve updated the patent license.

Comments are closed