The situation and rules for the Ballot Resolution Meeting (BRM) are starting to become clearer. As you may know, this is a follow-up to the September 2 vote where the Microsoft/ECMA OOXML (“Open” “XML”) specification failed on both measures to become an ISO/IEC JTC1.Given the way the system works, this means that we get to go on and try something else to make it a standard. And if that doesn’t work, try something else. Repeat until either the opposition goes away, gets bored, or no one remembers what you were trying to do in the first place (grin).
Anyway, the specification document was over 6000 pages long and got thousands of comments from around the world. Some of these were basic (“perhaps you need a comma here”) and some are potential show stoppers. The Ballot Resolution Meeting that will be held during the last week of February in Geneva will try to resolve as many of these as possible.
[This next paragraph was corrected by BRM convenor Alex Brown in the comments: there will only be this one BRM and probably no intervention by officials to kill the process.]
The representatives from the national standards bodies who attend may not have enough time to get through everything and so it is possible that the meeting might need to be continued at a later time. Or if proper resolutions to each comment are not made, the standards officials may decide that it will not be possible, and the whole process may be terminated, indicating failure to achieve a standard.
My understanding is that many of the “easy” comments are being worked on now and some of the national standards bodies have been engaged by the OOXML supporter(s) to start making them happier. Here’s something that’s important to remember: if you have 100 complaints and I fix the 80 most trivial ones, leaving 20 really hard and serious ones, I don’t get to claim that I’ve made you happy because I fixed 80% of your problems.
This is an example of the “weakest link in a chain” story. It should be the case that the proper resolution of all the hardest and most technologically challenging problems is key to this. If there is one critical thing that is so important that it must get fixed and it does not, then it does not matter how many other things were fixed. That is why this is such a daunting situation to overcome.
My further understanding is that Microsoft/ECMA will give their preliminary resolution of the comments to the national standards bodies in the third week of January, leaving slightly more than a month for them to make sense of them before the BRM. The people to whom I’ve spoken do not think that all national standards bodies will send representatives to Geneva. Indeed, only so many of them will fit in the room, and so the number of active participants will be limited. (Note that is not the same as what happened in Portugal.)
I’ve mentioned this before as well, but suppose that in the original vote in a country, 9 people voted against OOXML and only one voted for it. This means that 10% voted for it. If only two representatives can go to Geneva, it would make mathematical sense that you do not send one that is for OOXML and one that is against. That is significantly disproportionate to the original vote. Expect people to be tracking this sort of arithmetic. This is true as well, of course, if the original vote had gone the other direction.
The final fate of OOXML will not be decided at the meeting, just the production of a disposition of comments report. After that, and this is where it gets funky, the original voting countries will have 30 days to decide if they want to change their votes from what they cast this last summer. So expect the supporter(s) of OOXML to go all out in March to try to get the many NO votes to become ABSTAINs or YESes.
That said, I really expect this thing to drag out for a long, long time through 2008.
So it’s Microsoft’s ‘wall of money’ in one corner of the boxing ring, against … well … IBM has a wall of money, too, but isn’t very likely to deploy it all … so, perhaps, all the people who got together and decided that ISO26300 was the way to go, in the other corner.
By next March, there could be significant deployment of ISO26300-based solutions. ‘Trick-or-Treat’ (tonight), Christmas, and Valentines are traditional days for giving of large quantites of cheap gifts in the USA; might I suggest that http://symphony.lotus.com/ would qualify ? 4th-of-July, too, but the next one of those isn’t until after the vote.
There’s also http://www-306.ibm.com/software/lotus/notesanddomino/ , somewhat higher up the value chain, and priced accordingly. Same standard for what it keeps on your Personal Computer’s disk, though.
270 million Americans can’t be wrong. We just have to get IBM Lotus Symphony into their hands, and we’re done.
Bob hi
ISO/IEC will very shortly be publishing a FAQ on the Ballot Resolution Process which addresses some of the topics you have raised. I expect it to contain a few of points which bear on what you’ve said:
- it’s true the venue has a limited capacity (approx. 120 people). But there is room for a delegation from EVERY country that voted in the letter ballot. If restrictions have to be made, delegations will be reduced in size; countries will not be excluded.
- I don’t believe “standards officials” have any authority to abort the process.
- The BRM will last for five days, period. No re-convening. If NB’s think that is inadequate, their recourse is to vote against the DIS. The process will be terminated (one way or the other) 30 days after Feb 29th.
- The 30 days after the BRM in which votes may be changed are indeed “funky”. I expect there may be lobbying of all kinds in that time!
- Alex.
Alex, thanks very much. I’ll amend my entry and mention this comment as well.
Right on cue, Chris, as Jack Loftus reports on the increasing number of applications adding ODF support in the past two months:
http://enterpriselinuxlog.blogs.techtarget.com/2007/10/26/odf-alliance-hails-record-growth-application-support-for-odf/
IBM’s Lotus Symphony led and way along with Firefox allowing users to view ODF files in their browser. I’m betting a big “Hell freezing over” before Microsoft allows IE7 to do the same, even though you can view MS-OOXML files in it.
We had the local Rotary Club http://www.rotary1110.org.uk/ pitching in the cafeteria today. I gave them a copy of IBM Lotus Symphony http://symphony.lotus.com/ , marked up with its IBM price ($0) and with a suggestion that IBM might be willing to enter into some distribution deal. Usually, corporations do pay those who distribute their advertising, but he’d have to enter negotiations with the part of IBM that buys things. IBM Global Procurement. SWG/AIM is the wrong bit.
Rotary can reach places that IBM cannot. But, equally, Rotary as an organisation would be torn apart if they were to favour one corporation over another.
The Rotary guy said that Microsoft give them a pile of money to help in their mission to eradicate Polio … we don’t have a particular problem with polio in the UK, but in Africa it’s a very different story. http://www.gatesfoundation.org/default.htm have a wall of money, too, and that gives a certain amount of freedom.
So, we’re straight into the usual terms-of-reference problem. Do the Rotary distribute anyway, like the IBM Lotus Symphony web site says is to IBM’s commercial advantage ? Or does Microsoft’s money block the distribution channel ?
Huge forces involved. Very interesting times.
I have a question. Since I understand that only technical issues will be addressed at the February meeting, may I ask what qualifies as a technical comment? Specifically, what if a comment was sent in by a national body member regarding interoperability? Does that qualify as a technical comment? Where, exactly, is the line being drawn?
Also, who exactly is working on the comments, collecting them and cateogorizing them and addressing them? I know ECMA, but more fine tuned, please.
Is it time for pushback?
National bodies have an opportunity to change their vote — this not only means that bodies that voted ‘no’ can change to ‘yes’ — if those who ‘provisionally’ voted ‘yes’ don’t have their issues addressed, they can also be convinced to change their vote to ‘no’. (unless, of course, I’ve misread the rules).
Well, pj, it’s a bit like this.
IBM is IBM, and wants high prices for XBox chips http://www.ibm.com/developerworks/library/pa-nl19-calendar.html
Microsoft is Microsoft, and wants low prices for XBox chips http://www.xbox.com/en-US/ .
The actual price is commercially confidential, but what we can say for sure is that a dollar either way will transfer huge sums from the pockets of Microsoft investors to the pockets of IBM investors, or vice-versa.
I expect it’s the same with technical issues, ISO, and ECMA-376. The line will be pushed around, and it will be pushed around for commercial reasons. You won’t be able to please everybody all of the time.
“The Rotary guy said that Microsoft give them a pile of money to help in their mission to eradicate Polio”
Was it Microsoft that gave it? Or the Gates Foundation? They are not the same thing.
Hi.
Just thought we should chip in here. we are hosting a site: http://www.dis29500.org for the community to be able to delve deep and actually highlight the “issues of substance” for the BRM. We are aware that ECMA have stated that this process has been started – but in a closed fashion. This is open. we do not restrict anybody from choosing to contribute. We just want to make sure that in the very limited time available and the BRM, that all the cruft, duplicates and non-substantial issues can be cleared away before the meeting.
Anyone is free to visit and start filtering. All the comments have been separated and split by country and reference to the spec already and we have had quite a bit of activity (nearly 500 comments reviewed) in identifying dupes and non-issues. Anyone who is interested – please drop in!
Thanks
Alan
It was the Foundation .
You’re right, they aren’t the same thing, but the money all comes from the same place. They are somewhat connected.
IBM too has a ‘charitable’ arm http://www-05.ibm.com/uk/news/ondemandcommunity/ondemand_qa.html too; if I spend 40 hours a year helping out a school or charity, IBM gives the school or charity a piece of technology kit; whatever they want, usually a something from these guys http://www.lenovo.com/us/en/ but if they wanted something else they could ask.
It’s considered part of IBM Sales and Distribution; I presume another form of advertising, just as commercial as all the other things that IBM does. Revolutionary, though.
Looks like OS/2 is still chugging along http://www.os2world.com/content/view/15421/1/ slowly into the sunset. “eComStation Release 2.0 RC3 available”. Faithful servant, but unlikely ever again to be high-growth.
Linux does its job, for IBM. Knocked out by the forces of commoditisation.
I wonder if anyone will get Lotus Symphony http://symphony.lotus.com/ to work on it ? By all accounts OpenOffice.org http://openoffice.org/ is there.
Pj: Comments, issues and changes are usually divided into Editorial” (think typos), “Technical” and out-of-scope (which are usually rare, but for the recent ballot, numerous. A technical change involves some change (broadening, restricting, altering) in semantics such that might have some effect on implementations. Editorial changes would have no effect on implementations. An out-of-scope comment would be one that was not based on something that could result in an instruction to the editor. Any major technical change would expect hard scrutiny, especially if the standard related to an existing external technology (for the obvious reason that such as standard has to be about what is, not what isn’t).
This distinction is carried through in the maintenance process for standards, where there is a “Corrigendum” system that allows editorial fixes to be put through fast, while technical changes need extra care for review. Of course, where there is some o-my-god realization about a new standard, sometimes committees will try to be optimistic in what fits into an editorial change, but there still must be a ballot.
On interoperability, a standard for product conformance might indeed look at that issue. DIS 29500 is not a product standard but a document standard: like other SC34 standards (and like ODF) its primary objective test for conformance is to validate the XML sub-documents against the schemas to make sure the names, positions, structures and possible values of the document are correct.
Typically application conformance is a matter of national testing bodies, such as NIST; however these organizations often require specific client demand (i.e. governments putting their money where their mouth is) to develop and maintain these kind of suites.
To give you an idea of the difference between a document standard and an application standard, consider that neither ODF nor OOXML (nor HTML) gives specific details on general typesetting (how to break pages, lines, words). And application standard or test suite might indeed specify that kind of thing.
It is delegated to the editor to generate generate a disposition of comments document, which gives the editor’s preferred options (based on the editor’s judgement, committee deliberations, informal discussions, and so on) and this is usually used as the practical basis for subsequent discussions: the editor makes a good faith stab at resolving as many issues as practical.
For an example of a Disposition of Comments document from SC34, see http://www1.y12.doe.gov/capabilities/sgml/wg8/document/1934.htm which is one for the committee draft of ISO HTML. ISO HTML was tricky because of course it needed to conform to an externally specified and deployed technology in order to be useful, but it had to be restricted in ways to promote best practice. Note, for example, AUS-6 where the Australian comment (which would have been made by me from the year, but I have no memory of it) is rejected because of “expert advise from W3C”. The ball then comes back to the NB to decide whether they then agree with the disposition of comments, judging the whole thing as a package and not getting caught up on individual hobby-horses.
For an example of a large Disposition of Comments documents, see http://jtc1sc32.org/doc/N0201-0250/32N0237.pdf which shows Jim Melton’s diligent editing of the ISO SQL standards. This has about 520 pages of responses to comments (plus a few hundred for a version of the draft) . Note in particular on the third PDF page, where it says
“1548 comments resulted from the ballot, of which 527 were categorised as Major Technical, 658 were categorised as Minor Technical, 193 were categorised as Major Editorial, and 170 were categorised as Minor Editorial.
The Editing Meetings managed to resolve all of these 1548 Ballot Comments. A number of Ballot Comments required that any additional problems discovered during the editing process also be resolved. At least 118 of these were discovered and resolved. Some papers submitted discovered and resolved more than one problem but these have not been separately counted. I estimate that around 1700 comments and problems were resolved in total.”
Cheers
Rick Jelliffe