97 Romania says yes to OOXML once again

Posted: Mar 27, 2008, under Romania, Morals&Politics, Standards. Updated: Apr 1, 2008. Add a comment!

On March 26, Dan Matei, the president of CT210, the Technical Comittee in charge of OOXML within ASRO (The Romanian Standardization Association), made public the results of Romania’s decision regarding the adoption of OOXML as an ISO standard: 15 votes in favor, 6 votes against and 5 abstentions.

With that out of the way, let us get down to the dirty details.

Note: Since ASRO refuses to talk, a lot of the information here was assembled from the members of CT210 and from online discussions. Some details may be wrong. I urge those who detect mistakes to leave a comment or contact me. My personal position aside, I want this article to present all facts as they are. Please DO NOT whine on your own blogs (you know who you are) about how inaccurate this article is as long as you haven’t made any attempt to tell me to fix it.

1. A bit of background

I will ask readers who are not familiar with the OOXML vs ODF saga to read my previous article on the subject. It contains details on Romania’s September decision and links to make this whole thing more clear for you.

Those who understand Romanian can also read a thorough explanation of why OOXML should not be adopted. For English speakers I recommend NoOOXML.org.

In September, during the world-wide process of deciding whether OOXML should become an ISO standard, Romania voted a resounding ‘yes’, without any reserves.

There was no foul play involved back then, only a particular mixture of incompetence from ASRO’s part and of general lack of caring from Romanians everywhere.

Traditionally, CT210 had been staffed by people who joined because they wanted to work on “Character sets and codifications”, which is the committee’s official name and designation. Technically, however, CT210 is also the correspondent of ISO/IEC’s SC34, so OOXML was automatically entrusted upon it.

In a bizzare twist of events, the committee members, instead of admitting their own lack of interest and knowledge in the matter, chose to rubber-stamp the standard through, voting no less than a definitive ‘yes’, without any comments. Read all the gory details.

2. The post-BRM decision

2.1. The warm-up

After September, there was at least one well-known Romanian forum where the issue of OOXML was discussed at length: TIC-Lobby.

For months, OOXML and Microsoft have taken over any other subject on this forum. The pro and con debates were heated, numerous and lengthy.

In all this time, OOXML supporters have utterly failed to offer any solid argument for OOXML’s adoption, while opposing arguments were abundant. The only coherent pro-OOXML statements boiled down to the same tired FUD that Microsoft has put together, repeated ad-nauseam, in spite of all the rich and detailed counter-arguments.

The OOXML supporters dismissed the opposition, saying it was just a numbers game, that they are badly outnumbered on TIC-Lobby and that on a public forum with a more balanced distribution of positions the debate would not sway so strongly against the OOXML adoption as a standard. It is my personal opinion that quantity has nothing to do with quality, and that the OOXML supporters have failed to make a solid case for themselves.

At some point, as the ASRO meeting was getting closer, OOXML supporters were being practically begged to put together a succint pro-OOXML argument that would serve to enlighten CT210 members who read the forum on why OOXML should be adopted. They have failed to do so, reactions ranging through simply ignoring the request, lame excuses and taking offense at one thing or another.

2.2. The CT210 member list

The full member list of CT210 at the date of the vote has been made public by Dan Matei, the CT210 president, on March 27.

There appear to be 27 members, including Mr. Dan Matei’s institution, CIMEC (I remind you that the members are juridical persons).

I have marked in bold those who were already members in September.

(I apologise if I’ve misspelled any member’s name. Please leave a comment and I’ll fix it.)

1. ARIES (The Romanian Association for Electronic Industry and Software)
2. The Linguistics Institute (associated with the Romanian Academy)
3. CIMEC (The Institute for Cultural Memory) (the representative, Dan Matei, acts as comittee president)
4. DIMSOFT (company)
5. Fujitsu Siemens (company)
6. GECAD NET (company)
7. Genisoft (company)
8. Hewlett Packard Romania (company)
9. I Computer Solutions (company)
10. IBM Romania (company)
11. IBR (The Romanian Banking Institute)
12. Info World (company)
13. Magic Fox Media (company)
14. MCTI (The Romanian Ministry for Communications and IT)
15. Microsoft Romania (company)
16. Montero Soft (company)
17. NetConsulting (company)
18. Prosoft (company)
19. S&T (company)
20. SoftNet (company)
21. Softwin (company)
22. STARTX (company)
23. TotalSoft (company)
24. UNITECH (company)
25. The Slavic Languages faculty of the University of Bucharest
26. Wizrom Software (company)
27. Zando Computer (company)

2.3. How the meeting went

Several participants have reported that the actual meeting, which took place on March 25, was well organized and civil, and that heated outbursts were few and restricted to the debate portion of the discussions.

Not all comittee members were present. 5 votes were issued later, over email, during a 24-hour grace period.

2.4. Vote distribution

There were 26 votes (the entity holding the president chair does not vote, as per ASRO regulations):

• 15 in favor
• 6 against
• 5 abstentions

5 of the votes were not expressed during the meeting but later, over email, during a 24-hour grace period. It has been made public knowledge that these 5 votes were distributed as follows:

• 2 in favor
• 1 vote against
• 2 abstentions

2.5. Committee joining chronology

In 2007, before the whole OOXML affair began, there were 12 members in CT210. In March 2007 there were 27.

Two of the following 15 are outspoken ODF supporters and OOXML critics, who have been arguing their cause and trying to prove that OOXML is not fit to be a standard since long before September 2007. They had misgivings about joining CT210 especially for the vote, but felt justified since they had been deeply involved with this issue for a long time. They applied for joining the committee on February 13 (Zando Computer) and March 7 (Startx).

Genisoft applied for membership on November 15, 2007.

The rest joined in late March, the week before the vote, as far as we know.

Please note that ASRO is not offering this kind of information (or any information at all), so we depend on the members’ own claims on this subject.

Interesting fact: about mid-wave, ASRO stopped informing already joined members about new members as they joined. The last 6 joining members and the fact that the count had reached 27 were only revealed at the meeting.

2.6. Relations between CT210 members and Microsoft

What would happen if one were to question the impartiality of the CT210 members who voted in favor of OOXML?

Let’s say, for instance, that one would check to see what kind of relation exists between those members and Microsoft. That person would discover that:

• GECAD NET, Info World, Montero Soft, NetConsulting, TotalSoft and Wizrom are Microsoft Gold Partners.
• The ARIES president and GECAD NET owner are the same person: Radu Georgescu.
• Genisoft is a Microsoft Certified Partner and its representative, Ciprian Jichici, is a Microsoft Regional Director for the Timişoara region.
• Prosoft is a Microsoft Registered Partner.
• Dimsoft is an Authorized Education Reseller.
• S&T is a Large Account Reseller and according to their website they and Microsoft “have mutual synergies concerning Microsoft platform and server solutions”.
• Cătălin Gheorghiu (I Computer Solutions) and Zoltan Herczeg (member of the Developer & Platform Evangelism department with Microsoft Romania) often mention and support each other’s points on their personal blogs. I Computer Solutions offers quite a lot of services based on the Microsoft platform.
• IBR’s IT courses mention the word “Microsoft” a lot. They say that these courses offer “essential notions regarding computer use: the Windows operating system; office applications (Microsoft Office); Internet services.”
• HP and Microsoft: Microsoft Gold Partner, Enterprise Partner of the Year for the 4th time (in 2007; the only time in the award history a company has received it more than 3 times), officially support adopting OOXML.
• Do not forget Microsoft itself.

For the sake of completeness, I must also mention that SoftNet and Softwin, who voted against adopting OOXML, are Microsoft Certified Partners, and IBM Global Services is a Gold Partner.

My sincere thanks to Cătălin Gheorghiu for explaining the ever-changing nature of the Microsoft partner program to me and advising me to double-check the program membership of the CT210 members for up-to-date status using the Microsoft online catalogs. I can therefore state that the above partnership status comes from Microsoft itself. Feel free to look yourself in the Microsoft AER and Partner catalogs. (Note: these relations may expire at some point and these companies may not appear in the catalogs later. I have snapshots of the catalog webpages and will provide them upon request.)

2.7. The vote results

Who didn’t vote: CIMEC (Institute for Cultural Memory). The representative, Dan Matei, acts as CT210 president. According to ASRO procedures, his vote is kept in reserve and only used if the vote cannot otherwise be decided one way or the other.

I have always wondered about this rule. Basically, what it says is that ‘abstain’ was never an option for Romania!

Who abstained: the Ministry for Communications and IT, The Slavic Languages faculty of the University of Bucharest, Magic Fox Media, Fujitsu Siemens Romania and the Romanian Academy.

No surprises among the abstentions. I expected academic institutions and the Ministry to abstain. They were members of CT210 since before all the madness began and they have joined because they were actually interested in CT210’s regular work. They are also more likely not to be so involved with business world politics. Magic Fox Media’s representative, Cristian Secară, is a valued contributor to the Romanian character coding standards and also an old member of the commitee. As for Fujitsu Siemens, well, I guess big companies have more than just Microsoft to consider.

Who voted against adopting OOXML: IBM Romania, SoftNet, Softwin, Startx, Unitech and Zando Computer.

IBM, Startx and Zando Computer have been vocal opponents of OOXML. Softwin (while a Microsoft Certified Partner) is a long-time CT210 member. If their vote was dictated by a refusal to compromise their past work for the commitee and/or actual consideration of the arguments against OOXML, I applaud them.

The only odd goose here is SoftNet. When a Microsoft Certified Partner joins the commitee at the last moment you expect them to vote ‘yes’ for OOXML — based on similar scenarios in Romania and all over the world. For 10 other last-minute comittee members who are also Microsoft Partners, you’d be right. Hats off to SoftNet.

Who voted in favor: everybody else.

Who changed their vote: 4 members. Back in September 2007, of the 12 members 1 didn’t vote (president), IBM voted ‘no’, the Ministry abstained and the 9 others voted ‘yes’. Of these 9, 2 have now changed their vote to ‘abstain’ and 2 to ‘no’.

2.8. ISO notification

Reportedly, the result was sent to ISO as soon as all the votes were out, even before the result was announced to the public. Hot off the stove, so to say.

Why, I wonder. From what I understand, there was no real need to do this since Romania’s vote hadn’t changed. Perhaps to ensure that the affair was truly done with?

3. Conclusions

3.1. Committee stacking, ballot stuffing

In September 2007 there were 12 members in CT210. At the moment of the March 2008 vote there were 27.

About a dozen new members joined at the very last moment, the week before the vote (the details are murky, since ASRO is not offering this information officially). 10 of those voted in favor of OOXML.

All 15 members who voted in favor are in the Microsoft Partner or AER programs, or otherwise have a strong business or technological relation with Microsoft (or are Microsoft itself). They are all small or medium companies, or organizations or institutions strongly bonded to Microsoft technology.

By cross-referencing the 15 Microsoft-involved yes-voters with the late-comers we get 10 votes in favor of OOXML that arrived at the last moment, and all from Microsoft-biased entities.

Look at these facts and tell me that everything was fine, that there was no committee stacking, no ballot stuffing or whatever you want to call it.

If it wasn’t for those 10 votes, the vote would be 5 in favor, 6 against and 5 abstentions.

And if ASRO would have taken India’s example and not allow voting on this issue for members who weren’t already on CT210 for the first round, the result would have been 5 in favor, 2 against and 4 abstentions. That’s because CT210 was traditionally already populated with several Microsoft-friendly entities due to ongoing work on better Romanian language support in Windows.

Ironic, isn’t it? That a favorable result for OOXML could have been achieved anyway, without a shred of suspicion about tweaking the adoption process?

3.2. ASRO: a secret society?

After the vote took place, myself and other TIC-Lobby participants have publicly asked for the following information:

• The list of CT210 members at the date of the vote.
• Which ones were present at the meeting and who voted during the 24-hour grace period.
• How each member voted.
• What rules are used by ASRO to determine its official position based on the vote distribution.
• When ASRO’s decision will be transmitted to ISO.

On March 27, in two separate messages to TIC-Lobby, Dan Matei, CT210 chair, has provided the list of the CT210 members and, several hours later, how they voted.

However, this happened after he started March 27 by citing ASRO regulations that prohibit tehnical committee members from disclosing any such information. (He also questioned the need to disclose such information at all, in particular the identity of the CT210 members. He seemed to fear some kind of witch hunt, which to me implies acknowledgment of something wrong. But that’s another story.)

Other comittee members have also resorted to generally being very careful with their public statements. The reason? They were unsure of how the ASRO confidentiality rules applied to the situation at hand and what the fallout would be should they choose to break said regulations.

Which begs a very interesting question: why is ASRO so big on confidentiality? It’s an institution that discusses public standards. There aren’t any industry secrets or matters of national security or whatnot being discussed. If you want ISO standards you simply buy them, and anybody can become a technical committee member if they act as a juridical person and pay a fee per year and per CT. So what exactly are they protecting?

As a matter of fact, throughout the OOXML saga, ASRO showed less than stellar transparency and has shunned any form of direct communication.

It’s important to remember that Dan Matei, the one who acted as a public spokesman for CT210 through all this, is only another member of the committee. He holds the president chair for CT210, but he went beyond the call for duty by basically being ASRO’s voice in the matter and the buffer between ASRO and the public. He didn’t have to, he did it voluntarily (and I’m afraid that in the process he has taken some flack that was actually meant for ASRO).

That’s because ASRO itself has been extremely quiet and uncommunicative all these months. The ASRO technical advisor delegated to CT210 (Teodor Stătescu) has flat out ignored emails sent to him, or offered responses which evaded the issues put to him. The ASRO website is void of any relevant information. Their email address is not working most of the time. For all intents and purposes, ASRO has perfectly isolated itself from the public and has insisted on imposing secrecy and isolation on CT210 members as well.

The way I understand it (and I Am Not A Lawyer) ASRO seems to put its internal regulations above the laws of the State it’s very existence is based upon; one of which states: “the following basic principles must be followed in national standardization: […] transparency and public availability”.

3.3. ASRO regulations: bent to suit

The voting procedures in ASRO technical commitees in general and CT210 in particular remain a bit fuzzy. It appears that a simple majority vote is conducted and whatever position scores more votes wins (the president gets to veto the outcome, but only if the vote cannot be decided through normal proceedings.) This is how the March 2008 vote was decided.

However, that’s not what ASRO’s own rules are saying. My translation:

1.7 Consensus: general coming to agreement, defined as the absence of sustained opposion regarding an essential aspect of the subject, coming from an important part of those interested and as part of a process that involves taking into account the points of view of all interested parties and unifying diverging stands. NOTE: consensus does not necessarily mean unanimity.

So from what I understand, there was no consensus reached during the March 2008 meeting of CT210. And the same regulations state that any standard discussed must have consensus before it’s approved. ASRO broke its own regulations here, yet they are quick to insist on other parts of them, such as confidentiality.

I remind you that the September 2007 and March 2008 votes were the first and only occasions that CT210 has ever had to vote like this. Their usual work involved getting out working specifications, so they usually worked on something until they were all pleased with it. No voting was involved.

It remains unclear at this time why ASRO used simple majority when its own regulations and the ISO regulations require more than basic vote counting.

3.4. The bottom line for the OOXML affair

The public debate on TIC-Lobby in the months leading up to the March vote have shown a staggering amount of arguments against adopting OOXML as a standard. There were links, tehnical details, technical experiments.

The counter-arguments have been weak, obstinate and often repeated. Even when begged to do so, OOXML supporters have failed to put together even the smallest document containing arguments for their position.

The facts give strong indications that CT210 fell victim to committee stacking driven by commercial interests; that ASRO bent its own and ISO rules to facilitate the outcome; that ASRO has been very uncommunicative throughout this and tried to impose silence on CT210 members as well.

I guess that puts Romania in good company, from what I’ve seen all over the world.

I have come to see how ASRO does things only in the last few months and only from CT210’s perspective. I do not know if the way things were done there is a trademark of how ASRO does all things. But what I’ve seen is not exactly flattering.

At the very least, the lack of proper organization, the obvious bias, the tendency towards secrecy were all quite disturbing.

But this is the institution that the Romanian State is confortable entrusting its standardization to: a body who can act in a completely opaque manner should it choose to and defy the spirit of the laws, if arguably not their letter. A very strange and disturbing notion, but there you have it.

Perhaps the biggest loser in all this is CT210. It was a fine and hard-working technical committee until Microsoft decided it had to have a certain vote out of it no matter what. I’ll be curious to see, a few months down the road, how many of the current 27 members will actually get involved in the committee’s work. Because you can join with an agenda and you can pay a one-time fee, but you can’t fake actually doing some work.

3.5. What have we learned?

In these past months, the manner in which Microsoft understood to take advantage of National Standardization Bodies in Romania and all over the world has had many effects, of which only some are good for Microsoft itself.

In Romania, this way of acting has prompted people to react, to get organized, to become more aware of the issues involving ODF, OOXML and Microsoft. On a smaller scale than it could have been possible or I would have personally liked, but it happened nevertheless. And please note that Romania is chronically deficient when it comes to its citizens demanding and upholding their civil liberties. These things considered, it wasn’t bad at all.

The benefits of this grassroots movement are many. The TIC-Lobby archive is now the repository of a detailed story, documenting how FUD is born and how it dies at the hands of facts and logic. It is also the repository of a great amount of anti-OOXML arguments.

To CT210, this movement has managed to provide enough evidence and arguments to sway 4 members (who have previously voted ‘yes’ out of inertia) to ‘abstain’ and ‘no’ votes.

It has now become more clear that Microsoft is a ruthless entity that will stop at almost nothing in order to protect itself and its way of doing business. That fact that they were perverting the proceedings of a global organization (ISO) and destroying its credibility in the process did not stop them. Everybody, beware. It is time to serve back a statement that was once made by Microsoft about Linux: that it is a cancer. For me, today, the cancer is a corporation that spans the world and from which nobody is safe when they stand in the path of its personal gains.

We have also found out that our NB, ASRO, is in the best case miserably incompetent and in the worst corrupt and biased. I will give it the benefit of the doubt and assume it is simply a case of another ISO branch being utterly unprepared to deal with Microsoft tactics.

Was it all worth it? Let’s hope so. Because, should OOXML be rejected after all, it would have all been for nothing. The entire world now seeing Microsoft for what it is, and nothing to show for it.