Standards: We need to talk

Print Friendly

Last Friday I had the pleasure of speaking at Tsinghua University in Beijing at an IBM-sponsored conference attended by over 200 government officials, business leaders, and academicians to discuss the confluence of three important topics: open standards, innovation, and intellectual property policy. Below you’ll find the text of my address, though I did ad lib a bit and the recorded version is thus somewhat different.

Just to set the scene, there was an impressive group of officials in attendance from the State Intellectual Property Office of China, State Owned Assets Supervision and Administration Commission of China, Ministry of Science and Technology; and Chief Judges from the Intellectual Property Rights Trial Chamber of China’s Supreme People’s Court and Beijing High Court.

My colleagues were well represented there, too: John E. Kelly, head of IBM Research; Henry Chow, chairman of IBM China; and Dave Kappos, VP & assistant general counsel; and many others.

Before we get into my remarks there, let me say that the conversation and consensus in Beijing is made all the more poignant by Andy Updegrove’s report of how the events leading to the JTC1 vote on OOXML have since largely paralyzed one of the most important working groups in IT standards, SC 34.

Here’s what I said in Beijing, but see what follows below as well.

There are many different kinds of standards, but it is clear that they have greater global importance now than ever before.

In this global environment, it is important to take into account the needs but also the expertise of countries, regions, individuals, organizations, and companies around the world.

Hence the great need for open standards.

Let me ask a simple question: “What is our goal?”

In my opinion, it is this:

The highest quality standards, developed in an open, collaborative, and transparent manner, designed to create a strong foundation on which we all can innovate.

I want to emphasize the word “quality.” By this I mean

  • Quality of technology
  • Quality of standards process
  • Quality of membership and intellectual property rules

In my opinion, there is too much variation in standards organizations in all of these areas.

When looked at in total across the world, it is difficult for many users and consumers to know which are the best standards and which are the best standards setting organizations (SSOs).

If I may make an analogy, a five star restaurant is among the best, while a one star restaurant is among the worst.

What standards and SSOs deserve five stars? Which get only one star and are therefore to be avoided? How do we even measure this?

As we think today about standards and intellectual property, I would like you to consider different practices in different parts of the industry.

Software standards are different from hardware standards in many ways today, not just in technology, but also intellectual property-wise.

For example, in the area of software interoperability, IBM pledged several months ago that we would not assert our necessary patent claims on over 150 standards for areas such as web services and document formats.

The recent attempt to create an international standard from the ECMA OOXML specification exposed many problems with the standards process. Issues around quality, single vendor control, and intellectual property became clear.

What can we do to improve standards around the world?

We can start by having a dialogue, which is why I am so pleased to be participating today.

As Henry Chow (Chairman, IBM China) suggested earlier, we can and should use the discussion tools of the modern web and internet to collaboratively identify and start solving some of the most serious issues facing the creation and use of standards today.

If there is interest in this proposal, IBM would be pleased to help lead this and to actively participate with standards leaders and experts from China, other countries, SSOs, and other organizations.

As I conclude, let me remind you of the goal as I previously stated it:

The highest quality standards, developed in an open, collaborative, and transparent manner, designed to create a strong foundation on which we all can innovate.

Standards, especially open standards, matter to all of us, and will only matter more in the future.

Let’s continue to collaborate to ensure that the processes, policies, and technologies are of the highest possible quality, to the benefit of us all.

Let’s do this before the problems we have already identified get even worse.

Thank you.

There were so many things fundamentally wrong with how OOXML entered the standards process, was propped up and pushed forward, and just never seemed to go away even in the face of withering analysis and criticism, that it is now clear that things have to change. Rather than dwell on that, though, let’s use it as an example that we should ensure never gets repeated.

I’ve previously addressed the idea of measuring the quality of standards and will continue to look at that as well as the larger question of ranking standards organizations. Indeed, I think there is a fundamental confusion among many members of the public about what even constitutes a bona fide standards organization. As I’ve said before, not all standards are created equally and not all standards organizations are created equally either. The experience of the last few months has made this painfully clear.

But more than anything I can say, I think we need a fundamental discussion among those committed to improving and perhaps reforming the various standards processes around the world.

There is much that is right when we look across all the efforts at the consortia, local, regional, governmental, and international standards levels: some standards groups and bodies work extremely well. More than simply being congratulated, they are to be identified and emulated. There are best practices to be collected and more broadly practiced.

However, there are also inconsistencies, unbalanced influence, and a glaring lack of transparency in some other important efforts. While preserving the necessary autonomy of organizations, we can nevertheless work together to identify the models by which they can improve the service they provide to their constituents.

I will soon return to the proposal that I’ve touched on here, the start of a focused discussion by enlightened standards stakeholders to fundamentally start doing more of what we are doing right and to fix what needs improving. In the meanwhile, I welcome your comments on the topics for discussion and the goals of that effort.

Quality of standards process, quality of standards policy, quality of standards technology.

I think it’s time.


42 Comments

  1. I think that one thing that must happen is for interested users / consumers to be involved in the process somewhere, whether it is creation of a standard, approval of a standard, revision of a standard, or in testing / establishing a test suite for a standard (and every standard needs a test suite and an openly-licensed [that is, issued under an OSI approved license] reference implementation). It is one thing to have a standard that everyone in one room likes, but quite another to get buy-in from the end-users that actually determine how successful standards-conforming products will be in the market.

    I think that one of the issues with Office Not-so-open XML is that it meets the need of one company to remain in control of its users’ data (their business model depends upon that, really), but it quite ignores the needs of end users to be able to reliably open, process, and save their documents with any major product on any major operating system, as well as organizations’ need to extract or import data from those documents into a variety of home-grown and commercially-purchased systems. Above all, it ignores the necessity of government agencies to have access to the data stored in said formats long after MSFT has lost interest in OOXML. This means that the formats must be fully-specified and implementable outside of Redmond without any kind of legal backlash.

    Openness in standards, file formats, and protocols goes well beyond the office suite space. This is merely one of the early clashes in a struggle to free our data from captivity to corporate overlords. The standards process needs to reflect the need of end-users for free access to their data or the process and its outputs will be ignored.

  2. The problem with the standards process and OOXML is that the rules were bent and Microsoft threw money at it. I think that concentrating on standards bodies so that they are truly independent is something that has to be looked at. Most standards bodies are just social networks…we all know that social networks can be controlled by an underlying network of key nodes and that is what happened with OOXML (this is why thousands of soldiers will happily fight a ridiculous war on the say so of a few individuals!).

    Can we not enforce strong governing rules that can void votes if those votes are deemed to be breaking the rules.

    One rule that is obvious is … When a standard is voted for then all members must be present …

    Standards bodies must have “clear” rules for putting forward a standard.

  3. “The highest quality standards, developed in an open, collaborative, and transparent manner, designed to create a strong foundation on which we all can innovate.”

    Unless you address the problem of the participation agreement (the legal declaration of IP rights prior to contribution), you are blowing FUD, Bob. The problem is not open transparent processes: it is transparent policy. If you actually understood the legal implications of IP, you would know that unless there is a participation policy in place, it is very dangerous to create standards based on unvetted contributions. For that reason, MHEG closes their work and reserves licencing rights to the membership. The Web3DC takes the same view as the W3C and refuses closed license IP. But it closed working groups to members to prevent submarine IP as well.

    I suspect yours is the first shot in IBM’s war on ISO. You contributed as much to the OOXML debacle as any single company did. Now you want to shift blame to the victim. Other groups are able to create consortia with transparent policies and work successfully with the ISO processes. Somehow, IBM, Microsoft, Red Hat, Oracle and so on don’t seem to be able to do that.

  4. It would be interesting to see if IS 25000 Software product Quality Requirements and Evaluation could be applied for document, application, protocol or API standards.

    IS 25000 is the new generation of IS 9126. See Wikipedia under ISO_9126 for the approach.

  5. len: I respectfully disagree, but you are welcome to your opinion, of course. Let’s try to be constructive moving forward instead of talking about FUD and how uninformed people are.

  6. The most obvious problem is that voting can be manipulated so easily.

    Other organizations do not allow votes to be “bought” quite so easily.
    A simple remedy would require an individual/organization to have been a member for (say) a year, and to have participated actively during that period.

  7. Philip, right, we need to make the process much more deterministic in terms of who can vote and when. It really can’t be that hard to figure out a workable scheme. In the US there is a deadline for voter registration. We need something similar for standards orgs.

  8. A possibility that standards organisations should perhaps consider is that all participating members either the individuals on committees or organisations or countries depending on the level should only be given a vote after full participation for over one year.
    It should also be possible for a requirement for a participating organisation to grant a 100% sublicensable right to use any patent, design registration etc. required for a full or optimum use of any standard.
    Microsoft may not have been so keen to force things through if adoption of the standard would mean that O’Reilly could immediately publish a detailed description of the entire thing.

  9. … under the heading, “there ought to be a law”, standards should be non-patentable and under a free license. Period. Full stop. When I read about Rambus, Qualcomm, and others who have pushed standards that they just happen to have key patents on, it makes my blood boil. There should be a law, internationally recognized, that any patents affecting an ISO standard are immediately nullified.

  10. Jim: Under that plan, given an patent, then create an ISO standard that necessarily uses the patent, then kill the patent. Is this what you mean, and for both hardware and software?

  11. One problem with standards organizations is the absence of a braking mechanism. ISO, MPEG, and other organizations can be manipulated by well-funded organizations who push their technology through the standards organization despite the apathy or opposition of the industry that is being standardized.

    The example I am most familiar with is ISO/IEC FDIS 14496-23:2007, the symbolic music representation addition to the MPEG standard. EU-funded researchers established this effort to get their own technology into MPEG, despite the fact that the industry was already standardizing on another technology. The EU provided funding that the tiny symbolic music industry could not match. Participation in MPEG is too expensive for an industry whose entire size is far smaller than single companies like IBM or Microsoft.

    I can’t tell from the web site if it just went out to ballot or it just ending balloting:

    http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=45531

    But if you are indeed opposed to manipulation of the standards process, rather than just opposed to Microsoft, please consider advising your national bodies to vote no on ISO/IEC FDIS 14496-23:2007.

    If it passes, it’s not a big problem: it will join ISO/IEC 10743 as an ISO standard that the symbolic music industry will ignore. ISO/IEC 10743 happened about a decade ago. Braking problems within the ISO process far predate the OOXML controversy.

  12. There’s a sense in which Microsoft rely a lot on ‘the community’ to hold back http://en.wikipedia.org/wiki/Storm_botnet , and other manifestations of ‘OnDemand eCrime’, which rides on the coat-tails of the ‘OnDemand eBusiness’ that we know and love.

    Now, dominant vendors stuffing the ballots at standards organisations, and dominant vendors attempting to extract payments from others for patent licences, really doesn’t help in our quest for provision of useful goods and services to paying customers.

    Where it all leads, I do not know. But something is not right with the world.

    What should we do, as individuals and as employees ?

  13. I strongly agree with len for having participation agreement signed prior to contribution (think about patent trolls).

    The participation agreement must ensure that all Intellectual Property contributed must be a blanket grant and irrevocable (no neccessary Intellectual Property claim type language, no termination of Intellectual Property provision) and possibly the participate to indemnify the standard organisation.

    I think the standard organisation should make available a copy of its Constitution to the public.

  14. We learned the hard way, Bob. If there is no participation agreement and contributions aren’t controlled by legal agreements, the patent trolls have a field day.

    I am being constructive based on real experience and long time contributions to 3D on the web, XML, and efforts that predate the web.

    Did IBM invite the Web3DC to the Virtual Worlds Interoperability forum? No. Why? They have the ISO standard, they have market presence particularly in Europe. The Chinese saw to it that the ISO standard was translated for their use. You know all of this.

    If you are being respectful, be honest. That is the only definition of respect that counts.

  15. @Chris:

    You have to decide individually what fights are worth fighting and what are you willing to gladly give up. I don’t want to use this blog as a morality platform, but you asked. Did you read Grady Booch’s comments on CNet? Finally, at long last, senior members of the community are coming to the realization of what some have been trying to get across: all justice originates from individuals. There is no justice in nature; there is karma.

    In lighter terms, you have to fight the Blue Meanies.

    There is much wrong with how standards get created, but standards don’t have to be adopted. Eventually the power of the keiretsu dominates the market space. There is no other means that preserves competition AND innovation. On the other hand, it is wrong to take a handful of private agreements for technologies, paper over them, and call it a standard. MHEG/MPEG is a significant barrier to competition and start ups. We’ve had to innovate around those patents. I suspect when the Virtual Worlds whosis publish, they’ll have similar problems.

    X3D, on the other hand, doesn’t. Those participation agreements are already signed and that standard already is an ISO standard. The momentum of VW is nascent but not unproven or sequestered. Regardless of how much money gets tossed at this, Rick Jeliffe is exactly right about the quality of the work and what is required. It doesn’t take a lot of people but the ones involved usually commit a whole career to it (See IBM’s Dr. Charles Goldfarb and the enormous sacrifices he made personally to see to it markup remained unencumbered. Contrast that with his classmate, John Warnock, at Adobe and how quickly he closed every opportunity to open his technology. PAS my…). The only group that has done what is necessary is the group(s) that created X3D, H-Anim and DIS. There is more to come from them soon and that is the space I will create in.

    Why?

    Simple: they pass the lifecycle test. When working with expensive long lifecycle content, that is the only test that counts. My 3D work from 1995 and 1996 still works.

    As for the rest, it will be harder than they think and may not last as long as they say. Real standards for interoperability and lifecycle have exactly one solution: open languages.

    The Language IS the platform.

    And that, Bob, is as positive as I can be because I can back it up with proof. OTW, XML and HTML would have disappeared by now. You have enough experience to know this is so.

  16. Re Jim Stanley’s idea, what if patent holders have a one year period to object to a proposed standard if they think it infringes. In that case, the proposed standard must be changed to avoid the infringement. Patent holders who do not raise objections during this time forever forfeit the opportunity. After approval, the standard is, by law, immune to any and all claims of infringement. This does not invalidate patents, but exempts implementations of the standard.

  17. Perhaps a useful way to measure a standards org’s resilience to untoward manipulation would be to imagine subversion scenarios: if I wanted to ram through a rotten standard to achieve marketing and sales goals, how would I go about it? How could I take all the teeth out of a standards process, in order to render it formally independent yet market-ineffective? How do I get in the backdoor fast track? The voter registration analogy is useful: there is a long history, in the US and worldwide, of creative folk who game the system in setting up a political machine. Tammany Hall lives on when we witness the cynical OOXML shenanigans. Part of the problem I think is that the creation and maintenance of technical standards depends on experts drawn most often from major corporations. The result is that standards committees can become proxy battlegrounds for non-technical reasons, and the most well-intentioned participants could find themselves asking whose interest they serve. Governments have a role to play, but a recurring problem is keeping them clueful. Stakeholders whose interests are not necessarily aligned with big business and government are under-represented. I do think it’s possible to strengthen standards orgs: in the way judges are former lawyers, and sports referees are former players, perhaps what’s needed are more standards orgs presided by retired and not active (employed) engineers, relatively free of conflict-of-interest pressures. The stereotype says grizzled oldtimers obstruct innovative changes, but I believe the opposite could be true: seniors who would hold no truck with ballot stuffing. Or perhaps standards review should be designed to fairly represent all stakeholders? Some elected, impartial monitors representing consumer interests?

  18. Bob,

    Hope I wasn’t too rash. I just think no company or organization should be able to demand a license fee for any hardware or software that’s essential to implementing an approved standard. Frankly, I don’t think a “covenant not to sue” or “reasonable and non-discriminatory” is enough: all tools, hardware design specs, protocols, algorithms, etc. necessary for implementing a standard should revert to public ownership (perhaps by the ISO?) once the standard is approved.

    I apologize for sounding really leftist, but can you imagine the slower pace of innovation on the Web if you had to pay someone every time you wrote a line of HTML? I don’t mean making the *tools* free: if IBM or Microsoft or anyone else wants to charge for, say, a word processor that creates standard files and does it faster and better than the competition, more power to them. I just don’t want to be charged for *creating* a file in that format, no matter what tool I might use.

    (slightly off-topic, to Music Developer: are you talking about MusicXML? Michael Good has done a tremendous job, and it’s interesting how he’s played the standards game to good effect.)

  19. Jim: not at all, was just running through the implications, and the “state of the art” there is different for hardware and software. Also, thoughts on this are changing. We fought hard against the W3C going from RAND to RF. We lost, and we changed. Seems like a generation ago. We’re still changing. So keep the ideas coming.

  20. I think the current roadblock can be cured with a simple amendment to procedure. After 3 attempts to get action on a project, if not enough players partipate to create a valid vote, then the moderator can at any time after the 3rd vote, call for a vote of only participating members on the project. If they can’t come up with pro’s or con’s on a project to keep it moving, then they shouldn’t be allowed to vote on the project. It’s put up or shut up.

  21. There’s “The Good Ship IBM”, and “The Good Ship Microsoft”.

    It’s fairly obvious that IBM cannot buy Microsoft, Microsoft cannot buy IBM, and the two cannot merge to form a kind-of “IBM-o-Soft Corporation”. The US Department of Justice would surely not allow it, nor would the European Commission. Probably the South Koreans would have thoughts on the matter, too.

    So, the two corporations are destined to sail the seas for the forseeable future in a kind of ‘co-opetition’. There are going to be clashes when the economic interests of the one conflict with the economic interests of the other.

    OS/2 is history; you can read that one, from when the two corporations broke ground on its joint development, to when IBM sold it to Serenity Systems at the end of 2005. Now, neither IBM nor Microsoft has any economic interest in OS/2. “Follow the money” over the 2 decades or so of its commercial relevance makes quite a story.

    IBM recommends you look at Linux and Microsoft recommends you buy Windows. But OS/2 is over.

    Now, we’ve got supercomputer chips for XBoxes. IBM makes the chips, Microsoft makes the XBoxes.

    We also have ISO 26300 ODF XML (favoured by IBM for deployment in IBM Lotus Notes http://www-306.ibm.com/software/lotus/notesanddomino/ ) and ECMA-376 OOXML (favoured by Microsoft for deployment in Microsoft Office, but I don’t think Microsoft claim conformance to the standard).

    So yes, we do need to talk.
    Is Microsoft going to pay IBM for the privilege of implementing the standard ? Is IBM going to pay Microsoft for the privilege ? Or are you going to declare it Royalty-Free for each other (and for everyone else on the planet) ?

    Are both of you going to try to extract royalties from hobbyists ? How are you going to stop yourselves, because doing that will dry up the world’s supply of future engineers and scientists faster than anything else you might do.

    Oh well. I’ve got to go shovel some more coal into the fires of the development engines of “The Good Ship IBM”. So long as I stick to task, I should be OK.

    Talk, by all means. But no secret deals.

  22. @ Chris Ward
    ODF is already, and always has been free for the world to implement and use. MSOOXML never has been. It looks kind of like it is on paper, until you get to the parts of the MSOOXML spec that aren’t actually included *in* the spec, but are rather included by reference to behavior of past MS products, and therefore not actually covered by MS’s patent grant. Of course, the truely sad thing is that not even MS is compliant with their would-be standard.

  23. Well, yes.

    I know some people who are school governors, and I ask them “Why are you not moving your schools to ISO standards for reviseable documents ?”. Also “Why are you practically forcing the parents of the children who attend your schools to buy Microsoft software ?”

    The answer is usually along the lines of “Compatibility with already-existing documents”.

    It’s a bit like wondering why your old VHS videotapes won’t play in your new DVD player.

    Sometimes it is time to draw a line under things, and move on.

    Yes, VHS http://en.wikipedia.org/wiki/VHS (a single-vendor standard, JVC) did rule the roost for many years.
    Yes, it has been largely superseded by ISO standard DVDs nowadays.
    No, your VHS video tapes will never play in a DVD player.
    If you need to play your old VHS videotapes, get yourself a VHS videotape player. You can get them on ‘ebay’ http://www.ebay.com/ and you will be able to do so for many years after the smart retailers have stopped stocking them.

    But really, it’s not a clever idea to make new ‘videos’ on VHS.

    ISO standard DVDs are the way to go.

  24. When I sat on the BoD of the WeDC, I fought RF too. It wasn’t to get royalties. My objection was the 3D industry was at its lowest point in the market but that this would change when the then 13 and 14 year olds entered the market. Being life long gamers, they would be very 3D savvy and demand it. At that point, the market would explode and the big companies that had abandoned it would be back. Because of the problems we had experienced with Microsoft and Intel as members over tactics and IP, I felt that when this point came, they would do anything they could to lead the industry away from any consortium with RF policies and this would lead to the extinction of VRML and X3D.

    That was five years ago. Today, I get to find out if I was right.

    X3D is RF. The reason is because the BoD took the W3C lead (a liaison partner) which demanded compatible conditions. As a result, anyone can implement it (given skills) and any hobbyist can use it. The questions now for the big companies just entering this market is do they need the IP-free standards to get into business and do the authors accept the risks of creating content for closed systems. A third question is have the companies currently entering into this market to sell services, servers, and systems done their IP research to discover if any of the patents existing or applied for actually have value?

    ODF/OOXML is a dodo war. There are too many existing formats for the outcome of this to matter. In the virtual worlds market, it matters a great deal. So the fourth question is one of architecture. As Forterra is asking smartly, should the systems coalesce around a multi-speak client or a multi-speak server? What is obvious is that there will continue to be multiple formats in both markets so any solution that begins with the statement “there must be ONE format” is a non-starter marketwise. That ship sailed a long time ago.

  25. Regarding Jim’s patent exclusion comment & your reply..

    The simple answer is “Yes, exclude patents in standards”. Bear with me.. Patents are for the benefit of society. Standards are also for the benefit of society. The U.S. government (to name one social entity) has already affirmed that patents may be, on a case by case basis, nullified. I say that the benefits of a standard being free for all far outweighs the protection of a patent. It should go without saying that software patents should not exist. Copyright already exists to protect wording, said copyright only applying where the wording is made public (the ‘right’ to the ‘copy’ implies a public copy). Entities that wish to use secrecy explicitly give up copyright protection. Free choice exists. Protection should not be tailored to the greatest financial/political power. Patents in a ‘standard’ give the patent holder control over the use of that standard negating the value to society of said standard.

  26. Ken, I was referring more to the Rambus case, where they pushed for certain specs in the JEDEC DDRAM standard, and then turned around and sued everyone implementing the standard because they had patents on key portions and didn’t bother to tell anyone in JEDEC. The FTC came down quite hard on them, but didn’t go so far as to invalidate the patents:

    http://www.extremetech.com/article2/0,1697,2090333,00.asp

    This case has quite a bit to do with the general topic at hand: “gaming” a standards body, open standards vs. intellectual property, etc.

    OK, so maybe we shouldn’t go so far as to nullify all patents becoming part of a standard. I’ve often wondered if software patents would be workable if they were valid for only two years and non-renewable. That might strike a balance between a decent return on investment and not locking out innovation based on prior work. The tremendous pace of innovation in the web world makes a 14- or 28-year patent pretty ridiculous.

  27. Which particular dodo should win (in ODF and OOXML) ?

    Multiple standards is like having multiple kinds of plugs and sockets, multiple voltages, and multiple freqeuencies for the mains electric wiring round your house. That would be a pain for householders and appliance vendors alike.

    I know that in the US you are 110v 60Hz, and in Europe we are 220v 50 Hz; that does cause some interoperability problems, but at least we know what ‘110v’ and ‘220v’ each mean, and we can cope with it because they don’t keep changing.

    OOXML is baroque, and Microsoft-only. “Martian”

    ODF is simple, and vendor-neutral. “Esperanto”

    I think the ‘multiple formats which do broadly the same thing’ will coalesce. This guy http://symphony.lotus.com can read all the Lotus SmartSuite ones, and translate to “Esperanto”, all at no charge. A useful player to have on the team.

  28. @Chris:

    Neither will win if by winning you mean one will be the only highlander left standing in the market or the standards world. They will jostle one another, little bitter butter battles will erupt but the public and the customers will tire of these and they become like the boy crying wolf. The ‘one size fits all’ format ship has sailed. The anti-OOXML/anti-MS forces and the pro-ODF/pro-IBM forces are screamers at an empty dock. But by all means, do party on.

    @Bobby: Patents will go on because they do serve a purpose even if abused. Standards such as MPEG will continue to collect royalties because of the agreements among the members of that standards group. What you want is a standard form of participation agreement that is transparent and binding to members of the group to which you contribute your intellectual property be it unencumbered thought or previously patented material. Without such prior transparent agreements there will be mischief such as the submarine patents of Rambus and Intergraph.

    To determine which groups will be successful, look at the patterns of behavior. If a forum begins with a technical submission they have a much better chance of succeeding in writing standards then the group that begins with press announcements and exclusive recruiting efforts to brand them as the leaders. The second group will be a keiretsu or simply, a network of interlocking deals for goods and services. The first group will be a network of interlocking technologies. Both will attempt to lead but money is a consumable and once spent, irretrievable except in terms of the contract made. Technology networked together is a growing entity that becomes ever more powerful as it is installed and used. Its evolution depends on strength of the participation contracts as seen in the continued willingness of the members to make contributions that improve its features and therefore, expressiveness.

    I am impressed by Forterra’s opening at their Interop forum. The company officers signed up for the discussion and they opened with a page terrain technology. Now it is a matter of conditions of participation. The Web3DC liases with the W3C successfully based on compatible RF conditions and goals. It is the participation agreement one wants to watch because it determines the bandwidth of cooperation legally and ethically possible.

  29. @Chris:

    The similar standards don’t necessarily coalesce. VHS won not because of pressure on players but because of the low-rent conditions of neighborhood video rental stores. The weak force that turned the tide was shelf-space in Mom and Pop stores: VHS titles could be stored more densely and made cheaply. Pornography titles became the secondary weak force given production costs. People were less interested in the high dollar film releases than anyone knew.

    IOW: it isn’t the technology that coalesces. It is the buying pattern based on availability and the intangibles of taste. Note that these are also non-linear dynamic systems (choosers of choices, or second order systems).

    To understand how an ecosystem evolves, look carefully at the edges where the ecosystems overlap each other. In biology, this is called an ecotone. Conditions are hot at the ecotone because of the high rate of exchanges there and the intensity of competition for resources. To analyse document formats, be that office formats or virtual world formats, you need to analysze the forces dominating transactions. This is pure S-R modeling.

  30. Well, I don’t think it will coalesce to 1 standard. I think it will coalesce to 2, and they will be known as “The ISO one” and “The Microsoft one”.

    Salesman will use the Microsoft one for their highly-differentiated sales pitches.

    Scientists will use the ISO one for their long-lived documents.

    In between, well, who knows ? Salesman and scientists can both afford to pay for software-with-service. It’s a question of how to take the money, and how much to ask for.

    In the UK, Microsoft have got themselves referred to the “Office of Fair Trading” by BECTA, the government-run technology procurement body for schools, for some kind of behaviour which BECTA thinks commercially inappropriate. http://news.becta.org.uk/display.cfm?resID=34425 is the reference.

    I’m not sure it’s very helpful; I don’t think it will being OS/2 or SmartSuite back to life. Those ships have left the dock, made whatever profits they were going to make, and sunk.

    If IBM’s going to sell anything to schools, it’s going to start again with Lotus Notes http://www-306.ibm.com/software/lotus/notesanddomino/ . On the ISO standard. Lotus Symphony http://symphony.lotus.com/ is the advertising material.

    I don’t know if schools can afford the price. But they can certainly ask to speak to their salesman.

  31. IMO, OOXML will eventually become the alternative ISO standard for office document. My speculation is the customers of both corporations will barely notice. It is a rallying cry for web-centric marketing and sales politics in the halls of power, but very little else. The standards focus for next generation products is already moving on to other freshly invested money pits such as virtual worlds and interactive GUI editing of the new generation of documents that are one part information to be consumed (content), one part GUI and one part network plumbing. Who says this? Adobe.

    This will be where the alternating tactics of closed IP but open source vs open standards with IP-unencumbered agreements will fight for mindshare and eyeballs. … yet again. Someone wrote to me to say that this is where the dysfunctional aspects of our larger corporations with regards to these topics will become most evident.

    Customers will continue to choose along the continua of traditional cost to acquire, features needed and lifecycle commitment within the contexts of existing investments and new contract opportunities.

    And so it goes.

  32. By the way, while here, I want to offer a public apology to Bob Sutor for being a caustic ass in some of my comments of late. Bob has been most generous to provide all sides a place in his private blog for all sides of these discussions of standards and other issues. I am making a most amateur mistake to conflate the decisions and strategies of the company of which Bob is an officer with his private opinions in regards to decisions made about the future of virtual worlds and virtual world standards.

    At a recent interoperability forum sponsored by IBM, ALL of the existing standards organizations were shut out of admittance and discussions. That this rankles the supporters of existing standards and standards organizations would be an understatement. That Bob had no control or input to those decisions is not as obvious I am egregious in assuming he did. Over the many years of working on open lists on standards beginning in the SGML days when we were young turks, I’ve come to respect Bob’s expertise and professionalism with regards to this and I have no cause to doubt that now. Any of us who have worked for very large corporations know that they are hydra-headed and one group can often be making decisions at variance with the others wearing the same logo on their badges.

    Whatever the case in these current discussions, I’ve no right nor reason to assume Bob is acting differently or advising the corporation to decisions contrary to the spirit and fact of open standards for emerging markets and technologies. I do apologize for my remarks.

    I do also have to fight for the organizations that have proven by act and contract their past, present and future commitment to open systems and IP-unencumbered standards for 3D on the web and the implementations of this in virtual worlds. It is my belief that this is the fight for the rights of the generation of world builders now coming into the market, who are in fact, our children. Virtual worlds for them are what rock ‘n roll were for our generation. We may be the ones who kick the doors open as Bill Haley and the Comets were credited and the untold numbers of black musicians who did without credit, but they will be the Beatles, Dylan and Clapton of the art form. If we do not hold this door open until they can establish themselves, we have robbed them for the sake of our current careers already entering the last arc.

    I don’t think we want to be remembered for having done that.

  33. Well, we definitely are the ones who are creating the foundations for tomorrow.

    I think we are putting the ‘computer science products’ of the past 20 years … since the dawn of the Personal Computer era … very rapidly into the crusher, so that their ashes can grow the phoenix.

    That would be OS/2 and Lotus SmartSuite, wouldn’t it ? They have been cleared out of the way, to make room for POSIX-standard-conformant operating systems, and ISO26300-standard-conformant office productivity suites. And what’s more, the price of them is ‘zero’ for those of us who know where to pluck them from http://ibiblio.org/ http://symphony.lotus.com/

    Now, what a certain other dominant vendor is doing, we’re not quite sure.

    But we’ve got an Internet now.

  34. The problem is not putting the products in the crusher, Chris. It is putting the ethos of open systems and transparent policies and participation agreements in with them. While one side of the face talks the talk, the other side is walking straight toward reproprietarization of the web through next-generation products and the same claims that open source equals open standards and fast enough isn’t fast enough.

    Marketing a product, open closed or wrapped in tin foil is not a reason to call burglary “a liberation”. Interesting times, indeed.

  35. IBM’s a big company, and it likes to think that it will make profits from hardware, software, and services.

    What tends to happen is that when one part gets squeezed, profit moves ‘along the value chain’. And the squeeze against OS/2 and SmartSuite has been severe over the last few years; guess which large competitor squeezed them ?

    I think IBM is much happier with Linux than with OS/2; Linux can run all of IBM’s servers, and there really isn’t anything else that can run the ‘System Z’ http://www-03.ibm.com/systems/z/ mainframes in Chinese. The EBCDIC operating systems (MVS, VM) don’t have enough letters in the character set for the first-class Chinese business person to even write his or her own name on them.

    As to the future of SmartSuite, well, you can buy it, but http://www-1.ibm.com/support/docview.wss?rs=479&context=SSKTYF&dc=D600&uid=swg21253472&loc=en_US&cs=UTF-8&lang=en says that the writing is on the wall. It seems to work under Windows Vista but if you want IBM to support it there then you should ask your salesman for a ‘special bid’ quote. And it’s pretty unlikely that it will ever read documents in ISO26300 or ECMA376 formats. http://symphony.lotus.com/ pretty much torpedoes SmartSuite; the only question is how much other vendor software sinks with it.

    So, Linux, OpenOffice.org, and IBM Lotus Notes http://www-306.ibm.com/software/lotus/notesanddomino/ will suit IBM just fine. And if you pay IBM’s prices, IBM will warrant that it all works together for your business. If you don’t pay IBM’s prices, well, all bets are off. I’m sure we won’t deliberately upset any applecarts, but that’s as far as it goes.

    There’s a certain amount of what might be called “Business Optimisation” going on … we sell that, too http://www-935.ibm.com/services/us/index.wss/bus_serv/bcs/a1008891 … which, since IBM’s economic interests aren’t identical to Microsoft’s economic interests, is tending to produce instant globally-integrated competitive responses. Lotus Symphony is out of IBM Beijing labs, where the engineers are cheaper. We watch with interest and amazement (and amusement, sometimes). Unfortunately I’m not Chinese, and there’s not a lot I can do about that.

    Interesting times, indeed.

  36. So goes the pace of globalization. IBM surrendered to China and other companies are kowtowing as fast as they can send salesman. At Intergraph, we were working that market well over two decades ago. Markets seeks levels like water seeks low ground. The problem of annealing processes is sub-optimum minima that have to be socked really hard to restart the flow.

    This is normal and predictable. The US dollar is in the toilet for awhile, the Chinese are holding the notes, and just as we had to do in the late 70s and 80s to dig our way out of the oil debt and the falling dollar post-VietNam, we have to do some major kowtowing to rebuild from our latest misadventures. Again, predictable as sunrise.

    On the other hand, the meme boiling up from ostensibly multiple sources is that of single language ecosystems. This is the most pernicious one. It has failed repeatedly in our memory and probably before but it is the cut of the knot most inexperienced or otherwise misinformed sales people and politicians try. Markup succeeded precisely because it is the antithesis of the single language ecosystem. Think of DNA and you see how the engines of living systems actually work in cooperation or competition with the host environments.

    What is truly entertaining is how strange the bedfellows become and how former bad guys turn out to be the good guys at the denouement.

  37. Please don’t use the phrase “intellectual property” It allows too much bull**** and delay. Either a specific issue concerns copyright, or it doesn’t. Or it concerns patents (even sw patents) or it doesn’t. Or it concerns trademarks, or it doesn’t.

    Going on about “intellectual property” allows these specifics to remain unaddressed and progress to be delayed indeffinitely. Afterall, if the problem cannot be identified, then it cannot be addressed.

    That aside, your summary of the problem is the best to-date:

    “There were so many things fundamentally wrong with how OOXML entered the standards process, was propped up and pushed forward, and just never seemed to go away even in the face of withering analysis and criticism, that it is now clear that things have to change.”

    I would add further that it is now clear that vendors need to be called to task for foot-dragging on ODF support and called out on why they are spending resources on OOXML. Apple, for example, could have had ODF support a long time ago, but what went on that it had to be delayed for OOXML and the two appear side by side in *every* press release about formats?

  38. IBM didn’t surrender to China; IBM invested in China.

    It may be hard for those of us who aren’t Chinese, and who therefore don’t personally benefit from being invested in; but so long as it’s a good investement, coming up with marketable products and services that can be sold with the IBM logo on, it sounds fair enough to me.

    DNA works with mRNA transcription, sort-of like a zipper; then the Ribosome taking the mRNA like a knitting machine takes a pattern, and constructing proteins like the knitting machine makes jumpers out of wool. The proteins do the interesting stuff to make life work.

    It’s all very interesting, but a million miles away from the question of why Microsoft stuffed the ballot at ISO.

  39. All of the parties stuffed ISO. It’s a mess now but if you look at the numbers, a roughly equal number of pro and anti forces left damage and have locked up the voting processes. That’s real damage and blaming it on Microsoft is just another way to duck responsibility for setting attack dogs on the process to prop up one format and to damage the MS customer base.

    See, we won’t agree on that one. I’ve done this too long to be fooled by any of the players. It is bad behavior all around but IBM is fanning the flames and that is what will burn down the process. IBM stands to gain from that by pushing standards making into the hands of the MHEG-like organizations that keep the IP locked up among their members. If you contrast that to the W3C or W3DC where IP is held in the common to be freely used or improved by anyone, you see a sharp contrast in means and goals.

    The emerging keiretsu are acting like city-states where the power is not in the technology but in the interlocking contracts for goods and services. Language evolution is more like DNA for precisely the reasons you describe: pattern zipping based on a common syntax. The single language theory flies in the face of everything we’ve learned about language evolution. It never works. It treats the elite cultivar as the only exemplar and it soon becomes unviable in the face of environmental change. The elite cultivar relies on its genetic structure to be refreshed from other exemplar found in the wild.

    ODF as the only standard simply won’t work. And it won’t happen. It won’t even be something that IBM or MS or ISO determine. The natural market forces intervene.

  40. As I said, I think there will be 2 formats for XML reviseable office productivity documents; the ISO one (which is good for long-lived documents) and the Microsoft one (which is good for short-lives sales pitches).

    It’s rather like there being 2 formats for character representation; the EBCDIC one (good for punch-cards, now only honoured by IBM) and the ASCII one (good for paper-tape, the American Standard Code for … ,also honoured by IBM and by everyone else in the industry)

    So, Esperanto and Martian. Stray cats and pedigree cats.

    I’d prefer it if we all learned Esperanto. But if you’re a real good customer, and you want to speak Martian, I’ll make sure there’s someone on the staff who can engage in a dialogue in your preferred language.

    And of course I’d like to sell Lotus Notes to the Microsoft customer base. I’ll sell it to anyone who can agree a price with their salesman.

  41. And there we should rest the case. The problem of the Tower of Babel myth is no one actually knows how to interpret the will of the Divine. Was it a case of jealous rage against a single language entity, or a case of merciful salvation of individual thought and access? One can read it both ways.

    Esperanto perished because it turned out, no one really wanted to learn it. It is a good idea that turns out to have no lasting champion or need. Regardless of pedigree or lack thereof, if the cats won’t queue, they can’t be herded. Given the current majority for the incumbency and the switching costs, it will take some time before challengers will have the necessary credibility. Until then, it will be best all around if the electioneering doesn’t include burning down the courthouse. We should take great care of what we do to ISO. All they guarantee is professional editing and process, and we need those badly. In the last decade, we tried everything else and we come back to slow, steady, and in person.

  42. Thing is, no-one (except Microsoft) knows what the incumbency is. It certainly isn’t the ECMA-376 OOXML that they presented to ISO ; Microsoft Office 2007 is using MS-OOXML which is ECMA-376 with “Martian” extensions.

    IBM Lotus Symphony http://symphony.lotus.com/ , ‘vendorless’ OpenOffice.org http://openoffice.org which now seems to be offered whenever Sun release a fix to Java , and Google Sun StarOffice http://pack.google.com/, could upset the incumbency real fast for those who have Internet access and a little time on their hands.

    Then there’s the 120 million IBM Lotus Notes http://www-306.ibm.com/software/lotus/notesanddomino/ users, who presumably have money to spend with IBM and busy lives, because they work for ‘first class’ businesses. ISO26300 documets come as standard for them too.

    Yes, ISO does need looking after. How are they to know whether what they are asked to rubber-stamp is ‘technology consensus amongst suppliers, consumers, and other interested parties’, or ‘effects of a wall of money’ ?

    If all you want is “Professional Editing and Process”, just come into any IBM development lab. However, there’s a chance you might also require “Vendor Neutrality”.

Comments are closed