Friday 31 August 2007

Linux Beer Hike

I will be present at the Linux Beer Hike on 2-9 September in Crete, also enclosing an FFII board meeting on 7-8 September. Time for me to finish writing the minu... I mean the agenda. Will see if I can get some pictures up here next week.

SiS confirms: Sweden abstains from voting on OOXML

I just received confirmation from SiS that Sweden will not vote on OOXML. This is because scheduling a new meeting on such short notice (less than three weeks) would have required unanimous approval, and this was not reached in the working group. Therefore, a new vote is not possible.

Update 17:05: SiS has issued a press release (Swedish).

Thursday 30 August 2007

SiS retracts its OOXML decision!

SiS just published a press release saying that the decision on Monday is annulled, and Sweden will likely abstain from voting on OOXML, due to procedural issues. More info as soon as I've catched up.

Update 21:15: partial translation: (may contain mistranslated terms)

Office Open XML - SiS invalidates the vote
The Swedish working group of SiS, Swedish Standards Institute, Document description languages, SIS/TK 321/AG 17, decided on 27 August 2007 at a vote to vote for making Office Open XML an ISO standard. Today the board of SiS decided to invalidate the vote.

The motive of the decision of the board is that SiS has information suggesting that one of the members in the working group has participated in the vote with more than one vote. [...]
Update 21:25: the head of SiS says in IDG that this had nothing to do with the massive criticism, only a procedural matter. I say no more... ;-)

Microsoft acknowledged OOXML problems

On 15 June, at one of the first meetings of the SiS OOXML working group, presentations were held about ODF and OOXML, respectively. (The OOXML one was one of those "black is white" shows - after seeing it you'd be tempted to believe that all the problems were a good thing.)

Anyhow, Microsoft was present and a consensus was reached that ODF and OOXML are in conflict! (I didn't participate in the particular meeting, but I learned this later.)

I was also told that Stephen McGibbon from Microsoft had mentioned that the EMF / WMF issue is going to be corrected. An e-mail I received from him confirms that he "told the SIS meeting that this is an error in the specification that will be corrected at the BRM, and that Open XML has no requirement on EMF or WMF".

So he acknowledges that this technical problem (though I wouldn't be surprised to see a denial of this categorization) needs fixing. Therefore, we must vote NO, right?

Well, SiS had apparently (un-)done some homework in time for the vote, because they then - despite criticism - no longer seemed to support the earlier assessment that technical problems must be dealt with in order to vote YES (which is how I read the JTC1 directives, section 9.8)...

Now, with 170+ comments, one would think Microsoft would have some counter-arguments, no? Well, I talked about many of the comments on 14 August, with Microsoft present. The working group was given an answer to one or two, the rest Microsoft would answer "via e-mail". Interestingly, on 16 August the same representative pointed to a stack of paper, claiming he'd been working for (only) two days with it, answering all the comments. He didn't want to share them at that time, but said he would "consider" it later.

That was the last I heard about it, until 27 August. Apparently the response had been available since 24 August on LiveLink, a SiS / ISO system for communicating documents, but the working group was not notified (this 3-day gap is not the interesting part, though). One could wonder why he didn't send the information 8 days earlier, when clearly he could do so.

On 27 August, he claimed that "final edits" had been necessary before sharing it, and (since I even asked why they didn't send a draft) I can only conclude that - according to Microsoft - a preliminary response was somehow impossible or of low value to the working group. If this is indeed their opinion, I agree wholeheartedly. They had nothing to say about most comments except "it's wrong", just like in Norway where they hijacked the decision with no actual comments. (Update 3rd September: Norway finally voted no with comments.)

Wednesday 29 August 2007

Highlights of the SiS meeting on 27 August

The SiS working group received vast amounts of comments on OOXML, from several parties that had paid the required fee of 600 SEK / 65 EUR (yep...). All of them but one also submitted a recommendation for a resolution to OOXML; 27 companies suggested a YES vote, but did not write even one sentence to justify it. 2 companies that said YES made very brief statements - just as meaningless.

9 companies said NO and (in most cases) presented conditions for approval in appropriate detail (most were at least an A4). In mid-August, we finally concluded a list of 173 suggestions, not counting some general argumentation or the fact that some comments were collapsed into one.

Now, somehow the idea was presented on Monday that this "clear majority" in favour of OOXML as-is should help us in our decision, and that it's important to look at who submitted these comments; I get the feeling that they didn't buy my foolish thought of considering the comments as such (which, as it happens, were basically non-existent on the YES-side). It was also hinted that this approach would be justified (statistically or otherwise), in comparison with e.g. the Eurovision where anyone can call in... (Lordi, anyone?)

Anyway, then there were the thoughts of "diversity" as a reason for many standards, and that "we're not ready for a single language (Esperanto)", wherefore "we aren't ready for a single standard"...

I especially liked one of the companies' presentation:
  • ECMA is good (so there)
  • The interests of the customers is important (I hinted that maybe, possibly, it's Sweden's interest that's relevant (at the very least), please forgive me)
  • There's no time to deal with technical deficiencies (so we just ignore them? Convenient!)
  • There will always be problems, no standard is perfect (this is a good thing, right?)
  • The fast-track procedure has to be safe (therefore it is, Q.E.D.?)
  • It's about taking control from Microsoft (no matter such details as Office 2007 doing something else entirely, or the patent problems, or the un-implementability of OOXML for anyone)
  • We say yes to ODF too (therefore there is no problem with adding another standard)
I think that about concludes it; most YES voters didn't bother to justify their stance or to at least suggest that even one of the 170+ proposals need to be sent to ISO.

The only interesting speech at the meeting was given by Georg Greve of Google, who held a very thorough presentation of various problems. Their position is also available as a PDF document.

NyT: "Microsoft admits SiS voting coup"

Summary from an NyT article: (roughly)
Microsoft admits that it's behind the SiS voting coup that resulted in proposing the OOXML document format as a standard.

- Mistakes have been done at our end, says Klas Hammar, Microsoft.
Klas goes on to say that this action was done by a "individual employee" and that the action "was not authorized" by Microsoft.

I think this speaks for itself...

Update 17:25: the following people from Microsoft were present at the SiS meeting on Monday:
  • Jonas Persson, technical director
  • Klas Hammar, business area director
  • Peter Centellini
I'm sure they were all "individual employees" and had no clue what happened at the meeting...

IDG: "Microsoft put pressure on partners to vote YES"

IDG just published an update on the Microsoft scandal.

Quick and dirty translation of the summary:
Microsoft offered extra 'market subsidies' to partners that participated in the Monday vote about the Open XML format. This appears from internal communication that CS has seen. 'It was badly formulated and would never have gone out' says the business area chief of the company, Klas Hammar.
Update: a partial English story is available at OS2 World.
Update 14:18: Groklaw also has a story.

Microsoft's Office coup in Sweden

On Monday, 27 August, the Swedish Standards Institute (SiS) declared its coming vote for the "Office Open XML" (OOXML) standard proposal initiated by Microsoft. All such ISO participants in this matter must vote no later than 2 September.

The working group that recommended this decision to SiS originally had 12 members, where a NO vote was very likely. However, on the meeting appeared 23 new members, most of them Microsoft partners. (Many of those became members even when the meeting was about to start.) As a result, a YES vote was enforced.

This is just a brief update to existing information available from e.g. the FFII Sweden press release and countless blogs and articles.

I was present at this meeting and just received confirmation from SiS of the voting results: (please see the FFII PR for some more details)

Vote results: 25 YES, 6 NO, 4 members absent.

New YES votes: (19 members since 23 Aug. 2007 or later)
  • Camako Data AB
  • Connecta AB
  • Cornerstone
  • Emric AB
  • Exor AB
  • Fishbone Systems AB
  • Formpipe Software
  • FS System AB
  • HP
  • iBizkit AB
  • IDE
  • IT-Vision AB
  • KnowIT
  • Modul1
  • Nordic Station AB
  • Sogeti
  • Solid Park AB
  • SourceTech
  • TietoEnator

New NO vote: (entered 23 Aug. 2007)

  • Google

Old YES votes: (6 members)

  • Diamo AB
  • EPiServer
  • HumanData
  • IAMCP Sweden Chapter
  • Microsoft
  • WM-Data Sverige AB

Old NO votes: (latest member was registered on 9 Aug. 2007)

  • Illuminet
  • Kungliga biblioteket
  • Riksarkivet (chair)
  • Sun
  • Verva
Members leaving the meeting before the vote: (all but IBM (old member) filed applications after 23 Aug. 2007)
  • Cybernetics (Microsoft Gold Partner)
  • IBM (NO)
  • Readsoft AB (Microsoft Certified Partner)
  • Strand Interconnect AB (Microsoft Gold Partner)
I have some more info on the meeting etc., I will try to publish this shortly.