Report on Standards Australia industry forum on DIS 29500

by Rick Jelliffe

Australian national body Standards Australia had an industry forum today on Open XML. The agenda and invitation for this is up at Tom Worthington's website.

The Invitation


Here is the most interesting part of the invitation:

This forum is being conducted by Standards Australia as a courtesy to stakeholders. It is an extraordinary meeting that we are not required to hold, but do so to provide an open process. We appreciate your attendance and expect that you appreciate our effort in making this opportunity available to you.

Standards Australia values its vote as a participating member of all international committees, and does not exercise it injudiciously. We provide considered Australian viewpoints that are beneficial to Australian stakeholders, including industry, government, academia and the general community, through the facilitation of trade and the inclusion of clear Australian requirements in international standards.

The JTC1 process has established that the ECMA-376 document is not contradictory to existing standards and ECMA has responded to a number of technical considerations raised in the initial consultation period. This forum is not to debate the merits of the JTC1 decision making process or the validity of the ECMA response.

While technical comments are welcomed, it would be entirely counter productive to use this forum to reiterate technical comments that have already been raised and are likely to be debated in every JTC1 member body in some form.

We are looking for creative, positive contributions that emphasise our commitment to representing truly Australian views to the international community.


More on that later.


The Speakers



The meeting had 30 to 35 attendees (I didn't count, oops), based on the membership of existing technical committees and people who had sent in comments to the process so far. It was not a voting meeting at all, just a meeting to help consensus and to give more information to higher committee members. (However, participants can submit comments by Aug 21 for consideration by the Australian CITeC Standards Sector Board.)

The meeting was a three hour affair, with the first half invited speakers and the second half question and answer and commentary.

The first half started with an introduction by Standard Australia's Alistair Tegart, who provided good strong chairmanship that left most people frustrated that they had not had a chance to say more, but which gave everyone a chance to make their most important comments in the allotted time. The interesting thing was that discussion of technical minutae was strongly discouraged (wrong meeting for that), which is a nice break for me. Discussion was civil, everyone friendly in the coffee break, and frank in the meeting.

I had been invited to speak on the subject of General overview of the standards process because of my involvement as Australian delegate to (what is now) SC34 in the 1990s and my continuing involvement with standards. A nice comment afterwards (by a law professor!) was that mine was the only talk with new content. I tried to present an SC34-based perspective on standards: what SC34 standards are, how the preference for enabling standards rather than applications has been overtaken by the fast-track process, the basic standards posture for Australia in SC34 in the mid-90s (need for simplicity to suit our small development teams, I didn't mention support for regional neighbours though it was important) and how each different country has different requirements. (For example, some countries have a requirement that they do not want to be blocked out from international contracts because of the lack of standards.)

Then a quick mention of some of the issues that I prototype in this blog: that ISO standards for documents are voluntary, that standards form a library of choices, that the mere existence of alternative standards does not prevent any group from choosing one over the other, that standards such as PDF and Torx are not open in the sense of allowing arbitrary change but nevertheless valuable, and so on. I emphasized again that the ISO process is a win/win system in which attempts by one group to stymie another's needs does not fit.

That took about 15 minutes, then there were speakers on the case against the adoption of Open XML (the scheduled IBM speaker was hospitalized so we were treated to an emergency podcast from Rob Weir which was basically the same content as his Technical Case against OOXML.) and for the adoption: a quick tag team with an MS representative, then the local CompTIA representative, then a CEO. The CEO, Richard White from CargoWise EDI, was particularly forceful on how it would help his business.

The Discussion



Then after coffee we had over an hour of moderated discussion. By and large it went as expected: people from local industry welcomed it as solving a real problem, people from business rivals of MS didn't like it, people who identified themselves with Open Source didn't like it, people from academia or standards bodies seemed to think that having it as a standard would somehow force them use it (I didn't get this.)

I had to gag myself a few times. The local Google Maps operation was represented, but I was quite surprised to hear Lars Rasmussen say how difficult it would be to implement Open XML...surprised because he had told me last year how Google maps used VML to deliver to IE and how format was simply not a problem. (Here is the first line sent for a Google map, for confirmation: note the namespace declaration and stylesheet reference:


<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" xmlns:v="urn:schemas-microsoft-com:vml">
<head><meta http-equiv="content-type" content="text/html; charset=UTF-8"/><noscript>
<meta http-equiv="refresh" content="0; URL=http://maps.google.com.au/?output=html"/><<noscript>
<title>Google Maps</title>
<link rel="stylesheet" type="text/css" href="http://www.google.com/intl/en_au/mapfiles/86/maps2.css" />
<style type="text/css">body{margin-top: 3px;margin-bottom: 0;margin-left: 8px;}
#vp {position: absolute;top: -10px;left: -10px;width: 1px;height: 1px;visibility: hidden;}
#homestate {display: none;}
v\:* {behavior: url(#default#VML);} ...


It seemed strange to be saying that a technology was too big to implement when you were in fact using that technology successfully. Maybe the Google speaker didn't realize that VML is a part, though obsolescent, of DIS 29500. I think what happens is that "implement" gets stretched to mean "implement all the parts of a specification": so "It is too big to implement" means "It is too big to implement it all" in Googlespeak. But Australians aren't going to implement a full new office suite. It is too big, even if you just used ODF; and Open Source people will more naturally join the existing Open Source and Free projects rather than set up new ones, it seems to me. For Australian requirements, "full implementation from scratch" is an imaginary and spurious requirements.

What has maybe slipped Lee's mind is that most integrators will use DIS 29500 in the same way that Google Maps would be using it: just cherry picking the parts that are needed (in their case, the subset of VML.) And, in particular, when you are using it as an end-format, you only need to "implement" (i.e. generate) the elements that correspond to your input. Not the whole thing.

When I was talking to the local Google people last year, they told me that Google doesn't actually have any fulltime people allocated to standards work in general. I gathered that was a little pedestrian for them, because they made their money by innovating not by following the pack: sounds like a recipe for QA disaster to me. I don't know whether their foray into web-based applications will make them a little more savvy with standards.

Another Google guy (who turned out to be a ring-in: Georg Greve, initiator and president of the Free Software Foundation Europe, who gfim says was flown in from Switzerland by Google especially for the occasion) stood up and recommended we should track what the Indian standards body's concerns about binary mappings. Again I had to gag myself (actually, Alistair did it for me) because I believe I actually was present at the meeting in Delhi where that issue was raised: by the Indian representatives of Sun and IBM. I am afraid I couldn't help thinking this was the classic Colbert "Echo Chamber" effect, similar to Wikipedia's astroturfing: one member of a collective puts something up in one forum, then other members of the same collective bring up the first as independent evidence. (In this case with the added twist that Sun and IBM were not mentioned: a lay listener could easily have had the impression that this was some kind of position adopted by the BIS, whereas, as far as I know, it has not so been. I hope the Georg will be a little more careful with attributions in the future, because people can so easily get the wrong impression.)

An interesting comment from local IT29? committee head, Jamie (surname illegible sorry), that for educational users, they needed to guarantee interoperability and could not force students to purchase particular programs. I didn't uite see the logic of how this meant that DIS 29500 should not be adopted at ISO. Marcus Carr from Allette Systems (who I consult for and teach standards seminars by) sits on a local IT committee too and responded that students would be better served by PDF if guaranteed interoperability were the issue.

A few comments later, I got a chance to mention that none of the XML formats today provide guaranteed interoperability in the sense of visual fidelity, for the reasons that readers of this blog will be familiar with: every application supports different feature sets, has different fonts, hyphenation dictionaries, kerning tables, line break algorithms, and so on. Plus the formats are extensible, so can have all sorts of strange media types. Standard documents may perhaps be a necessary condition, but they certainly are not sufficient. What is needed are profiles, which restrict the features, requires certain application behaviours and require certain fonts. And for uncramped page designs that reduce the chance of page overflow on different systems.

Several speakers raised the issue of IP rights and worries about the MS Covenant Not to Sue and the Open Specification Promise. MS gave the usual response: we have run it past lots of external lawyers who say it is fine, and since OSP is so similar to Sun's equivalent, why don't you have the same concerns about ODF. I think Standards Australia has been playing a little coy here, because they are trying to be scrupulous not to be seen to take sides, I guess. I had asked them in email to have a clear position on standards and IP from the legal perspective, and they ended up saying, in effect, that for Standards Australia, it is JTC1s responsibility and competency to evaluate the IP issues of drafts submitted for Fast-Tracking, and not a technical issue for voting.

I think a better and more complete answer would be better. People who are interested in this are should first read the excellent webpage by OASIS lawyer and anti-Open XML conduit Andy Updegrove, especially on the Allied Tubemakers and Dell cases. Standards don't exist in a vacuum, and MS standard's participation and the very strong and constant statements that MS would be considered in any court.

Another aspect of the IP discussions is that typesetting and desktop applications are not now a new thing. With a 20 year limit, patents before 1987 have expired, which is well after the invention all of the basic ideas in office suite software. (Last week in Thailand, MS' Oliver Bell was asked a question on this issue, and IIRC he said that actually MS only uses its patent portfolio defensively and has never sued on IP. Does anyone have a list on this?)

Jamie ? pointed out that participation in a standards body does not nullify the IP; however, the issue is the scant chance that submarine patents are enforceable. So add the two covenants, external legal opinion, vetting by Ecma and ISO/IEC JTC1, the age of likely patents, the recent stronger court awareness of junk patents, the difficulty of enforcing submarine patents, the multiple statements made by MS executives and staff to the highest level, and the basic fact that a document standard is more concerned with schemas and general description and no of methods or algorithms, and I don't know how much more would be possible to satisfy someone.

One comment mentioned the idea that MS covenant not to sue etc does not cover external technologies. Of course, not, but that is no different for ODF and HTML.

Other speakers brought up a few of the usual suspects. autoSpaceLikeWord95 made its scheduled appearance, along with statements that made it clear that the speaker had never read the spec and was parroting. There is an easy way to tell a parrot in this area: they will say something like "The standard is full of compatibility elements like autoSpaceLikeWord95 which are undocumented and prevent implementation". In fact, there are 64 compatibility elements, and IIRC correctly all but two are adequately document with explanations of their general functionality. AutoSpaceLikeWord95 is optional and is clearly marked as deprecated: it seems to be a warning flag that some document was originally created by Word95 with this bug and it had never been corrected. The bug is related to the treatment of Fullwidth character used in East Asian typesetting (zenkaku): certainly for Australian users it is utterly extraneous to our national requirements.

The issue of the definition of various functions in SpreadsheetML came up too, from a localization perspective. (Again, irrelevant to Australia.) If the moderator hadn't been so tough, I would have liked to have asked whether the speaker wanted to remove or rename the existing function (and break everyone who used this function's spreadsheets) or merely to add better localized functions (which belongs in a maintenance phase.)

On the issue of maintenance, I did get another opportunity to spout. I said that it is too early to tell what effective systems for maintenance will occur. When OASIS and ECMA submit their standards, they also submit information about how maintenance will occur. There would be collaboration with JTC1 SC34, for example. I said I thought this was only practicable for fast corrigenda (which don't add functionality just fix the text and clear mistakes) and that the approach that OASIS seem to be taking, which would involve resubmitting ODF 1.1 and ODF 1.2 etc for fast-tracking each time, was probably the more realistic thing to expect. However, I noted that DIS 29500 has a quite strong extension regime, indeed a whole part (Part 5) and starts from a much more complete position that ODF: so one would expect complete updates to be rare events, perhaps aligned to the three-year product cycle.

The main Google guy at some stage made a good point about overlapping standards, along the lines that having multiple standards for programming languages was OK because the differences could be justified, but he had not heard arguments why Open XML was so different from ODF that it could be justified.

A few others also had comments that could be fairly reduced to "We don't need it, therefore we do not support it becoming an ISO standard, therefore we appose it becoming an ISO standard" which is a non sequitur.

Baseline formats and downstream formats



A very interesting point was made by the National Archives representative. They don't have the resources to cope with Open XML and ODF, he said, so they would adopt ODF for their future format and didn't support Open XML becoming a standard. Again, I don't see how the standardization of Open XML forces their policy in any way. Standards Australia is not even a government agency, and has no legal clout on the National Archives: moving to ODF where possible seems a reasonable choice (well, ODF 1.3).

Marcus Carr objected to this. He spoke from the perspective of document processing from the early 90s, and the difficulties in practice of dealing with Word documents (with the various hijinks: converting .DOC to the Rainbow DTD, converting .DOC to RTF then processing that, etc) and brought up the key processing issue that I think almost all the commentators on Open XML miss. He brought up the issue of the need for a full-fidelity baseline schema to allow the most flexibility in downstream processing.

Now this is a pipeline approach that has proven itself to work over the last 15 years we have been using it. Elephantine readers may remember a blog of mine a year ago:

A typical strategy when converting from XML into some structured text format is to have three transformations:

* first, convert the XML into ideal XML: resolve links as needed, remove extraneous elements and attributes, convert cases, generate headings and other things that need to be generated
* second, convert that ideal XML into an XML-ized version of the output format
* third, convert the output XML into the text format, delimiting and indenting as needed


If the input data is non-XML, then we have an additional stage where we first convert the data into an "baseline" XML format that maintains all the information from the data source (it could be a database, another format, a binary, no matter.) You never know what information you need, and you don't want to trust someone else's abstraction but work with as unmediated form of the data as possible.

So Marcus' comments was that we (the system integration and document processing community need Open XML as an ISO standard, because it alone provides an adequate baseline format for subsequent transformations. So the Australian National Archives could well decide to archive data using ODF, but they may well decide to implement their conversion to ODF by going through Open XML. So one standard would be useful for one purpose (saving future archives), the other standard for a different purpose (opening existing files in the archive.) If the Australian National Archive is moving to ODF 1.0 fast, I hope they don't throw away the original binaries...

Breadcrumbs


So all in all, I think the day was a worthwhile exercise, and a good opportunity to help us all escape groupthink.

I suspect, from the tone of the invitation and comments made at the meeting and elsewhere, that when Standards Australia looks at the comments that people send in (deadline August 21) they will be completely disinterested in comments that question JTC1 decisions and comments on issues that have no local relevance. I gather they may not be much impressed by arguments that can be refuted by precedent: for example, that there should be no overlapping standards. However, we shall see, and I don't know anything about the CITeC Standards Sector Board.

The most enlightened part of their approach, and I think this is pretty novel, is that Standards Australia seem very aware that the role of an individual standards body in vetting a standard when there is a multi-national campaign to discredit it (on the one hand) and promote it (on the other) changes the requirements for a review. In the case of a normal standard review, you raise as many (sensible) flaws as you can, because you don't know whether the issue will be addressed by anyone else. In the case of a global campaign, it is clear that almost every National Body has been mail-bombed with the same speil and that therefore those are issues that we can actually ignore, unless they have a clear national significance because we know that other national bodies will be examining them. I think that is what may be behind the last line of the invitation
We are looking for creative, positive contributions that emphasise our commitment to representing truly Australian views to the international community: they want to husband their resources to what is important for local industry and local requirements. They don't want to succumb to a Denial of Service attack where by concentrating on sorting out edge cases and typos they miss out the big picture of national interest.

I'm preparing my comments to the CITeC Standards Sector Board at the moment, and I will put them online here too, if anyone is interested.


11 Comments

Benjamin Henrion
2007-08-09 15:13:22
"In the case of a global campaign, it is clear that almost every National Body has been mail-bombed with the same speil and that therefore those are issues that we can actually ignore, unless they have a clear national significance because we know that other national bodies will be examining them."


I think the issues the National Bodies can ignore is those standard letters sent by Microsoft puppets:


http://www.noooxml.org/forum/t-11857/microsoft-puppets-spamming-ansi


For your information, I suspect Microsoft is doing that also in Australia, as they were looking for companies to send such template letters to SA.

Rick Jelliffe
2007-08-09 18:32:11
Benjamin: I think you misunderstand: the mailbombing I am talking about is the anti-Open XML one. As I have traveled around the world, it is really obvious that the same talking points are brought up again and again by people who don't seem to have actually read the standard and are parroting.


You can see the same thing in comment responses to blogs, where someone will, after some token sentence say "Do you know that..." or "I found out that..." then parrot one of the Grokdoc claims, often utterly tangential to the thread topic. After a few hundred times, this kind of thread hijacking gets tedious.


Form letters are really honest, because they just show that the person concurs with an opinion, and the recipient does not need to progress specially with the technical details. To call it astroturfing, as some have, is insulting to the writers and our intelligence: astroturfing is a kind of fraud, while a form letter is clearly someone joining a campaign.


Furthermore, for anti-Open XML person to complain about form letters smacks of self-rightousness, I am afraid to say. For example, the NOOXML site gives a list of canned responses as "arguments that can be used to formulate your submission". When someone sends in a cut-and-paste email, it has as much effect as a form letter, because its original content is the same.


When the Standards Australia body said "creative" and "truly Australian views", they were rejecting both canned material, whether in the honest format of a canned letter or the faked up format of a cut-and-paste letter.


Rick Jelliffe
2007-08-10 12:32:04
UPDATE: Some other reports from the meeting:


Lee Welburn (Groklaw)


gfin (Groklaw)

(These provide some names I missed, which I will correct in the blog.)


Andre
2007-08-11 17:00:17
"The most enlightened part of their approach, and I think this is pretty novel, is that Standards Australia seem very aware that the role of an individual standards body in vetting a standard when there is a multi-national campaign to discredit it (on the one hand) and promote it (on the other) changes the requirements for a review. In the case of a normal standard review, you raise as many (sensible) flaws as you can, because you don't know whether the issue will be addressed by anyone else. In the case of a global campaign, it is clear that almost every National Body has been mail-bombed with the same speil and that therefore those are issues that we can actually ignore, unless they have a clear national significance because we know that other national bodies will be examining them."


LOL.


That's a funny defense.


Don't point out the flaws because others will do.


Silly.


If Australia does not point them out they will get away with it.


We have the ECMA specification of Open XML. Microsoft wants something from ISO: approval. So they will need to make offers to improve the shit.


For an approval without comments a nation as Australia would get nothing in return. Microsoft does only answer pressure.


As I wrote somewhere else: "Free hugs for Microsoft" is no rational strategy.

Rick Jelliffe
2007-08-11 23:14:22
Andre: By "ignore", I didn't mean "ignore for all time", I mean ignore as a priority for ballot comments. It is a waste of review effort to merely go over the same issues as every other body rather than start with your local concerns.


Indeed, National Bodies need to make sure that they are not spending all their time hijacked on the talking points of large closed-source multi-nationals and people who are not even potential users of the standard.


The first thing that is involves is understanding the big picture (what are the goals?, what are the precedents?, what are the legal checks that ISO makes?, what is the maintenance regime?, what are your particular national requirements?, what is required for conformance and how is it tested?, what skills are available for review so that our national body makes its distinctive contribution?, etc.).


The second thing is to understand the technology well enough to be able to know what are edge cases; if you look at the differences between the early Grokdoc material and the material that Patrick Durusau put before INCITS/V1, you can see that the former had no idea of what an edge case is, while the latter clearly does.


But it is the former's issues that get parroted: all these claims that OOXML is bad because developers cannot implement things like "autoSpaceLikeWord95" when the documentation clearly says it is not intended for implementation, indeed deprecated, because it merely signals that the original file was made with a version of software that had a typesetting bug or problem that might affect the layout. It is a stretch to say that a document format needs to document bugs.


That "autoSpaceLikeWord95" affects only East Asian fullwidth characters. Few people outside China/Korea/Japan have any idea that such alternative alphabets/syllabaries exist, let alone use them. It has nothing to do with their national interests, and their National Bodies only waste their time by looking autoSpaceLikeWord95 once they realize it is a CJK issue; instead of trying to learn East Asian Typesetting, a National Body should press on and look at problems that affect their documents. For China/Korea/Japan, it is a different matter: they have a national interest. However, it is still an edge case.


Certainly, it is completely inconsistent to say "We cannot have OOXML because it does not define line-breaking alogithms" and then say "We like ODF not OOXML" when ODF does not define any particular line-breaking algorithms too.



2007-08-12 03:56:57
/*The second thing is to understand the technology well enough to be able to know what are edge cases; if you look at the differences between the early Grokdoc material and the material that Patrick Durusau put before INCITS/V1, you can see that the former had no idea of what an edge case is, while the latter clearly does.*/


Grokdoc is a wiki collection. The essence was to table "technical comments". See
http://www.noooxml.org/how-to-communicate-with-standard-organisations
http://www.noooxml.org/arguments


We have two sides of the debate: technical and political. The petition is for the political angle.


/* But it is the former's issues that get parroted: all these claims that OOXML is bad because developers cannot implement things like "autoSpaceLikeWord95" when the documentation clearly says it is not intended for implementation, indeed deprecated,...*/


Why not say "undocumented"?


/* It is a stretch to say that a document format needs to document bugs.*/


full specification.


/* That "autoSpaceLikeWord95" affects only East Asian fullwidth characters. Few people outside China/Korea/Japan have any idea that such alternative alphabets/syllabaries exist, let alone use them. It has nothing to do with their national interests, and their National Bodies only waste their time by looking autoSpaceLikeWord95 once they realize it is a CJK issue; */


We are international. Take FFII's Hartmut Pilch from Munich as an example, he is translator for asian languages. Asians are everywhere, Muslism are everywhere, companies enter these markets. ISO is about international standards.


/* instead of trying to learn East Asian Typesetting, a National Body should press on and look at problems that affect their documents. For China/Korea/Japan, it is a different matter: they have a national interest. However, it is still an edge case. */


It's not.


/* Certainly, it is completely inconsistent to say "We cannot have OOXML because it does not define line-breaking alogithms" and then say "We like ODF not OOXML" when ODF does not define any particular line-breaking algorithms too. */


A standardisation process is about raising issues. There is no such thing as an "ignore comments" campaign. It was only invented recently.


Microsoft or ECMA messed it up. They proposed an immature standard and now they want to get all valid technical comments ignored. It is plain nonsense to say that CJK is only a CJK issue. Apply that rule to ISO standardization and all procedures will overrule CJK interests and fail to produce international standards.

Rick Jelliffe
2007-08-12 08:24:17
Anonymous: But at ISO, there is no "political" only "technical". People who go at it as a political campaign are in for a nasty shock.


All this hype, the inflation of the importance of the contradiction phase, the ridiculous lessening of definition of "contradiction" to include anything that might cause confusion, the bogus invocation of parts of international law, the name calling, the automatic knee-jerk reaction that any committee that does not vote the correct way must have been bribed, the edge-cases being treated as showstoppers, all these kinds of tactics are futile. They wast my time and they waste your time.


I don't know FFII's Hartmut Pilch: is he a typesetter with knowledge in this area? A translator qua translator has no expertise relevant to evaluating the impact of this flag, any more than a person who buys a meal in a Chinese restaurant, with no disrepect to Herr Pilch.


The standardization process indeed has a phase about raising issues, but 1000 people raising 1 issue each is better than 1,000 people all raising the same issue. Especially if that issue is an edge-case. Surely there must be lots of text improvements that can be found. The thing I was most disappointed about the Standards Australia meeting was that, even though it was not a technical meeting, when any participants used examples, it always was the same examples. I ended up suspecting that few of the attendees had even read much of the spec; it is incredible arrogance to oppose or support a technical spec on a public forum without having more than browsed it.

Rick Jelliffe
2007-08-12 08:30:22
Anonymous: "Its not"


A deprecated element, in a group of 64 compatibility elements that are optional, in a spec where conformance is syntactic and does not require any semantics be adhered to anyway, that reports that the text may have been originally layed out so that some characters in one or two's countries' documents had some visual problem, related to spacing, because they were created in 1 version of an old application--a visual problem that a modern application would not duplicate--, that is not an edge-case? Pulease!

Anonymous
2007-08-14 23:47:56

Rick, I see some disingenuousness in your posts here. You accuse the anti-OOXML group of a lot of actions that the pro-OOXML group are just as guilty of, if not more so. How about a bit of balance?


You claim the same talking points are being raised repeatedly. Not all points are equally important and it's a typical bureaucrat's response to complain about minor points while allowing important points of principle to be drowned out by the minutiae.


One principle at stake here is that at a bare minimum SA should only be giving their imprimatur to standards usable by the entire community. It is disingenuous to claim that nobody in Australia will implement all of the word processor standards. People in Australia will certainly be parts of teams that implement substantial parts of these standards, including modifying existing software packages over the years.


Another principle is to not have mutually conflicting standards. Despite their complaints Microsoft could easily implement ODF in full, far more easily than any other party can implement OOXML in full.


Don't be a bureaucrat, implementing complexity for the sake of complexity. There is very little valid technical argument for having a second standard. If Microsoft have any problems with ODF nothing is stopping them proposing and implementing changes to ODF that are compatible with the community at large. They have already had plenty of opportunity and have chosen not to do so. That in itself speaks volumes about their respect for the ISO/SA standardisation process.


If there are edge cases that stop the standard from being fully implemented then there should be no problem removing those edge cases, right? If you can't remove those edge cases, edge cases that favour one vendor over another, then maybe they're not so "edge" after all. You can't have it both ways; either it's an unimportant edge case that can be removed to create a truly portable standard or it isn't. Which is it?


You claim that responses have been political rather than technical. Of course, the entire point of OOXML is political, to give Microsoft favoured status on government and other contracts with an ISO/SA approval only Microsoft can fully and easily implement. The claim that ISO/SA giving their seal of approval to Microsoft's format will somehow allow ISO/SA to "control" it is silly at best. Microsoft will quite happily implement the standard and then add "optional" defaults (such as non-portable fonts and operating system dependencies) that insure the vast majority of OOXML documents are non-portable. They have a long history of doing this. Have you checked the OOXML standard for the ability to do such things?


You claim that most responses haven't read the OOXML standard. Of course, the standard is thousands of pages, possibly the longest that SA has ever looked at. The comments timeframe is entirely inappropriate for the study required and another piece of dishonesty is pushing it through the process while ignoring the fact that it is excessively long.


You claim to have been "mail bombed". Different people and organisations use different avenues for expressing their opinions. Microsoft spends some of that $50B/year they earn to employ "handlers", create nice presentations, give junkets and publish spin. You should be bending over backwards to avoid being manipulated in that way and accept the fact that community input, as unpolished as it is, is just as important.


You quote "local concerns". Local concerns include all the global concerns that affect the local community, including implementability, interoperability, completeness and everything else. Global concerns do not disappear in a local context and framing the debate to ignore them does everybody a disservice.


Please, step back a little. While the technical detail is of course important there is also an important political dimension to this standard. SA not unreasonably does not want to be treated as a political football. Unfortunately, by even considering OOXML when an ISO word processor format already exists, SA has taken a political action and should not be surprised at the response. Microsoft and SA should be proposing changes to the ODF standard as needed, not proposing a new, incompatible one-vendor OOXML standard.


Since the reason for OOXML is almost entirely political if SA approves OOXML as an Australian standard don't be surprised if SA loses some standing in the community as a result. Please think carefully about all aspects of the standard, including things other than the technical minutiae, and whether it should exist at all.


Rick Jelliffe
2007-08-15 00:52:43
Anonymous: I don't think there is a single sentence in your comment that I agree with! Well done!


Standards are decided on technical issues (after the IP and market requirement issues are sorted out, at the beginnning, and after the drafting issues are sorted out, by the end), not purported political or economic claims. (Even then, the kinds of technical issues that are appropriate for consideration for an external (fast-tracked) specification are more limited than those appropriate for a standard developed or collated in-house by SC34.)


he reason that I support DIS 29500 becoming a standard are firstly that my business and the business of my Australian sector (industrial publishing and markup) would benefit from it (indeed, we have been calling for this for years), and secondly that the ISO rules need to be administered fairly (if ODF can get in, so should OOXML.) (B.t.w., if I thought that ODF was technically ready to be used as a baseline format for conversions from legacy Office documents, I probably would not have this business need.) I hope that through the ISO review process we will have enough corrections to make an important but flawed draft into a workable and acceptable standard.


A standard can address sectoral requirement.


ISO and Standards Australia's standing will be maintained by them following their procedures. It is the standing of childish sour-grapes campaigners who impugn or threaten to impugn people and bodies whenever those bodies don't deliver the results they demand that will be diminished.

hAl
2007-08-15 07:57:34
The grokdoc list of comments is a joke.
It are unverified issues collected when the site held a rally for so called contradictions. Allthough it has correctly found a lot of editorial errors in the spec (which I hope Ecma suggests to be corrected in the ISO ballot resolution meeting) it is also full of errors in interepretation or legal issues like patent law and licensing and totally ridiculous on interpretation of ISO rules and directives. Also it is complelty blind to the fact that a format allthough preferbly faultfree might have some elements in it that are known issues and still make a valid standard.


I have no doubt that the anonymous FFII member (they do like to link spam their 'hoorah for ooxml'' site) is also completly deaf for any argument you could provide him Rick as the FFII president on wikipedia already said before that this debate is "a fight against Microsoft".


I do find it amazing that a group of what should bei ntelleigent people just keeps repeating the same crap they are fed by this anti-ooxml campaign. I am 100% sure that besides Microsoft not a single organisation even remotly considers implementing rendering for autospacelikeword95 even it its rendering issue was fully documented however it is brought up by every ODF supporter of Microsoft hater at every oppertunity without them realising how totally insignificant these items in the spec really are for real use by people that are actually going to implement this standard.
Also peope worship ODF because it is supposedly less complex but ODF has not seen a full implementation of it's specification stil after more then two years of being an OASIS standard (and OOo were working on implementing ODF long before that even) and when ODF v1.2 will add spreadsheet formula's the spec will be even a whole lot more difficult to implement fully !!!