The international standards body ISO has offered to help maintain the ODF document standard alongside its work on the rival Microsoft-originated OOXML specification, saying its creator Oasis is not dealing with defect reports quickly enough.
At a meeting in Korea last week, the International Organization for Standardization (ISO) committee for document standards, SC 34, issued a liaison statement to Oasis, the body that created ODF. It requested an "alignment" of maintenance of ODF between the work done at Oasis and that within ISO.
According to a member of SC 34, this move was prompted by fears ODF was not being updated properly by Oasis. However, it has been described as a 'coup', aimed at putting ODF under the control of a body dominated by OOXML supporters, by critics such as lawyer Pamela Jones, writing in the Groklaw blog.
Delegates from 15 national standards bodies unanimously agreed to send the message to Oasis, because they are unhappy with its work maintaining ODF, said Alex Brown, convenor of WG1, a working group within SC 34. These nations include strong ODF supporters such as Brazil, India and South Africa — countries whose national bodies objected to the approval of Microsoft Office Open XML (OOXML) as a standard, Brown said,
ISO put the SC 34 committee in charge of standards for 'Document description and processing languages'. That means it has been instructed by JTC 1 — the top of the ISO hierarchy — to handle both the ISO standard based on OOXML (known as ISO/IEC 29500), and that based on ODF (known as ISO/IEC 26300), Brown wrote in an email to ZDNet.co.uk. "We did not ask for that responsibility; we have been given it: JTC 1 has included ODF maintenance in our work programme", Brown said.
Brown said Oasis has not been acting on reports of defects in ODF from standards bodies, some of whom are in the process of applying the document format as a national standard. "When flaws are found in an international standard, nations… want to see the problem fixed pronto," he said. "Now they find the mechanism they thought they had for this (Oasis via SC34) does not appear to function. Their defect reports are being shelved."
The ISO move has come in for criticism: "I call it a takeover attempt of ODF," Groklaw's Jones wrote. "If the takeover were to succeed, SC 34 would get to maintain ODF as well as Microsoft's competing parody 'standard', OOXML." The ISO committee is dominated by Microsoft supporters, Jones said.
But Brown says Jones's post is "chock-full of misinformation and spin". ISO/IEC standards have to be managed and published by ISO/IEC committees, even if they are created elsewhere, and ODF cannot have a free ride any more than OOXML, he said. "Ironically, during the OOXML furore there was much talk of ISO 'rubber-stamping' OOXML," Brown noted. "Now the dust has settled, it turns out the rubber-stamp accusation is more justly levelled at ODF."
At the same meeting in Korea, SC 34 applied tighter control of OOXML, Brown said, writing the private standards group Ecma, which first promoted OOXML as a proposal, out of the process of standardising it and giving it only a secretariat role. Objections to ODF getting the same treatment would harm the ODF standard in the long run, he said.
The SC 34 committee has declared it is "open to suggestions" to reach a "mutually acceptable" resolution.
Whatever the future path of the standards, the OOXML adoption process is still causing trouble. Thirteen members of the 23-strong document standard committee at Norway's standards body, Standard Norge, have sent an open resignation letter. They protested that Norway decided in favour of OOXML, when only two members of the committee voted in favour of doing so. Standard Norge followed 37 identical letters from Microsoft partners, instead of the advice of its own expert committee, according to the resignation letter.
Meanwhile, the OOXML standard has been leaked, breaking copyright statements.
Oasis has not made any official response to ISO's move. It appears, however, that the body will welcome ISO involvement, even though some ODF supporters may object. Patrick Durusau, ODF Editor and a member of Oasis's ODF technical committee, welcomed the regularisation of ISO's management of the OOXML standard. He also warned that at least one ODF supporter, IBM, might resist efforts to apply the same strictures to ODF.
"An early objection to ISO/IEC 29500 was that Ecma was going to [maintain the standard]," Durusau said. "Having lost that objection, IBM wants to avoid discussion of the same question for ISO/IEC 26300. Sauce for the goose isn't sauce for the gander, at least at IBM."
Talkback
"Delegates from 15 national standards bodies unanimously agreed to send the message to Oasis, because they are unhappy with its work maintaining ODF, said Alex Brown, convenor of WG1, a working group within SC 34. These nations include strong ODF supporters such as Brazil, India and South Africa — countries whose national bodies objected to the approval of Microsoft Office Open XML (OOXML) as a standard, Brown said,"
That's an interesting comment, knowing that
"At this meeting, Brazil was represented by Microsoft. I and several other members of our committee protested against this indication, but we had our protests silenced by the Director of ABNT (Brazilian NB) that said that this decision was not our prerogative and that if Microsoft would pay for their own representative’s trip, their indication was approved." (see http://homembit.com/2008/10/meeting-of-jtc1sc34-in-korea-it-is-the-end-of-the-world-as-we-know-it.html).
It would be interesting to know how many of the 15 national standards bodies were represented by Microsoft employees, but ISO do not publish this information...
"Sauce for the goose isn't sauce for the gander, at least at IBM."
The same can be said for ISO, as the "sauce" used at ISO for ratifying ODF was the PAS ("Publicly Available Standards") path, while OOXML used the Fast Track path.
And indeed, PAS and Fast Track provisions for maintenance of the resulting standard are not the same...
"Brown noted [...] it turns out the rubber-stamp accusation is more justly levelled at ODF."
In the case of ODF, it is not an accusation, it is a fact: the purpose of the PAS procedure is precisely to rubber-stamp Publicly Available Standards already reviewed by another standard body, without extensively rewriting them, as it was needed for OOXML.
I think that what is happening here is that when a group bends the rules, for whatever reason, they sometimes find that they subsequently suffer from a lack of credibility, especially when it comes to activities in that same area. Clearly a lot of people believe very strongly that the process was abused in approving OOXML. To have that same group now come back and want to take over responsibility for the alternative standard, would require considerable trust and faith in the best of circumstances, and that is obviously lacking in this case.
jw 7/10/2008
I thought, with the fiasco regarding OOXML and Microsoft, that ISO had shot their creditability in the foot. It might have only exposed procedures that quietly went on for years but when the approval of any standard hinges not on the standard but on the procedural processes surrounding the decision making, it all becomes a bit pointless and irrelevant.
That's a good point, Luc, that I will be taking up with people.
PAS specifications are created and completed outside, and can be rubber stamped by ISO, but Fast Track specifications expect ISO to complete the specification, so they should not be rubber stamps.
This whole thing is complex, or did I say that already?
"Fast Track specifications expect ISO to complete the specification, so they should not be rubber stamps."
I would not say it so. Indeed, these ISO procedures are very complex, probably much more than they should. Here is my vision of the whole story.
PAS and Fast Track have more or less the same objectives: having an existing standard "co-opted" by ISO. Most standard organisations have to use the PAS process, for which a number of criteria have to be met. Fast Track has been designed specifically (by and) for Ecma, and has no entry criteria.
If the system is not abused, Fast Track has basically the same result as PAS: a text extensively reviewed by a standard organisation, stable and with several implementations is endorsed by ISO as being an International Standard. So, rubber-stamping is not an issue. Accordingly, the Fast Track process is very lousy about the review process, because it is not designed for this.
The problem with OOXML was that it had not been seriously reviewed by a standard organisation, was huge and far from stable, and had not a single implementation released when it was send to ISO by Ecma for the Fast Track process. This was clearly an abuse of the process, made possible because of the lack of entry criteria.
So, rubber-stamping this text, as requested by Ecma (and indirectly Microsoft) was a problem. The obvious solution was that ISO asks Ecma to use the normal track, where a clear and proven review process is described. But despite the claims of many National Bodies, ISO refused to do this.
ISO then had to invent (interpretation of) rules on the fly, that resulted in an extensive rewrite of the 6000+ pages text to be done in a hurry. The result is that nobody implements the current text, and Microsoft warned that they have so many changes to be done in their MS Office implementation that it will not be completed before the next major release, for which no date is available yet (probably sometimes in 2010).
The ISO goes away and sorts OOXML out first. It certainly sounds as if it's got enough on its plate already without taking ODF on too.
Not just loss of credibility. As for the idea of them taking over the successful standard - ODF - one immediately thinks of foxes and hen-houses.