Another leap forward for openness?

By Rick Jelliffe
March 25, 2009 | Comments: 14

The chairs of the ODF Technical Committee at OASIS have made a new response to my recent blogs which query the concentration of voting power in the ODF TC by...wait for it...not showing the corporate affiliations on the rollcall list in the latest minutes. A cunning plan indeed!

Either that or all members of the ODF TC have resigned from their day jobs.

This comes after my blog on the last meeting, where application makers outnumbered representatives of users 10:1. In the comments to that blog item, one of the ODF TC co-chairs characterizes my comments on the desirability of broader participation by non-vendor participants as like "broad participation from vacuum cleaner salesmen, Egyptologists and professional oboe players." Not exactly the warmest of welcomes.

My patient readers will no doubt be interested in the tally from the latest meetings. Here is the annotated rollcall:

* Rollcall

Bob Jolliffe, Republic of South Africa, Department of Science and

+David Faure, KDE e.V.
+David Wheeler, Institute for Defense Analyses
+Dennis E. Hamilton
+Donald Harbison, IBM
+Eike Rathke, Sun Microsystems
+Eric Patterson, Microsoft
+Florian Reuterr, Novell
+Jomar Silva, ODF Alliance
+Michael Brauer, Sun Microsystems
+Ming Fei Jia, IBM
+Oliver-Rainer Wittmann, Sun Microsystems
+Patrick Durusau
+Peter Junge, Beijing Redflag Chinese 2000 Software Co., Ltd.
+Robert Weir (presiding), IBM
+Stephen Peront, Microsoft

Voting Members are indicated with a + before their name.

- 13 of 17 voting members (88%) were in attendance, so quorum requirements were met.

I think there is a typo there, because there are 15 members, not 13.

So that gives the ratio and percents of voting members of the TC as

1) Office suite developers: 11/15 = 73%
2) Commercial voters: 10/15 = 67%
3) Voters associated with a single code base: 7/15 = 47%

which is pretty much inline with the figures from the previous 12 weeks.

You might also be interested in:


So the ODF TC has healthy participation levels from those individuals who have the greatest interest and expertise in the technology undergoing standardization, as well as a strong commercial motivation to ensure that the work is completed in a timely fashion and that the resulting technology meets user needs.

Golly gee whiz, Rick, that sure sounds like a recipe for disaster to me. Thanks for pointing it out!

Of course, 99.99% of all standards are created this way, and you are merely spreading another dihydrogen monoxide rumor. A bit early for April Fool's Day, don't you think?

Rob: No. It is dominable by representatives of the supply side of the markup, by big business commercial interests, and frequently by a single code base.

By this stage, you are probably thinking, "Gosh, Rick is right you know. We should do something to be more actively balanced. We must try to do better. But what can we do?..."

So here are some suggestions:

1) Actually bother to promote broader representation by the demand side (users, integrators.) Have you ever written about it in your blog (you are asking for ideas at the moment)?

2) Allow all individuals who don't work for US companies and who have recognized expertise to join as voting members free. Keep on adding them until there is an average of 50% TC voting members who don't work for a US corporation.

3) Voluntarily get IBM and Sun to join Microsoft in agreeing to limit the numbers of voting members in the TC to 2. Actually: follow the algorithm I gave in the previous blog to determine numbers.

4) Take responsibility for achieving breadth.

Perhaps you could follow that Portuguese model, and make sure there is enough floating membership so that there are always n representatives from government, n from academia, n from small business, n from big business, for the same n.

Of course, I welcome any additional members, individuals, oboe players, whatever. Non-aligned members of the TC currently hold leadership positions, such as Chair of the OpenFormula SC and Chair of the ODF Interoperability and Conformance TC. They have also been quite success at proposing changes to the ODF standard and seeing them through. This is a process that works well. It is odd that I only hear you complaining, not them. You really haven't shown a problem. All you do is show numbers and claim that something is not right. But other than the numbers not matching your a priori model, you have not demonstrated a problem.

In any case, I hate to give you the bad news, but I am working on getting even more commercial office suite vendors on the TC, like Apple and Corel. I think getting them to the table, where we can work on a common approach to document interoperability is what will benefit the average user most, in a direct and tangible way. Although I welcome the end user, there is no substitute for having the participation of vendors whose products are used by millions. Since I can't have millions of end users on the TC, having a vendor who serves millions is the next-best thing.

Rob: Total public silence is a funny way to express welcome.

No complaints? Well, you certainly are aware of comments such as the following (since you responded to them):

"I am a former member of the OASIS ODF Technical Committee. I left two years ago because of that big vendor-dominated TC's obdurate refusal to get started on make the ODF Interoperability Myth that the big vendors spread come true."


"Apparently the co-chair of the ODF TC prefers not to discuss here how to make the ODF Interop Myth come true. Unfortunately, that is how it works on the ODF TC as well."

Now I am not interested in the merits or otherwise of that comment; but you just cannot say there have been no complaints. There have been complaints, just not ones you see fit to acknowledge.

Rob #2: Membership from Apple and Corel would help in the third of the three concentration issues I identified, which concerned voters associated with a single code base. So, yes, that would be good.

But does seeking out Apple and Corel involve so much effort that you cannot write a blog calling on broader non-vendor participation? Why not offer waived membership fees for some additional OpenDocument Alliance people from non-G8 countries for example?

As one of those formerly-non-aligned participants, I can say that I feel good about what I accomplished on the ODF TC (ODF 1.2 would probably not have enhanced emtadata support if not for my participation). You selectively quote Paul, but why do you appear to take his position at face value?

I think Paul's view of this, which you are effectively propagating, is entirely too simplistic. He conveniently omits contrary perspectives. For example, ask Thomas Zander from KDE about some of the conflicts that Paul associates only with "big vendors" (or in my case, people who he suggests, without any kind of evidential support, colluded with them).

I do think there may be real tensions between, say, office suite engineers, XML processing geeks, etc., and that it's good to have diverse representation. But I really don't think the numbers you're presenting prove anything except who is participating; it doesn't say why, or how that might be changed.

Bruce: Yes, and I have not cited Marbux's comments as evidence there had in fact been collusion in that case, but as evidence against Rob's claim that no-one had ever complained: I am interested in improving the process in the future, not in making accusations. I think ODF needs it. In any case, what would be the point? You cannot change history. Now for all I know, Marbux may be a complete nutter (or he may be spot on), but that is irrelevant to the point (a specific response to Rob's amnesia.)

And actually I think I have repeatedly given the way the process can be changed: broader participation...a more diverse range of voting TC members. Not impractical, not punitive, not accusatory.

E.g. Get more non-supply-side participants, and adopt (even voluntary) limits to prevent numerical domination by particular interest groups.

Now I does not surprise me if you did good things at the ODF TC: in fact, it is the core of my point. There has been a long-running drought of yous on the ODF TC; this can be objectively measured; the chairs seem perfectly content about this (they want more thems); it needs to be pointed out and concrete steps need to be taken.

The diversity of the ODF TC is essential to its mission, I would have thought.

And what has the actual response been? To deny that supply-side participants are valuable, to remove the affiliation information from the rollcall to make the drought less obvious, and to deny that anyone else sees any issue.

(Indeed to claim that I am talking for some mysterious corporate master, when my company does not have any current or planned development projects for Microsoft, and Microsoft are not interested in the issue: indeed, as commercial vendors, they are part of two of the three imbalances I identified.)

You will remember a year ago there was some talk about standards for standards bodies. Surely getting metrics should be an utterly uncontroversial thing in this age of evidence-based management and ISO9000 QA? I think the numbers show that there is an imbalance, and one that can be rectified.

Bruce: To say that quoting something, even with a disclaimer and only as a specific response to a specific comment buried in the comments section, is the same as propagating it is to weigh propaganda more highly than debate.

My readers know I have views on many issues that are eccentric compared to conventional wisdom: I think that is why they read. I always assume that they weigh it up and make up their own minds; what else could they be doing? If a reader wants to immerse themselves in kneejerk zeitgeist groupthink, they can go to Slashdot or Groklaw or NOOOXML.COM.

Now I do have a blog item The Education of Gary Edwards in which I do bring up some of these issues in the body of the blog.

For example, I quoted Edwards: "The last official ODF TC vote I reviewed had Sun with six votes, IBM with three, and one independent." Surely that concentration cannot be healthy for an open committee, and surely the answer is broader participation? That Edwards et al may be toxic and doctrinaire panic-merchants (or not) is absolutely no response to the objective concentration numbers.

OK, fair enough Rick; let me retract the "propagating" word. You were not actually attributing anything in particular to Marbux's quote.

I agree the removal of affiliation information on the rollcall was odd/strange/not a good idea.

I also agree with the diversity issue, but am still wondering how to address it. Many people I know who have been interested in ODF and who may have briefly joined the TC simply don't have the time to dedicate to it. I don't think there are issues particular to the ODF TC that preclude anyone from participating. Indeed, as Rob suggests, one way to change the balance is for you to join (which would be a good thing).

But there are larger structural issues that probably do explain the current composition.

So I guess I'm just trying to say I think both you and Rob have reasonable points, and whatever truth may lie somewhere in the middle.

Bruce: I would be happy to join if someone took over my editing of Schematron and its related open source coding. (And if it were free: the OASIS fees were one reason I did not join the original ODF TC when kindly invited by Jon Bosak. Fees are a rationing mechanism that work against breadth of membership in favour of US corporate participation; they need to be accompanied by a pro-active and common non-fee participation system, such as W3C, Ecma and ISO have in practise.)

However, I do suspect that in the whole world it would be possible to find five or six people of adequate expertise and language skills and time who could represent users. The world is full of people. (Indeed, there should be thousands of names available from the ODF/OOXML debates.)

My suggestion? The chairs issue a call for participation including a willingness to pro-actively waive fees to get better balance.

Rick, if you cared about facts or logic, you would notice that the membership of the ODF TC consists of 61 persons from 22 companies/organizations and 7 non-affiliated members. I think this exceeds the breadth of representation in any SC34 WG. This is all listed quite clearly on the page called called "Membership" at the TC's home page:

All of these are ODF TC members. All of them have paid dues or have been granted fee waivers. All of them are eligible to participate on the mailing list and attend meetings and contribute to the TC's work. So your argument about needing to recruit more members or offer more fee waivers is unsubstantiated. The ODF TC is a very large and broad committee already.

Of course, not all of these members attend meetings regularly enough to gain voting membership. That is their choice and their choice alone. Some members participate purely via the mailing list. If they wanted to attend meetings they could. We offer a toll-free number around the world. We post agendas in advance of the meeting. We meet at a time that is workable, in practice, for members from Beijing to Hamburg to Dublin to Boston to Seattle. If a member wishes to attend, we have made every reasonable accommodation to ensure that this is possible.

In any case, you seem to have a problem with what subset of members shows up for meetings. But meeting attendance is purely a voluntary action and is not in my or your control.

Let me get this right: having a whole lot of members who have paid fees but are not actually participating in the ODF TC is a good thing? And that it doesn't matter how limited the affiliations of voting members on the TC if there are a broad range of non-voting, non-attending members?

(It is not as if, under the accreditation rules, they act as a reserve army who could step in in an emergency to ensure broader voting on a contentious issue.)

Actually, we have members who participate on the mailing list, even though they are not interested in attending weekly status report calls. I think this is a good thing. Everyone should be able to participate at the level they wish. Their are no barriers to participation here, only individual preferences and priorities. Do you see a problem with that, Rick?

And remember that "voting membership" is not a class of membership at OASIS. It is a status. A typical TC member will see their voting status come and go based on their attendance at meetings. This happens automatically, week-to-week. When we are nearing completion of an item which a member has a particular interest in, then they tend to be more attentive to their attendance, so they don't miss out on discussion and votes. Then their attendance wanes and they participate less frequently, and more via asynchronous means, such as the mailing list. With a 700 page standard, there are relatively few people with an interest in every topic.

Of course, a vendor who implements all or most parts of the ODF standard will typically have a motivation to participate more fully in discussions related to all parts of the standard. But non-vendor members often have interests that are more scoped to a particular area, based on their specialized expertise. I think this is a good thing. I'd rather have 20 outside specialists on-demand to help us with particular areas than have 10 more outside XML generalists.

You really should join the ODF TC, or any OASIS TC for that matter, so you can get a better sense of how it works. Although your screeds have been a constant source of entertainment for us, they have not served your readers, or O'Relly Media well as a source of accurate information.

Why is it that IBM and Sun together have been holding such a tight grip on the ODF TC with continuosly having a majority or near majority of voting membership.
Why is it that IBM needs to have 14 members on the memberlist you posted and Sun 10 members.
from 2002 to 2007 it seems that noone participated from the big office suites MS Office, Apple iWorks and wordperfect and alm ost almost all activity in ODF has been based around for most of its years.

Also the rand this continues up to this day. The release of OOo 3.x has effectivly limited any activity to improve ODF 1.2 because it was not in Sun and IBM's interest to add any features after OOo features set the target in stone. Next time just let the OOo team decide which features they are going to support in OOo 4.x beforehand so the TC knows what to add to the spec and not waste any time.

News Topics

Recommended for You

Got a Question?