The ISO approved ISO 29500, otherwise known as Microsoft OOXML the other day. This is of course a travesty as the ISO fast track process has been soundly hijacked and the format is not ready for the big time as it is currently unimplementable by anyone other than Microsoft and maybe Novel.
Do read on to find out not only what’s been going on, if you haven’t been following too carefully, but also why this is important to all our digital futures. And if you feel it is important, feel free to use it to construct your own letter to you appropriate representative.
As you may be aware, Tineke Egyedi the president of the European Academy for Standardisation is highly critical of the approval process and believes that the standard is contrary to WTO Technical barriers to trade, and that the EU’s regulatory body has sent queries to several countries investigating irregularities in the approval and voting processes.
I am concerned that this investigation needs to be conducted with extreme vigour, I have been following the process closely and the simple fact of the matter is that Microsoft has conspired to buy an ISO standard which does not begin to meet the requirements for which it is designed and is, in fact, unimplementable by anyone other than Microsoft and is therefore designed to give the illusion of being an open document format whilst maintaining their monopoly in the office software market. The EU investigation has been directing questions to the various persons who are complicit in buying the process rather than the technical bodies behind them who’s true position has been circumvented.
Let me next expand on the background of what the OOXML format is and is supposed to do. The OOXML
standard is an XML based document format that is supposed to allow any person to reliably read and write data in the format and reliably render the document as it is intended to be seen. The standard was developed by Microsoft under the auspices of the ECMA , an allegedly independent organisation, under the name ECMA-376. This was in fact an almost entirely Microsoft owned project. The EXMA-376 format was then pushed through the ISO fast track process by ECMA on behalf of Microsoft.
The goal of OOXML is laudable and very important. In our increasingly digital world, the need to be able to exchange data reliably is very important for business to business communication, business to consumer communication and government to citizen communication. It is not simply a question of being able to open a document in Word Processor A that someone has written with Word Processor B (though that is very important as no one should be bound to the market leader/monopoly holder for any piece of software) but that any other application (including web based services) can also interpret data within documents, produce them and send them between various entities without having to rely on any patent or proprietary third party software. Indeed this is the foundation of Tineke Egyedi’s objections to the format. OOXML does, apparently, contain Microsoft patented methods which it has supplied with an Open Specification Promise however this does not (and indeed seems to carefully exclude) permit any implementation released under the popular open source GPL license. And it is not just open source implementations that should be considered under threat.
That businesses, individuals and governments are able to freely exchange information with each other is already of vital importance and can only grow more important as more and more of our business and government services are conducted via computers. In Scotland where we continue to try and develop the "silicon glen" end of our economy, I feel it is vitally important that international standards are available and usable so that we can freely join and compete in the global marketplace.
Of course it would be wonderful if there already were an ISO approved document standard that could fulfil the objectives of OOXML that Microsoft could freely implement in their software. And of course there is; The Open Document Format (ODF) which is approved as ISO standard (ISO/IEC 26300). There has been a fair amount of Fear, Uncertainty and Doubt (FUD) spread about the ODF all of which can be easily dismissed, and it has been implemented in a range of applications:
- Sun StarOffice Office Suite
- Sun ODF Plugin for Microsoft Office
- IBM Lotus Notes / Symphony
- Corel WordPerfect (beta)
- Apple Mac OS X Leopard / TextEdit
- AbiWord / Gnumeric
- Google Docs & Spreadsheets
It has also been implemented, via a third party plug-in for Microsoft Office, however this is not distributed with Office nor, when installed does it get added to the "Save As…" menu option without manual modification of your Office installation.
Let me now tackle the deficiencies in the competing OOXML standard itself. There are a couple of online resources that document the technical deficiencies with the standard, GrokDoc compiled a list to the specification as it stood in January last year (some of which have since been addressed) and DIS29500 has a
compilation of objections drawn up by ISO National bodies including 635 from the UK . Identifying problems with a specification and correcting them is, of course, an important part of the process and many have been addressed although some of the more esoteric instructions like "Format as Word 95" remain without being properly defined and are therefore unimplementable unless you have access to Office source code or assume that the Open Specification Promise is sufficient to prevent you from being sued should you try by any other method.
The culmination of the ISO approval process was the BRM meeting which the ISO declared a resounding success.The story from inside the BRM, which was supposed to analyse all of the remaining objections to the specification and resolve them, is somewhat different. Standards Malasia, for example complained that 70% of their comments were not addressed [PDF] and in short there was insufficient time to tackle all of the comments on file. Andy Updegrove covers the details of what happened here and here. A delegate from Brazil was so disappointed by the process that he broke several rules about non-disclosure in his blog post on the subject. The upshot was that the vast majority of comments that needed to be addressed were bundled up into a single approval vote instead of being properly discussed and vetted.
After the BRM the member bodies were given until March 29th to either confirm their vote given at the BRM or change it to Yes, No or Abstain. This final stage of the process is where the real skulduggery has taken place. The Open Malasia blog has covered the issue there excellently however the four locations I shall concentrate on are Poland, Germany and the UK as these are EU member states and on Norway as the vote rigging example there is extraordinary.
In Poland the national body met to vote on approving the standard. Of 45 members, 24 were present. Of those present, 12 voted for, 10 against and 2 abstained. With the vote unresolved the chair person elected to receive further votes from the absent members by email. However any non-replies were counted as "yes" votes and some members were thus not aware that they had voted yes. polishlinux.org has the story here, here and here. They also cover an interesting admission by Jason Matesow of Microsoft that they manipulated the process.
In Germany members were actively prevented from voting "no" and had their votes counted as "abstain" instead in order to carry the vote over all as "yes" this was a result of wilful misinterpretation of the ISO rules. Andy Updegrove again has the details and the rules in question are here where there is no mention that a vote can not be changed to "no".
In the UK the BSI Technical Committee IST/41 which is in charge of the UK process recently changed their overall vote from "no" as cast at the BRM unexpectedly to "yes". There really is very little information available about what exactly has happened here, the committee has worked hard on trying to correct problems with the standard (remember those 635 comments?) but consistently said that it should not be approved. The sudden last minute change of heart is unexpected and unexplained. I have no evidence of wrong doing however suspicions have been voiced that skulduggery has been going on. I believe it would be an excellent idea for the EU to request information on what exactly happened during the meetings where this last minute change of heart was agreed and to ask if they are satisfied that all 635 of their comments have been properly addressed.
Finally the true scandal of the process in Norway; the technical members of the committee overwhelmingly voted against the standard and recommended that Norway should vote "no". Once these members were out of the room however the vote was changed by the remaining members unilaterally. This has prompted Steve Pepper the chairman of technical committee to submit a formal protest about Norway’s vote to the ISO. This Norwegian article’s headline is "Scandal in Standards Norway" and that is no exaggeration.
To summarise, the entire ISO fast track process has been abused not least in the post BRM meetings I have highlighted but with bullying and bribery along the way plus character assassination of opponents in the process and Microsoft’s own material suggests that the best way to influence decisions is to buy the chairman. Any EU investigation of this process should certainly investigate if any chairperson on any National Body was unduly influenced.
Were all the technical problems with OOXML to be properly resolved, and the various issues around patents – like the Open Specification Promise being extended to explicitly allow any implementation of OOXML regardless of the license the implementation were released under or the removal of MP3, a patent encumbered audio format as the default for embedded audio then there is no reason (other than the question of why a second ISO document format is required) that OOXML could not become a usable document interchange format. As it is, this is not the case. The only people who can actually completely implement this standard is Microsoft itself, it has carefully bought an ISO standard so it can claim to meet the requirements governments are, quite rightly, setting about interchangeability and open standards but also protecting it’s monopoly in the Office software market which can not be tackled with fair competition unless there is a truly open standard in use.
I hope that this summary has made clear the fairly tangled web of what has been going on here. I have had to summarise quite brutally, there have been many more questions around the legitimacy of the process from its very inception which I have not had time to cover. For further information I recommend Groklaw’s excellent summary page and especially their chronology of events in the saga. Andy Updegrove has also been closely involved in these events and covers them on his website as does Rob Wier of IBM on his website.
My objective was to let you know the extent to which the ISO process has been hijacked in order to force the approval of a standard that does not meet many of the basic requirements such a standard should have and makes something of a mockery of the recent fines and mandates imposed by the EU on Microsoft for other uncompetitive behaviours not to mention putting the ISO itself potentially into disrepute, in the hope that you will pass on my concerns to the relevant persons within the EU charged with investigating this. I would very much like to hear your response on the subject and for you to keep me informed of any developments in the EU investigation into this matter – there is precious little about it being published that I can find.
Thank you for taking the time to digest all this, if you have any questions about anything I have mentioned or outlined I will be happy to clarify them for you. I look forwards to hearing from you.
Jamie A. Thom