DIS 29500 Meets All JTC 1 IPR Requirements for Approval as an International Standard

The FUD machine from the anti-Open XML camp has been running on the high-octane fuel of the Software Freedom Law Center critique of the Microsoft OSP. While there is absolutely a debate to be had about the role that covenants not to sue or open specification promises play in the broader world of standardization - they are irrelevant in the conversation about the approval of an international specification under the ISO/IEC JTC 1 directives.

Collectively, it is everyone's best interest to have the JTC 1 rules applied in a dispassionate and equal manner for all specifications under consideration for international standardization. The idea that Open XML should be held to a different standard than any other specification is damaging to all JTC 1 business.

Gray Knowlton put out a blog posting where he addresses many of the issues raised in about the OSP - and this is a topic I have addressed in previous blog postings as well. I certainly will write on the topic again - but there is really one one thing to keep in mind if you are part of a national standards body considering the fracas being raise by the anti-Open XML community.

Open XML (DIS 29500) exceeds any ISO/IEC requirement for intellectual property rights for ANY JTC 1 specification. The discussion of covenants not to sue and open specification promises have no standing in relation to the ISO approval or disapproval of DIS 29500. The only measure for consideration within the context of the process is the IPR policy of ISO/IEC. Period. Full stop. End. Arête. Halt. Alles klar?

For the sake of clarity, the reason I say that Microsoft "exceeded" the ISO/IEC JTC 1 IPR policy is that there was a clear intent from Microsoft to promote the adoption of Open XML broadly and mechanism to support that goal was the use of the CNS and OSP options. JTC 1 asks for RAND either with or without royalties. Ecma 376 complies with that requirement. For the broader community interested in developing Open XML implementations, they want to know that they can do so without concern of Microsoft pursuing them for patent infringement. Any implementer must consider what licenses they are using for their implementation and how that interacts with the terms covering the specification. This is true for all specifications being implemented by someone.

So to recap - if you are just generally wondering about DIS 29500 – the ISO/IEC IPR requirement for a minimum set of IP terms being Reasonable and Non-Discriminatory (and specifically in this case without royalties) - that has been met. There is NO OTHER CHOICE for any contributor to the ISO/IEC process – one of the boxes on the form must be checked. ODF is under THE EXACT SAME REQUIREMENTS. What happens beyond that in terms of OSP, CNS, etc. is not relevant to the approval of an ISO/IEC specification because national bodies should be under the the guidance of a consistent set of rules when they consider their position on any specification going through the process.

Comments (10)
  1. Andre says:

    "Open XML (DIS 29500) exceeds any ISO/IEC requirement for intellectual property rights for ANY JTC 1 specification."

    Maybe true, but mind the competition! An international standard for an issue as ubiquitous as document formats needs to be royalty-free. Anything else is just unacceptable, like a corporate tax, a technical barrier to trade.

    When you read the ISO/IEC rules on patents carefully it is not that ISO endorses patented standards. It is rather that they say 2.14.1:

    " If, in exceptional situations, technical reasons justify such a step, there is no objection in principle to preparing an International Standard in terms which include the use of items covered by patent rights – defined as patents, utility models and other statutory rights based on inventions, including any published applications for any of the foregoing – even if the terms of the standard are such that there are no alternative means of compliance."

    Where are these "exceptional situations" and the "technical reasons" for your trivial patents on OOXML?

    ODF is sufficiently indemnified by SUN. Why didn’t your company just sign the very same patent pledge? There is no point in insisting on ISO/IEC minimum requirements of RAND licensing because it does not apply in this case. No government user can accept a RAND standard for this purpose.

  2. carlos says:

    Jason, if this people did an analysis and found this issue… respect this and not throw more PR over it.

    Each NB will have what to do ( they are all grown up people ). You seems to underestimate them.  

    Probably each NB have done his own analysis on this question and has settled up an opinion.

    So please … calm down and let people work …  you have disturbed too much all this process …

    After all, isn’t ECMA who submitted this specification for fast-tracking and you plan to release control over it .. so let the world work and review this in peace.

    Thank you


  3. jasonmatusow says:

    Andre – thanks for the comment. Open XML is under royalty free terms. It is under far more liberal terms than the JTC 1 directives require. I really don’t understand the point you are trying to make. Contributed standards often have patents associated with them, and participating parties are expected to disclose the patents if they know about them, and JTC 1 allows for this. In this case, MS declared there are patents and has put them under royalty free terms for everyone to use in any development model. I am not going to argue over your interpretation of patents and their import or not. The point is that the IPR requirments for a JTC 1 specification have been met and, in fact, exceeded in requirment for adoption as a standard. The OSP and Suns CNS are very similar instruments. In fact Microsoft’s contribution is under both a CNS and the OSP – at your discretion which to use.

    As for your absolutely bizarre statement about governments use of RAND standards…I think you should do some homework on standards before you make a statement like that. i would venture to guess that the VAST MAJORITY of standards employed by governments are under RAND terms.

    Carlos – thank you for the comment. I have no problem with the fact that someone did an analysis of the OSP and formed an opinion. I do have a problem with the anti-Open XML community trying to use this to create fear, uncertainty, and doubt about the specification when there the analysis has no bearing on the JTC 1 process. The rules are meant to be applied fairly and evenly – and those that are sending letters (I have seen them myself) using this as a reason to vote no are purposly trying to create confusion. Believe me, IBM understands the difference and are choosing to ignore it into order to create confusion.

    Yes, the NBs will do their own thinking on it, and everyone (even little old me) has the right to express an opinion on it.



  4. zoobab says:

    Jason, can you send me a copy of the RAND-Z license?

    I am interested to see what kind of restrictions people like the GPL implementors who does not trust the OSP for penny will able to get screwed.

    My email address is zoobab@gmail.com.

    PS: I already asked many of your colleagues, no one was able to send me a copy of this RAND-Z licence.

  5. Andre says:

    Thanks for the comment.

    It is true that RAND is quite common. I would like to know in particular when Microsoft will make available its RAND-Z license for OpenXML.  

  6. jasonmatusow says:

    Andre – Open XML has been under royalty free terms for more approximately 2 years. From Microsoft it is not under RAND terms, it is under both a Covenant Not To Sue and an Open Specification promise (the CNS came first…we upgraded to the OSP based on feedback at the time but if you like the CNS more you may use those terms).

    There are no royalties on Open XML. The OSP was recieved very well at the time of its release – including from Free Software folks – and is now already being implemented in GPL projects and on GPL-based platforms.

    You are looking for controversy that is simply not there. If you want to have a detailed conversation about the OSP – there are always nuances to every discussion – but the real truth is that Microsoft has made a promise not to sue for use of the patents. End of story.


  7. jasonmatusow says:

    Zoobab – I am about to get on a plane from Munich to Sydney (urg) and will dig up what I can later this weekend.


  8. Andre says:

    The reason why we ask for the rand-z license is that for many people it is found to be more reliable than an OSP or CNS and could be a valid alternative. It was announced that a RAND-Z license is made available for interested parties by Microsoft but no one published it. A Rand-Z license could please many critics that are still sceptics of the OSP.

    I talked to potential third party implementers of the format before and they lacked trust in the OSP. Industry players were reluctant to say it was reliable. I don’t know if its applicable under our domestic law as it is so deeply tied to the US legal system. Its a real pity that we have no "industry standard" for patent indemnification yet. the truth is that no one can be sure about worldwide applicability or whether the text fails to address legal scenarios nobody even thought about before.

    If Microsoft is honest about the OSP it should consult as widely as possible for version 2.0 to make sure that all possible issues are addressed. It is always a surprise to me how different private law is on an international scale. For instance the European Union consulted widely for their European Union Public license and although we have EU copyright harmonized our legal systems and traditions differ very much. A consultation process is very helpful to improve such a text. We also see from DIS29500, specifically DIS29500.org and the OOXML review process how helpful crowdsourcing can be.

    Btw: Munich’s mayor wrote a letter against Open XML adoption.

  9. Today, the US INCITS voted to approve the Open Office XML document format specification for fast-track

  10. hAl says:

    An interesting quote on the SFLC article is

    "So any code written in reliance on the OSP is covered by the promise only so long as it is not copied into a program that does something other than implement the specification"

    Strangely they do not tink the same of the Sun covenant not to sue allthough the Sun covenant explicitly states:

    "it will not seek to enforce any of its enforceable U.S. or foreign patents against any implementation of the Open Document Format"

    So apparently the allthough the limitations in the Sun CNS on only covering OpenDocument implementation is as explicit as can be and similar to the limitations of the OSP the SFLC has stated that :

    "Sun’s terms are not in conflict with Section 7 of the Free Software Foundation’s GNU General Public License, and are not otherwise incompatible with the GPL"

    It must be a strange world that mr Moglen lives on. The legal interpretations of his organization apparently vary with the winds. Or mayby with who pays the bills ?

Comments are closed.

Skip to main content