Microsoft XML Fast-Tracked Despite Complaints

Lars Skovlund writes "Groklaw reports that the Microsoft Office XML standard is being put on the fast track in ISO despite the detailed complaints from national standards bodies. The move seems to be the decision of one person, Lisa Rachjel, secretariat of the ISO Joint Technical Committee, according to a comment made by her."

There are lots of bad standards.

There are all sorts of ISO standards that people refuse to use in their current form. Not seeing this one as that big of a deal however. I'd rather have a published standard for microsoft interoperation via XML file formats then the old .doc & .xsl files.


Oh yes, "Groklaw SMASH!"

Re:All the more reason to not push new ones.

And I'd rather have Microsoft use the already ISO-standardised and widely used ODF standard.

You won't get what you want from MS Office XML

I'd rather have a published standard for microsoft interoperation via XML file formats then the old .doc & .xsl files.

This too seems to be the M$ party line - the magic of XML is better than their old secret formats. It's bogus, of course, because their new XML is as poorly defined as any of their formats [slashdot.org]. If M$ was interested in interoperability, they would use ODF and make a converter using their knowledge of their crusty old standards. It's an impossible task because their old "standards" were contradictory to begin with [slashdot.org]. At the end of the day, the old formats are doomed to well deserved neglect, and there's no reason M$ could not just publish everything about them and let their former users translate things for themselves.

There's so much double talk around this issue, it's not even funny.

The new references the old and is just as bad.

What do you mean "as poorly defined"? With the binary formats there was basically no documentation: now we have detailed vendor-supplied documentation of virtually the entire XML format.

As you will note if you follow the previously supplied link [slashdot.org], MSOfficeXML references the results of their old binary cruft without further definitions, which is no better than nothing at all.

If they really cared, they would reveal what they already know and quit keeping those old secrets. They don't and all their efforts are just so much PR, aka a big lie. You were lied to before and you are being lied to again.

M$ will tell you soon.

You have pointed out that there are a few, legacy, parts of the specification that aren't defined. What we have for XML is several thousand pages of detailed specifications, compared to close to nothing before. How is that not better?

Soon enough M$ reps will be FUDing it up with the same old noise they've always made about "partial" implementations. All day long, you can hear them say that Open Office is not up to snuff because it does not "properly" translate all of those crusty old formats. Their new XML will be much the same, so it's no better.

If they get an ISO stamp, it will be worse because they can claim some kind of reputability and "openness" that they don't deserve.

Re:"Cruft", cute

Their "old binary cruft" preserves backwards compatibility. Are you against that for some reason?

No, I think he is against the failure to document the expected behavior instead of merely mandating mimicing of legacy applications behavior without specification of what that behavior is.

One would facilitate implementation. One is a barrier to implementation. Microsoft, unsurprisingly, chose the latter, either through incompetence or desire to produce a standard that could not practically be implemented by third parties.

Published Standard != Transparent or Open

It's kind of like .doc only with obfuscation and litigation clearly called out.

What you fail to realize is the published standard in this case is handcuffed to an arsenal of undocumented licensed components.

From http://www.microsoft.com/interop/osp/default.mspx [microsoft.com]

Q: Why doesn't the OSP apply to things that are merely referenced in the specification?

A: It is a common practice that technology licenses focus on the specifics of what is detailed in the specification(s) and exclude what are frequently called "enabling technologies."

Hmmm... So the specification alludes to closed and undocumented "enabling technologies" without specifying them OR licensing them. Same old Microsoft.

No teeth.

"despite the detailed complaints from national standards bodies."

So what is the point of these national standards bodies? Standards without a method of enforcement, are called "suggestions".

The great thing about standards...

...is there are so many to choose from. Yes?

no big deal

This is likely just a fast track off a short pier.

Fast tracking it could be a good thing

Assuming that the groups that had all the problems with it are not swayed by something between now and then, the end result looks a bit more like it would be a rejection than an approval... and if it's an approval, it will be a squeaker, not a landslide victory.

That said, it should be noted that the MSOXML does not fully expand out the data. When you read the article, you find that there are still things that are binary-encoded and proprietary.

As for standards, especially ISO ones, using the words of one of my graduate class professors when he was referring to stuff from OSI: "They're camels. A camel is a horse designed by committee."

How it works

Guess we all know what she got for Christmas.

Re:How it works

A blood-covered chair nailed to her front door?

Limited number of choices here

Well, I suppose a standard could be created based on the documentation from Microsoft. It is hardly an independently-implementable standard, however.

Alternatively, a workable standard that is truely interoperable could be accepted that is not anything Microsoft would implement.

I seriously doubt there is much middle ground between these two positions. Microsoft is after all in a position to just say no.

The real problem is that even with (X)HTML/CSS it is not currently possible to take two different implementations and produce the same printed output from the same source material. This is a far, far simplier standard than anything being discussed as a word processing format, and yet there is no common implementation. I am not even sure there is today an accepted "correct" implementation for printing HTML.

How are we going to have a multi-implementation standard for word processing that produces identical formatted documents? I would say it is clear we are not going to have this. This makes the "standards" process a joke.

If you somehow believe that the "presentation" can be separated from the "content" in important documents, you probably need to have more familiarity with government processes.

This is to get past the pending laws

The only reason that Microsoft wants this to be a standard is to get past the proposed laws that specify that government documents use an open standard. That's why these proposed laws, like the one recently introduced in California, need to specify that the standard must have an open-source reference implementation.

Re:This is to get past the pending laws

"need to specify that the standard must have an open-source reference implementation"

and that's where Novell comes in, adding MS's stuff to OpenOffice.org.

Re:This is to get past the pending laws

Much like the POSIX compatibility layer originally in WinNT. Effectively worthless as an implementation, it did allow them to get contracts that required POSIX compliance.

This 'could' be a good thing!

I know no slashdoter wanted this (too much anti-ms in the air), but think of the bright side.

MS has the market by the balls with the only real competition being the WordPerfect suite...Personally I do not like it, but it is fairly widely used in School in Canada. Anything that allows Word documents to be a bit easier to convert to other formats is a good thing.

so...

how much did Lisa get paid for her efforts? Was it cash or "perks"?

Yeah, mod me flamebait. I'd prefer having that checked anyways, even if just to be sure there was no foul play. With MS, the safe assumption is that someone involved didn't play by the rules.

Have you read the ECMA responses?

Having read TFA and the PDF of the ECMA responses [computerworld.com] to the complaints, i can see why they decided to fast-track it, many of the complaints by countries are thoroughly debunked as misunderstandings of the specification. The rest are supposed to be resolved during the 5 month process.

As for TFA, they started out talking about fast-tracking the standard, then went on about totally unrelated and unsubstantiated stories about intimidation.

I may be flamed for it, but i call FUD on the part of Groklaw for this "story", the process is working as intended.

Re:Have you read the ECMA responses?

Having read TFA and the PDF of the ECMA responses to the complaints, i can see why they decided to fast-track it, many of the complaints by countries are thoroughly debunked as misunderstandings of the specification.

That's fine, but it only takes one complaint ('contradiction' in ECMA parlance) to stop the process, and there was one such provided by three separate national bodies. It stated the objection, raised elsewhere in this thread, that elements in the standard such as autoSpaceLikeWord95, which basically state, 'do things like we did in this version of this application', are contradictory to the the very essence of a document standard.

ECMA's response is not at all satisfactory. First, they provide the self-serving argument that they're reproducing the state of the art, then they say that they can throw in any missing details later in the process, then they conclude with a statement that is patently absurd:

As already discussed, the OpenXML committee chose to take a different route in defining document settings. If, however, it is decided that more documentation should be provided on the elements in question, or if the elements should be removed from the standard, that is a more appropriate matter for the 5-month ballot, and is not, in fact, a contradiction.

We can sum this up as 'We accept that nobody has ever done this before, but we don't think that contradicts other standards. Anyway, even if it does, let's just agree to talk about this later.' Ultimately, ECMA is saying, 'Whatever faults may exist, even if they're unprecedented, let's just get on with it. We'll figure things out as we go.' That is hardly what one would expect of any self-respecting standards body.

oh, no!

oh, no! ... well, let's all put our hands together and pray.

Gotta agree with the Opera guy

I remember awhile ago an employee at Opera pointed out that using html and css would create a much easier to adopt document standard. Since it is well understood and universally used. There are a half dozen html renderer's that could all be used to read content on all platforms.

This has many advantages over everything that is being offered now. A universally viewable open well understood and easily learned document standard? That makes too much sense to go anywhere.

Standards Precedence

Microsoft have already worked with standards...

Office 97 saved Word 95/6.0 documents as RTF - and that is as close to a standard as Microsoft will ever get...

Remember MS POSIX Compliance?

MS Windows is POSIX compliant as well, though in practise, you need Cygwin for any POSIX programs. Just like the POSIX trick, the standardization is just to get a tick mark on government RFPs.

Which one do we use now?

So which standard do we use, wait... How can two standards that cover the same thing, while being incompatible, be standards?

Lisa Rachjel needs to be fired... I wonder how much bribery money was involved to get this bitch to do it.

Fast Track process.

Most people do not understand that the fast track process is only a way of accelerating the process, it is not a mechanism of rubber stamping the standard, there are still six months of work ahead before the actual voting can begin.

Details about how this work can be found in Brian Jones' blog:

http://blogs.msdn.com/brian_jones/archive/2007/01/ 29/explanation-of-the-iso-fast-track-process.aspx [msdn.com]

Miguel.

This could actually bite MS in the a$$

Think about this... they release the nonstandard standard and push the software out to their large corporate vendors... then one of their vendors tells one of it's developers to design a tool that can parse the 50,000 documents and find all instances of the word "porn" for example.

The poor programmer causes a stink and shows his boss the set of encyclopedias that lays out the "open" format and tells him that he cannot do it because there are important peices missing.

The large corporate customer demands a refund and damages for the cost of migrating to MS Office... and switches yet again to an ODF standard. There's another major migration, this time telling the world that MS's "open" format is damaged.

MS will not die overnight... it's gonna take a constant beating before it ever adapts to, or folds under, the demands of the people to keep their information accessible.

I say, let MS keep manipulating the market. People are not stupid, I'd be willing to bet that a larger part of the people making IT related decisions are anxiously awaiting the time they can kick MS out their door... and as alternatives continue to grow in power and reputation, and MS continues to tarnish their reputation by misleading consumers, eventually they will be replaced with more consumer friendly alternatives.

A few (10 or so) years ago, I'd never touched a Mac or Linux box... now I have a couple of each, and every job I have had in the last 10 years came with the rollout of Linux server and a bunch of open source software. Even non technical people I meet in my consulting work are asking about Linux servers and Mac desktops because the heard from someone that had a good experience moving from windows.

MS is not dead, nor are they close... but they are closer than they were a few years ago. It takes a LONG time for power to shift when so much of it is wielded by one entity... but it does shift eventually.

The decision of one person

The move seems to be the decision of one person, Lisa Rachjel, secretariat of the ISO Joint Technical Committee, according to a comment made by her.

The implication being that committees make better decisions than individuals? Please, be serious!

Nobody seems to get the point

The main reason you should refuse to adopt Open XML is because it is not 'OPEN' in the true sense of the word. If you read Groklaw's points on it, you'll notice that the current format is closed but there is a 'covenant not to sue', the problem is that as soon as you (the programmer or your company, or OOo) extends the format or Microsoft comes along and extends or changes the format slightly for Office 2008, all those license and covenants go down the drain and the format becomes closed again, but then it will be a closed ISO standard.

Re:hmm

If people didnt jump on whatever the newest Microsoft software is they wouldnt get away with this sort of thing.

What? You mean that there should be some drawn-out process to keep the most-commonly-used XML format from being standardized?

MS's XML should be marked and tagged as standard ASAP -- that way, when Office 2010 rolls around, OpenOffice 3.0 can simply say "we put out docs according to MS's standard. If it doesn't work, it's THEIR fault."

Re:hmm

MS's XML should be marked and tagged as standard ASAP -- that way, when Office 2010 rolls around, OpenOffice 3.0 can simply say "we put out docs according to MS's standard. If it doesn't work, it's THEIR fault."

The problem with Microsoft's "standard" is that in many places it says things like "do what Word 5.0.3 does in when in double-line-spaced mode" without saying just what that means. The specification for Microsoft's XML format is not in the standards documents, it exists in only one place - the source code for Microsoft Word. Making a fully compliant implementation of Microsoft's XML format when you haven't got access to the Word codebase is therefore virtually impossible.

Re:hmm

That is true. It is however less of a problem for a program merely wishing to *write* a document that MS-Word will (well, let's be realistic -- SHOULD) interpret correctly.

True, there is a tag for "Do Line-spacing the way Word version x.y.z used to do it on a Mac" (with no further specification what exactly that was), but if you're just *writing* the files there's a simple solution to that: don't use that tag at all. (it exists only for backwards compatibility anyway, I very much doubt that it's possible to make a new version of Word write that tag if you're starting from a clean new document)

If you need to *read* the stuff though, you're out of luck, because you can bet someone is gonna complain if you're able to correctly read only 99% of all Ms-office documents, despite the documents themselves being the insane ones.

Re:hmm

Isn't that just for use when converting documents from Word 5.0.3 format?

No, it's for use when not bothering to convert documents from Word 5.0.3 properly. If you were really converting a document, you'd implement the behaviour of Word 5.0.3 using the new tags. If Word 5.0.3 in double-line-spacing mode did 1.97x line spacing and added a 0.05 inch extra margin at the bottom of the page, you should code that, not just have flag which says "be like Word 5.0.3". The place for details of legacy file formats like that is in a conversion tool, not the specification.

Re:hmm

The problem with Microsoft's "standard" is that in many places it says things like "do what Word 5.0.3 does in when in double-line-spaced mode" without saying just what that means. The specification for Microsoft's XML format is not in the standards documents, it exists in only one place - the source code for Microsoft Word. Making a fully compliant implementation of Microsoft's XML format when you haven't got access to the Word codebase is therefore virtually impossible.

I found the answer from the reply from ECMA to ISO (here: http://www.computerworld.com/pdfs/Ecma.pdf [computerworld.com]) very enlightening.

As it turns out OpenOffice has a similar feature the "config:config-item" XML property, and there are a number of these config properties that remain unspecified (from page 14):

The ODF committee chose to exclude the list of settings (many of which are commonly used in a variety of applications) from the ODF standard, which has resulted in a large number of separately defined application specific settings which is an actual barrier to interoperability. For example, the following are a small selection of properties that OpenOffice saves into ODF using application specific settings (all of which affect the display of the document):

• 
  
• ChartAutoUpdate - specifies if charts in text documents are updated automatically.
  
• AddParaTableSpacing - specifies if spacing between paragraphs and tables is to be added.
  
• AddParaTableSpacingAtStart - specifies if top paragraph spacing is applied to paragraphs 1 on the first
  page of text documents.
  
• AlignTabStopPosition - specifies the alignment of tab stops in text documents.
  
• SaveGlobalDocumentLinks - specifies if the contents of links in the global document are saved or not.
  
• IsLabelDocument - specifies if the document has been created as a label document.
  
• UseFormerLineSpacing - specifies if the former (till OpenOffice.org 1.1) or the new line spacing
  formatting is applied.
  
• AddParaSpacingToTableCells - specifies if paragraph and table spacing is added at the bottom of table cells
  
• UseFormerObjectPositioning - specifies if the former (till OpenOffice.org 1.1) or the new object positioning is applied.
  
• ConsiderTextWrapOnObjPos - specifies if the text wrap of floating screen objects are considered in a specified way in the positioning algorithm.
  

It seems that more effort has gone into finding faults into OOXML while the same faults exist in ODF.

Miguel.

Re:hmm

I think you miss the point.

{If|When} "Open XML" gets set as a standard, Microsoft will claim that Office is "standards-based and open". Which, by definition, it would be.

Open Office et. al will implement ODF. It will also implement a partial version of Open XML - as best as it possibly can do, given the vague nature of some of the Open XML implementation points.

Microsoft Office will only implement Open XML.

Now, which format is a consumer to choose? Obviously Open XML. Put simply, we'll be no closer to a real-world, workable word document standard than we are now.

Open Office will say "we tried to implement the standard as best as we could". Normal consumers will hear essentially "Open Office wont open my documents properly".

Re:hmm

That may be true, but there was strong opposition to the OOXML standard by many countries. In fact, when the standard went up for comment the ISO received more participation from commenting countries then it has seen with previous standards. Although Microsoft has actually created a competitor to ODF (which from an archival viewpoint is a good thing, the more ways we have of doing the same thing means a bigger safety net for our data) it's done so in a pretty ass backwards sort of way. This move to fast track it makes me uneasy, as there are some huge glaring holes in the standard which need to be fixed before it can really be declared as a good standard.

Fifteen years late

The industry is fifteen years down the wrong path. We (many of us) tried to warn our nontechnical peers before things came to this point. We tried to express the benefits of a diverse field. We tried to illustrate the merits of alternative technologies. We tried to sing the praises of other operating systems and other companies. The sad fact is that computer technology was wrestled away from the true technologists who invented it and was thrust headlong to the public sector by the businessmen, politicians, stock brokers, and bankers who saw a massive profit potential in it but had no real knowledge or appreciation of the intellectual advancements which created it.

Billions of dollars in taxpayer money were funnelled, through government grants, contracts, and subsidies, into social circles and corporations who had demonstrated a willingness to put aside the morals and values of the true scientists in favor of ensuring their own priveleged paychecks, pensions, and long term profit margins. The American taxpayers subsidized the startup of the .com bubble, we paid for the infrastructure on which the rest of the internet was built, and we paid for the products, the software, and the services on the consumer end. Where, then, did the profits from the .com bubble go? The profits went into the hands of the same major investment groups who have been carefully profiling and controlling the market for generations--people who, when the .com bubble became the .com bust, shrewdly bought the real estate being sold by the common people seeking to ameliorate their losses (which had been carefully planned by those people who were now buying their real estate at dirt cheap prices). When America began to return to consciousness after the .com blackout we now find that the same real estate which we sold to keep ourselves from bankruptcy is being rented or sold back to us--as condos, apartments, are housing communities--at three, four, ten, even hundreds of times the cost.

The pyramid [slashdot.org] scheme [slashdot.org] is so beautiful we could almost cry for joy if we were on the financial winning side of it. As it is we have no choice but to cope with a world where Motorola is relegated to handhelds, HP has partnered with Compaq and become just another x86 retailer, and Microsoft holds a betting majority of the chips when it comes to influencing the direction of software development and globally recognized protocols.

Re:Opera is no better...

What makes you say opera was 'fast tracked' by that email (other than the author's email ended with @opera.com, and it was CCed to w3.org people/lists)?

Re:hmm

All 6 of the complaints were about technical issues. The 1-month fast track approval is not the correct place to raise those types of issues. The only thing that can keep something from getting fast track approval is an objection that highlights why it conflicts with standard that has already been approved. None of the 6 complaints did this, so it was pushed through.

They can of course, raise the same complaints during the 5 month ballot process, which is the correct time to raise such concerns. Although, 6 out of 100+ is still a fairly small number.

Re:hmm

Funny, I thought this standard conflicted with the ISO standard for time because it incorrectly treats 1900 as a leap year in spreadsheets.

Re:Opera is no better...

Opera just fast-tracked their "HTML5" proposal with W3C: http://lists.w3.org/Archives/Public/www-forms/2007 Mar/0019.html [w3.org]

HTML5 doesn't say things like "render like Opera 7 does"