Matusow's Blog : Open Specification Promise

A Simplified Discussion of “Open Standards” – Welcome to the Warehouse

jasonmatusow — Tue, 04 Aug 2009 00:33:14 GMT

My last post has driven some great discussion around what an open standard means. Heck, I think I was writing about “open” issues back in March of 2005. One could almost say it has become thematic.

Given the comments from my last post, I thought it would be useful to write out the analogy that was used to help my over-ripe grapefruit of a brain to understand the basics on standards. That way we’re all talking about the same thing. Here is how a very, very, very smart lady (thanks Michele) made a complex system into a simple analogy (paraphrasing of course).

The Warehouse

Think of standards organizations like a warehouse. There is a front door, interior workspace with many tables distributed around, and a loading doc at the back. People who want to work in this warehouse walk in the front door and find the table that interests them. Some of the people walking in the door have buckets full of ideas. Those buckets get brought in and dumped on the appropriate table and the people at that table use the contents of the bucket, plus their own ideas, as the basis for a discussion that ultimately ends up in a specification…a stack of paper that describes a technology. Once the spec is ready, it gets sent to the loading doc where anyone can drive up to the loading doc, pick up the spec and drive away to go build the technology (finally resulting in some software).

In this picture, the “warehouse” is really a legal framework that protects all parties in the process. The people walking in the door who want to participate, the people bringing contributions, the parties involved with the creation of the spec, and the people coming to the loading doc who want to implement. The framework creates a trust model that encourages contribution, participation, implementation, and long-term improvement of the specification. The warehouse allows competitors to work together and to feel comfortable that they are protected from the ideas shared being used against them, enables them to work together without running afoul of anti-trust laws, and creates the mechanisms of trust that encourage the implementers to implement as they are dealt within a uniform and consistent way by the contributors.

Every warehouse is different (lots and lots of standards orgs out there), some have strict rules, some more lax. But they all (in theory) have rules and restrictions designed to foster an environment where standards work happens.

Yeah?–So What

The point I was making in my last post is that this idealized “warehouse” works best when in balance. People seem to be very wrapped up in the discussion of royalties, but they make up only a part of this discussion. The analogy I used in my last post about limitation of scope is critical to this. The same is true for defensive suspension.

To dig deeper on that point, (I’m not a lawyer), my understanding of defensive suspension is that it creates a disincentive for litigation. Which is a good thing. But it is also a mechanism where the contributor (if sued) can revoke rights to the covered IP. So that breaks the idea of “open standards” = no limitations on IP.

The warehouse will work best when it is full of people, rather than when the front door is only opened to a few people. As I said in my last post, I think Rick was right to argue for the fact that standards orgs are better off with more participation. More people around the table, more interested parties willing to put their resources toward working on interoperability. More opportunities for people to go build great solutions thus leading to those standards having greater marketplace relevance.

The “so what” here is that the structure of the warehouses we all work in matter. If you only look through the lens of implementers, the system becomes hostile to contributors. If you only look through the lens of contributors, you end up with standards that no one uses, and the implementers will find other ways to solve their problems (probably with a good deal less interoperability). Microsoft is both contributor to and implementer of hundreds and hundreds of standards. The same is true for many software firms. It is better for everyone if both sides are considered when thinking about the future of standardization.

For the past 8 years the entire software industry has been moving to the middle on the hybridization of business and development models as everyone continues to look for the…dare I say…balance…between core assets, complementary assets, services, and the advantages of community. One of the more interesting (for industry wonks like me) by-products of that process is a clash between traditional standardization models and the modern meme of collaborative development.

Balance of Contributors & Implementers: A Blog Answer to Rick Jelliffe’s Post

jasonmatusow — Sun, 02 Aug 2009 14:00:44 GMT

I love reading smart people’s blogs – and Rick is definitely among that group. I have really enjoyed reading his post, “Standards media formats and licensing: JPEG vs. MPEG” and then the associated post, “Balance of Interest.” In the first one, Rick referred to my recent post “Contribution, Collaboration, & Implementation” and has spurred me to comment further.

First let me say that I agree with Rick’s essential premise that having greater representation in standards working groups is a good thing. There is no question that vendors tend to dominate most working groups quite simply because standards are an expensive game. Between travel costs and time that equates to opportunity costs, staying engaged with a standards working group can be…well…costly. In fact there are some interesting solutions out there (if I’m not mistaken, the Australia national standards body had a fund set up from the coal industry that they manage in a neutral way for ISO participation by smaller interested parties).

If you look at many orgs, the rules are structured so that the more participants one org may send, the more votes they have. Of course, the rules may say that the individual is a member rather than their company – but who is paying their travel? their salaries? I’m not saying this in a pejorative fashion – just pointing out the reality of the environment in which many standards are created.

So – onto the point that Rick was questioning me about. He and I are using “balance” in different ways. He is speaking of balance on committees for the standards-setting process. No disagreement from me there. I’m speaking of the balance in respect to the discussion of “open standards” meaning that there should be no IP restrictions in the standard in order for it to be open. This is different.

The vast majority of ICT standards are created based upon some original contribution from a commercial entity. These submissions are done in order to lay the foundation for a standard, and that decision to contribute is based on the idea of getting some return on the contribution. The return is NOT necessarily about royalties. The return may be that a product that includes the standard as part of it will do better in the marketplace because of the broader adoption of the baseline technology as a standard. The return might be in improved interoperability of a given product or service due to the adoption of that standard. There are competitive reasons for contributions – the hope may be to displace a competitor who is using a non-standardized solution. There are many, many reasons. And yes – some seek royalties from the IP included in standards. (For Microsoft, this is very, very rarely the case. There have been a few examples, but we almost never seek royalties from IP in standards. Historically we have primarily earned our money based on packaged products.)

The balance issue I speak of stems from the idea that many in the “open standards” discussion have placed an over-emphasis on implementers of standards – particularly those who want to implement under licensing that conflicts with patent terms. Essentially the simplified argument is that because those licensing models are not compatible with patent terms, then there should be no IP protections in the standards process so those implementers can implement. That is not balance – that is imbalance.

As a society, we want contributions to standards to continue. If you make the standards environment hostile to contributors…they will contribute less. I don’t think anyone who supports the current “open standards” arguments I have suggested above would like the idea that the contributors start relying more on SIGs than formalized orgs. Probably sub-optimal.

I have no problem with the concept of royalty-free patent licensing. In fact, nor does my employer (MS). I favor the idea the the organizations themselves, and their members, have the ability to choose which model makes sense. I think that things like defensive suspension are really important. Those terms have a dampening effect on litigation in a given sphere – that is a good thing. I think that limitation of scope is reasonable. If my contribution is about a given protocol, but it turns out that same technology is also the world’s greatest aphrodisiac (going for the over-the-top example here to make a point), then my royalty-free contribution should reasonably be limited to the protocol. I may well want to keep the super love-stuff (sorry – this is a really tortured analogy) to myself, or release it as a completely proprietary invention. I fundamentally still believe that innovations are opportunities…and that is a good thing (for the inventor and for society). But the “no IP restrictions” concept of “open standards” does away with too much. Out of balance.

Two final points: 1) There is no evidence to date that a GPL product can’t implement a standard with IP restrictions. There are many, many RAND standards implemented in GPL-licensed products today. 2) I need to write more about this, but I am all for the idea of “open standards,” but to me it is all about process and participation. I like Rick’s points about balance on committees, and am actively working on methods with my team to figure out how we can facilitate brining more voices to the table in a neutral context.

Thanks for the thoughtful posts, Rick. I really enjoyed reading them.

A Macro Question About Microsoft And Standards - Oh Yeah - And Is Microsoft Really Committed to Open XML?

jasonmatusow — Wed, 25 Jun 2008 04:35:29 GMT

A colleague of mine, Stuart McKee, sat on a panel recently during a Red Hat event. His comments have drawn some attention, and now some FUD that I really feel needs to be addressed.

First of all, you can see the coverage here, here, here, and here. But the one I most want to comment is this one from ZDNet in the UK.

This really is the case of a tempest in a tea pot that has boiled out into the general discussion. Stuart is a great guy, very bright, and was on a panel that was demanding. Okay, no big deal. However the inaccuracies came about, either through misconstrued comments and/or simply misspeaking on something, they are none-the-less inaccurate.

The first set of stories focused on whether or not MS is still committed to Open XML. It is - it will be - and we have never said anything different from that. Check out Gray Knowlton's blog as he is in the product team, and the group, that is working on this exact issue. We have always advocated choice in the marketplace, and yes - recently we announced that we would support ODF in Office. That does not mean we are stepping away from Open XML in the least. In fact, it is more of a statement of a deeper commitment to XML-based document formats...but that is a discussion for a different day. I'll leave the rest of this discussion to Gray.

The bigger issues come from the ZDNet article. Has Microsoft been working on standards before Open XML? Has the company done work on standards in the past? Will we work on more standards in the future? The answers are yes, yes, and yes.

Every big software producer works on standards as they relate to the products they are producing. There are essentially three scenarios on a given product. First, to implement a standard. Second, to work on a specification that is in process of becoming a standard. Third, to contribute a technology specification to a standards body so that it may become a standard. Microsoft has been doing all three of these across THOUSANDS of standards for decades.

More than eight years ago, a corporate standards organization was formed in the company to help product teams be better participants in standards orgs, to make more strategic decisions about what and where to contribute specifications, and how to deal with the legal issues surrounding standards bodies (there is an entire specialization in the legal field for this kind of work believe it or not).

Currently, the standards organization at Microsoft has more than 25 full-time employees in it and is focused not only on standards, but how the company thinks about interoperability and standards as a whole. What's more, because we are active in more than 150 standards orgs at any one time, and more than 400 overall - we have more than 600 product team and field employees who have been internally certified for standards work (and most of them are active in some committee or other). Our products have supported literally more than 10,000 standards and we have contributed specifications in the areas of development languages, runtimes, networking protocols, systems management, hardware, mobility, document formats, security,...the list goes on.

Anyone close to the industry knows that Sun, Apple, Microsoft, IBM, SAP, Adobe...any of the big players have people sitting in dozens to hundreds of standards organizations at any time. Companies in the hardware space like Toshiba, Sony, Intel, Samsung, Nokia, Phillips, Siemens...they have even greater standards engagement. There is a reason they are known as industry standards...it is because industry is the major factor in behind the investment of engineering resources to work on these issues. There is often close communication/collaboration with academia on standards-setting as well, but the big money and people investments are from industry.

I think it is very important that we take a step back from the self-perpetuating cycle of reporters and bloggers quoting each other to really think through what is being said. There are big differences in opinions about the role of standards, about the future of certain technologies, about the balance between IP protection and "openness," about the types of technologies that should be contributed...and more...but to work under the assumption that Microsoft (or any other big software company) is not committed to long-term investment in standardization is simply incorrect.

IP, RAND, Standards, OSP, ISP - the conversation continues...

jasonmatusow — Sun, 20 Apr 2008 06:14:00 GMT

There have been numerous comments to my last blog posting, as well as a long response from Groklaw. Overall I am encouraged by the conversation as a whole because I think it is touching on some foundational issues.

One of the most challenging aspects to the threads I've been reading in the responses to my post (and I see this in the Groklaw post as well) is that many issues are getting squashed together - and that is the very basis of misunderstanding these issues. One more thing, there are a few people who put some really excellent stuff in the comments to my last blog post, and I recommend them as reading. Andy Updegrove’s comment is probably the most thoughtful and thought provoking of the lot.

I’m going to try to pull out some of the most commonly touched on points and themes. This is long...sorry.

Standards and business models

At the core of the discussion both on this blog and over at Groklaw is the question of how evolving business and licensing models blend with the existing standards world. The fact that Free Software as a concept has been around since the mid-late ‘80s (if I remember correctly) is not particularly germane. It was really only following the massive commercial investments of the early 2000s that we see the growth of some very fundamental issues in software business models. (I know there was significant dev work happening in the 1990s – but this is a meta point about the industry.) The hockey stick effect of OSS adoption and real business momentum has brought about the discussions about RAND, GPL, covenants not to sue, etc.

You will get no argument from me on the point that the adoption of FLOSS in big business and small are changing the face of the software industry. But by the same rational point of view, the growth of the overall software industry continued to be fantastic during that same period of time. My concern has always been with people who argue that because the Free Software approach is predicated on certain beliefs, and that it is being used, that everyone should have the same beliefs. I think it is ultimately unhealthy for the industry for everything to shift to a single approach – we are all better off with choice of business models, choice of licensing models, and choice of solutions to acquire for addressing business needs.

(Caution – major simplification here) Standards bodies are meant to be the place where all parties come together in an environment that is safe and beneficial for all participants. And where even those who were not involved in the process, may utilize the output from the process in a safe and beneficial way. Additionally, it is in all parties’ best interests to see sustained contribution of innovations into the various standards orgs (no matter what model they were developed under). Moreover, it has been considered highly desirable to have both standards and solutions compete in the marketplace which pushes innovation and opportunity even further.

So in the discussion of RAND and GPL and even in the case of Open XML and the OSP – I think it is worth considering how it is that GPL implementations may be done. But I disagree if the argument is that they should be done regardless of any consequences to those who do not share the same model. There are ways of building bridges between models – to me, that is the most important place to start.

Free Software Licensing and RAND

I do not want to belabor this point. Groklaw pointed out, as have many others, that Free Software licensing conflicts with RAND. Not just on royalty terms, but on sub-licensing and limitation of scope as well. This is not the end of the world, nor is it a show-stopper (obviously – Unicode, for example, is shipped in Linux I believe). So the question then becomes how do software producers who have an interest in seeing their technologies broadly adopted (even in Free Software implementations) find ways to bridge the gap. Of course, it would be good to see the Free Software community seek to meet in the middle with all parties – consensus helps (IMHO).

ISO and RAND vs. Terms From Vendors

Throughout the comments people are still saying that the JTC 1 RANDz declaration for Open XML is somehow different than other JTC 1 RANDz declarations. It is not. The same rules apply to all specifications – uniformly. ODF and PDF and others are under the same terms. The assertion that Microsoft could somehow “wield” the JTC 1 declaration in a means different from any other contributor to an ISO spec is false. What’s more, the Open Specification Promise goes significantly further than the JTC 1 RANDz declaration and was created to enable the full spectrum of implementations. (Notice Gnumeric has included Open XML support even though it is GPL licensed.)

Vendor Promises – OSP/ISP/CNS

The vendors who are making promises, or covenants, on their IP separate from the terms directly associated with the specification are doing so to build the bridge I spoke of earlier between the traditional IP-holder structures and the Free Software model. We can (and probably should) debate the merits and drawbacks of each of these, but you should first take a step back and recognize that these promises are a positive step in the direction of enabling the highly divergent licensing models to coexist. It is also important to note that while a few of the big players have taken this step (logical because they are the ones holding more patents or who have more directly involved product interests), it is not common practice in the software industry. There are thousands of rights holders who participate in standards, whose contributions merit great respect, and who wish to maintain control of their intellectual property. For smaller software vendors whose business model is dependent upon the direct commercialization of their software, a mandate to exhaust their patent(s) (forget royalties – there are other issues worth considering here) will act as a disincentive for contribution.

Is the OSP from Microsoft global?

I saw this questioned a bunch of times so I think it is worth clarifying. The OSP is a global promise – to everyone, in every country, and in every line of business or personal interest. People are confusing whether or not there is legal recognition of software patents with the reach of something like the OSP. Those are different concepts and worth separating in your thinking.

Should standards orgs change their policies to accommodate free software?

This is a decision that each and every one of the hundreds of standards organizations who handle software standardization will need to consider. There is absolutely nothing inherently bad about working with it or not – it is a choice based upon the charter and membership of that organization.

In many ways, my answer there is a punt, but it is true. My opinion is that effective bridges need to be found, and done in such a way that things remain fair for all parties. Companies that sell software directly must clear their product for patent royalties and other IP considerations before selling it into the marketplace. The standards bodies have IP policies so that their output too is copasetic in regards to IP before people begin to use the specifications (thus the concern about trolls). There is no reason that orgs commercializing Free Software should not respect other’s IP as well.

I think an interesting way to consider this is to take a logical extreme of the argument – what if every standards body mandated Free Software-compliant (not MPL, not EPL, not MsPL...only GPL) licensing terms. What would happen? Some might say we would be one step closer to nirvana as a race. Some might choose to drop completely out of the standards game and seek individual licensing agreements and/or cross-payments etc. to bring their product to market with interop features. Certainly there would be those who would be forced to fracture their product (again, good/bad...all depends) and ship some components separate from others to stay compliant with the source license. Others might pursue a whole raft of SIGs (special interest groups) that are exclusionary and outside of the standards world. Still others...I could go on. To me, that is not a good outcome. I think we are in a better state if standard bodies continue to serve their purpose and be a place with highly competitive parties can come to the table to collaborate on specifications even if they have diametrically opposed business/licensing models. Ultimately, the marketplace will speak louder than licensing mandates.

Should there be multiple licensing schemes?

This is an extrapolation on my part from what I’ve been reading in the comments of my last blog. I’ll first say that it is a moot question because there are and will be multiple licensing schemes. How many OSI licenses are there? And why is that? Because IP holders all have strong opinions about how their IP should be handled. Secondly, I am always a fan of choice...and am nervous when zealotry begins to suggest that any alternate choice is heresy. This leads to a very scary place.

Conflict of reciprocal licenses – was I speaking correctly?

I spoke of the fact that Linux ships with code that has conflicting licensing agreements. And many people asked me to “prove” it. I’ll let you all tell me if I’m nuts. All reciprocal licenses by definition conflict with each other. The GPL claims to cover the “aggregate” product, the CPL/EPL the “component,” and the MPL the “file.” In all cases though, if there are conflicting terms they all say you can’t distribute. So – is there any mingling of GPL, MPL, EPL code in Linux? How about the rest of the licenses purported to conflict with the GPL?

By the way, this is also true for the point that I made about RAND terms on code shipped with Linux today – how about Unicode for example? Last I check, IEEE 1394 Firewire was similarly licensed. There are literally THOUSANDS of RAND-covered specs out there used commonly in operating systems throughout the industry.

Now, on to the leap of faith on the part of my critics – I did not raise this point to spread FUD about Linux. Why can I say that? Because this point has been raised in public events in the past...and it continues to not matter. So my point is something else entirely.

Legal snags like the ones I mentioned only matter if someone presses it in a court case. No one can say if these issues will ever become an issue but that has never stopped a single person from using Linux. So, when people then say that the MS OSP, or IBM’s ISP, or RAND terms, or whatever means that Free Software developers can’t develop something, I find it hard to take seriously when the intent, and all of the materials surrounding these actions speak of building bridges and enabling...not shutting down or threatening. Those same developers are willing to take those exact same issues as no concern on one hand and then scream foul on the other. It’s worth thinking about a bit.

Ending this long post...

Too much here already – thanks again folks for the good comments. We are at an inflection point in the industry...and that is exciting to be a part of. If you are going to comment – let’s shy away from the zealotry, and stay focused on lucid arguments.

More Open XML Discussion - more misunderstandings about standards and IP

jasonmatusow — Wed, 16 Apr 2008 08:28:58 GMT

There is much to talk about based on what has been happening in the discussion of Open XML these days. First, ISO has posted a FAQ about 29500. The approach taken in the FAQ is both direct and simplified. The rabbit hole on standards goes deep indeed, and any such FAQ must simplify to make it digestible for the layman reader.

An example of someone misunderstanding standards and IP issues would be Groklaw in this post. The post displays many of the misunderstandings about IP in standards that I see echoed around the blogosphere.

I will say that the author of the Groklaw post got it right when they wrote that, "Patents and Linux and FOSS don't mix." But the author jumps off the tracks completely when the observation is followed up with, "So RAND terms in Microsoft's hands are discriminatory."

I think it is important that anyone offing an opinion on the OSP should first ask him or herself a question: do I really understand IP licensing in the world of standards? I'd suggest that the author the Groklaw post didn't ask this question up front and would benefit from some boning up - it would help get the story straight.

Here are some things I think may be helpful in this discussion. By the way, my answer to the question I just posed in the previous paragraph is, "sort of."

1) RAND (reasonable and non-discriminatory) did not originate with software. It is a commonly used acronym but has different meanings in different standards bodies. It has many parts - more than just patent rights, more than royalty terms...my understanding is that it is a framework designed to provide both incentive for contribution and implementation. The reason I agree with the statement about patents and Free Software not mixing is that there have been terms written into GPL licenses that explicitly conflict with software patents. Okay, that is the choice of the authors and users of those licenses. But the fact that those licenses were written does not mean that the RAND terms that have evolved over the past century of standardization are suddenly invalid.

2) The ISO/IEC JTC 1 patent policy is applied uniformly to all standards in the ISO/IEC JTC 1 arena. The idea that the RAND declaration regarding Open XML is any different than a RAND declaration for ODF or for any other ISO Standard (such as...oh I don't know...how about PDF just for fun. Remember the huge list of patents that Adobe used to put on the welcome screen of the Acrobat reader alone?). The terms provided for the Microsoft patents in Open XML are legally irrevocable. They are global. Since they are broader than the RAND declaration for JTC 1, the attempt at FUD by the Groklaw post should be recognized for what it is...FUD.

3) The various covenants not to sue or similar promises from a few (mind you, this is hardly common practice in the software industry today) of the big vendors are all born of the same desire - to help bridge the gap between Free Software and software patents used in specifications. Forget the philosophical and ontological debates about software patenting for a moment and just deal with the fact that the law in a number of countries recognizes software patents as valid IP today. And, judging by the continued pace of patent filings, it would seem that more than a few people out there are seeking to use software patents as a legitimate means for protecting their innovations. Heck, even some of the long-time Free Software advocates are looking at the positive implications of patenting software as a means to encourage greater innovation. Yet all of these bridges run into some snag or another on various points. So the question that comes to mind is how important are those snags?

Before you get your hair in a knot over this, consider for a moment that all Linux distributions come with code licensed under reciprocal licenses that conflict and (in theory) should block the distribution of the software. Oops, no one seems to care about that snag. Okay, Linux also ships with code that implements standards specifications covered by RAND terms from many parts of the industry that should block the distribution of the software. Oops, no one seems to care about that snag either. Hmmm - I'm wondering if those complaining about the OSP understand the nature of IP terms in standards to begin with, since they don't seem to mind these other, potentially more-serious issues.

4) Many of the terms included in standards IP licensing are there for a rainy day vs. for any practical use. In many cases, IP holders will provide RAND terms that include royalty terms yet never, ever collect a penny in royalties. Why, you might ask? Those terms are there as a defensive mechanism. This is true for the entire concept of defensive suspension and to a large degree the concepts around limitation of scope as well.

5) The concept of "discrimination" in the standards world of RAND is another one where you really need a lawyer to talk you through what it means and/or doesn't mean. I am not a lawyer, but to my understanding it has a great deal more to do with the uniformity of your application of terms to like parties than it has to do with any one party being able to take those terms or not. If there are lawyers reading this thread I'd love the input. Everyone should have the choice of what license they use - but there may be implications for the use of those particular terms as they will interact with the rest of the world.

Concluding this too-long blog post...

As you can see, looking at the real issues on the table about IP and standards is a good thing to do. I am on board with the idea that the established world of IP in standards needs to be looked at carefully in context with the amount of work happening under the concepts of reciprocal licenses. It would be helpful if folks who speak with a loud voice also seek to do so in an educated fashion.

I am a serious proponent of participation in the process - but that needs to be for all parties, large and small alike. But that means keeping an open mind while you are also advocating open standards.