Material on Standards and IP

by Rick Jelliffe

The licensing of IP for standards has four aspects: what the (case and statute) law says, what the standards bodies require, what the IP owner grants, and how the developer (adopter) is acting. Standards themselves never seem to have useful information about patent IP, and even their copyright boilerplate needs to be checked against licenses given by the copyright holder: W3C and ISO don't like you copying their standards, Ecma does, for example.


For an introduction to the legal aspects, see, which is by a lawyer for OASIS. The Dell case is pertinent.

For an introduction to the standards body aspects, see Standards Law, which is by a lawyer for Microsoft. It has a reference to the ISO requirements. For the boutique standards bodies: OASIS, Ecma, W3C

For examples of the kind of grants that companies make see
Microsoft Open Specification Promise, IBM Open Source Portal, Sun's OpenDocument Patent Statement. Adobe has not put their equivalent online if it has been finalized, as far as I can see. (Microsoft also has a "Covenant not to sue", however this seems to have disappear from its website in a rearrangement of links. They need to get it put back online.)

So what does the user have to do with it? Some licenses provide particular conditions relating to private or not-for-sale use: the GNU licenses for example. Other times licenses are revoked if you try to sue the IP owner: these defensive patents are bargaining chips in legal wrangling.

One key term to understand is RAND: Reasonable and Non-Discriminatory Licensing. It is pretty much the bottom line for standards organizations. However, RAND licenses are controversial, and in the views of many of us, something that should be avoided by modern standards bodies in the age of Open Source and Free Software which, like standards, have strong counter-monopolistic and even communitarian aspects.

Another concept to understand is the Open Standard. Not all standards from standards organizations are Open Standards under anyone's definition, especially older standards and standards which involve semi-scientific research and development (compression patents, for example) where the IP holder would only license a vital technology under RAND or not at all. (There is some creep on what an Open Standard is, to conflate it with Open Source or free implementations.)

And it should go without saying that someone cannot grant a license to IP they do not themselves hold. So all covenants and licenses only extend as far as the material in question. This is important for extensible formats such as ODF and Open XML, because the ZIP container allows any kind of media or binary file.

See the IBM material for a definition of Necessary Claims and Required Portions.


Rick Jelliffe
2007-08-14 03:42:23
There is an instructive example of the way things work on the W3C REX patent page. A member of the committee was not willing to license its relevant patents. So the work was halted. Another effort has been started in a similar area, presumably to workaround the details of the patent (or perhaps to demonstrate to the patent holder that their IP is not necessary to any solution and therefore better shared.) The REX work may start up again in the future, it the patent holder agrees.

2007-08-14 06:41:55
"Some licenses apply to private or not-for-sale use: the GNU licenses for example."

What did you mean by that? The gpl does not prevent you from charging, nor does it prevent you modifying the software and no distributing those modifications

2007-08-14 07:30:38
Anonymous: Sorry for my clumsy construction: I was not trying to characterize or summarize the GPL, nor limit it. I didn't mean "applies (only)" but "applies (also)", just point out two topics of interest. I'll fix the sentence shortly. From the Q&A:

Does the GPL require that source code of modified versions be posted to the public?

The GPL does not require you to release your modified version. You are free to make modifications and use them privately, without ever releasing them....

Does the GPL allow me to require that anyone who receives the software must pay me a fee and/or notify me?

No. In fact, a requirement like that would make the program non-free...

(Of course, as always, I am confident that smart readers will read the license and the whole Q&A for the full story.)

Rick Jelliffe
2007-08-14 07:54:51
Update: Hat tip to Bob Sutor for link to PeerToPatent, which is a website for prior art for patents.

In the document processing world, we are fortunate to be old in computing terms: the basic techniques of typesetting were done in the 60s and 70s and any patents long expired. Though some recent patents still crop up: sub-pixel rendering for example.

2007-08-14 23:19:43
Do not be afraid of standarts. Some definitions of the term "open standard" permit patent holders to impose "reasonable and non-discriminatory" royalty fees and other licensing terms on implementers and/or users of the standard.
Rick Jelliffe
2007-08-18 19:44:32
RTF to XML: My understanding is that the ITU definition of "open standard" (which is the only one that allows RAND) is out-of-step with all the others. We need to have a term for a standard that incurs no royalties etc, and "open standard" is it: the ITU should change their definition.
2007-08-22 06:29:33
This is a good post, Rick. To make progress on establishing open systems, users and customers must clearly understand that openness is a condition of the agreement for participating in the process and contributing to the content of the standard. It is not a quality of the standard itself nor does such agreement affect the technical quality except insofar as it blocks or enables the opportunity for better technical content to be submitted.