Whilst the letter itself might be unusual the contents are not.

Is the standard more open as a result of going through an ISO process? Yes.

Has it been improved because folks bothered to take it seriously, pass comment *and* vote "no" in September? Yes, it would be hard to be worse than it was (although in some areas the responses have succeeded in that seemingly difficult goal).

The key question now is whether it is or will be in a fit state technically or otherwise for an ISO standard? No-one will really knows until after the BRM and we finally get to see the re-worked document. The letter offers no opinion on that. It simply addresses a single issue of "openness" which is a bit of a straw man anyway.

Rest assured, however, NBs will be following your advice and thinking long and hard about this letter.

Don: Yes, ultimately the thing that matters is the concrete editor's instructions that come out of the BRM.

You would expect them to be even more acceptable than the Editor's Disposition of Comments proposals, and given that the Dispositions are already acceptable for some (see, for example, http://xmlguru.cz/2008/01/ecma-response-to-czech-ooxml-comments) I think saying "No-one will know" is a bit strong, because I people can have a good informed guess even now, given that Ecma has agreed (in principle) on most NB's comments, including many of the showstoppers AFAICS.

Patrick's comments really focus the issue on maintenance, which in the long run is just as important as the BRM. ODF and OOXML are formats in evolution, not static things like the meter (thoguh even the meter has been updated!)

I tend towards the idea that Ecma should be the body handling corrigenda (small problems and errors where you want fast review and action) and Ecma-driven ammendments while the slower ISO process is better for NB-driven ammendments, convergence, mappings, round-tripping, harmonization, etc.

To discuss the technical issues in OOXML is a moot point, because we should have never gotten here anyway.

People have to realise that OOXML was never needed anyway. ODF is already here, it's XML, it's an ISO standard, it has multiple implementations, both open sourced and commercial. It was developed cooperatively by multiple vendors and independent developers.

Although it was based on Open Office's file format, it evolved from the collaboration of what others had to add. As much that even Open Office had to write code to become really compatible to ODF 1.0. It's a standard that is consistent between Office products, a table uses the same tags on the word processor and on the spreadsheet, which makes sense.

ODF is a standard that builds on other pre-existing standards, such as SVG, MathML and ISO 8601. It's a standard that was approved in ISO using the normal time for standardisation, instead of trying to rush it through the standard bodies.

But then, Microsoft realised that the standardisation of ODF, associated with the fact that companies won't implement Office 2007, and then decided to turn on the big FUD marketing machine to get out of that situation.

They document (if that can be called documenting it) their format full of holes and then buy Ecma to submit it to ISO. Then they start with the FUD. They say that OOXML is needed because it permits compatibility with Microsoft's brain dead binary formats.

Of course!!! How would I live with a format that doesn't allow me to use a calendar with days missing and extra days? Or how could I go on if on my documents from 10 years ago the footnotes appear on the page they're supposed to instead of the wrong page? And what about that crappy autospacing of that version from 13 years ago? If I open the documents today and that crappy spacing was different, all my work would be lost!

All I can say is: COME ON!!!

And then Ecma (puppet of Microsoft) starts saying that the standard cannot be changed or fixed, otherwise it will lose its huge feature of being compatible with... well, BUGS!!! A standard compatible with bugs, who needs that?

But I diverge.

The fact is that Microsoft started this process with one or more of these objectives:
- Have an ISO stamp on their format (though what ISO standardises will certainly be different of what they implement, see what they've done with HTML on IE for some historical perspective), so that they can still sell their products to governments that will require standards;
- Create FUD to try to avoid people from adopting the standard that already exists, ODF;
- Make people lose time discussing this issue (what a waste of time and money!), when they could be improving and implementing ODF;
- Make some PR telling the world they are so good for "donating" this standard to the world (what a lot of BS!).
And here, Rick, YOU are losing a lot of credibility, by defending this format, saying it has merits, and even trying to discuss the technical issues.

I still don't understand why, specially considering you have a blog at O'Reilly Net, which usually is a very respectable site. I also saw your previous work on XML and I can say it's good. So I don't see why did you start defending OOXML. I just can't believe that you can't see what Microsoft's underlying intentions are, that they just want to disrupt the market and this whole process is, well, FUD!

Anyway, in this case, I think I should say that you are not part of the solution and, so far, you have been mostly part of the PROBLEM! If your conscience lets you sleep this way, ok... but please stop to think about what you are saying. It's not too late for you to change your mind.

Thanks,
Phil

Phil: Thank you for your concern on my sleep.

Your argument seems to be that it is not a real use case to be able to open up old documents with good fidelity. In some cases that is true, but not always.

By the way, your claim that OOXML is a conspiracy to make people lose time instead of implementing ODF is a great new thought. If ODF implementations are lacking in some way, it is really MS' fault! Marvelous!

@Phil:
false

In your ODF standard, where does this element fit in?
It was found in an OOo-specific settings.xml file, but looks like it would be of interest to any other program implementing ODF. This is from the latest version of OOo I could get my hands on. I haven't yet bothered to try creating a really complex ODT document to see if ooo and ooow namespace elements will start creeping into the main content, but on the face of it this seems far less interoperable than OpenXML.

looks like my xml got chomped... (should have previewed):

Here we go:
<config:config-item config:name="DoNotJustifyLinesWithManualBreak" config:type="boolean">false</config:config-item>

Nksingh: I agree that ODF interoperability may be lacking some features, but introducing another standard is not the answer for it. If you introduce another standard, you'll end up requiring all Office suites to interoperate with *both* of them, do you think they'll interoperate more easily?

Also, you mention that ODF is "less interoperable than OpenXML", but right now there is *nothing* that interoperates with OOXML, since the only product that implements OOXML is MS Office 2007, and even then, it doesn't implement it exactly, doesn't implement the version ISO has, and hasn't promised to implement the version ISO approves (if ISO approves it). Even if there were multiple implementations of OOXML, none other than Microsoft could implement autoSpaceLikeWord95, because only Microsoft knows how that buggy version did that, it's not on the standard. The others would have to guess, and probably wouldn't be able to keep the so much desired fidelity.

And then, what's the point on keeping fidelity to bugs?

Rick: My argument is not downplaying the importance of fidelity, it's that what Microsoft calls "fidelity" is actually "reimplementing bugs". The same thing they did with IE, they put buggy implementations out and then they have to stick with the bugs, because they don't want to break existing websites, and then every web developer in the world has to work around IE's bugs. Everybody is sick and tired of that.

And if you can't see that Microsoft's effort is nothing but FUD, that lots of people are fighting this bastardization of standardization efforts instead of improving the existing standard, well, you must be blind! Or else the same reason that makes you sleep so well is making you deny this too.

And it's good that Microsoft puts out in the open the specification for their formats, both binary or XML, but there's no need to standardize them. There *IS* already a standard for that. Bringing another standard only makes it harder on every developer out of Microsoft that will have to implement two formats, spending double the time. If Microsoft implemented ODF as their native format, this whole process could had been avoided. But of course they won't do that, because interoperability is *NOT* what they want, what they want is to create FUD, to make people waste time in this pointless discussion, to discredit ISO and standards, and to continue with their lock-in.

And if you can sleep, well, that only can mean one thing to me, and it's so obvious that I don't even have to say it out loud. But I'll say it anyway: you're a Microsoft shill, either they're paying you, or hiring you, or they invited you to go to Redmond, or something like it. I just cannot believe that you are unbiased, you certainly aren't. If you claim that you are unbiased, please prove it.

Hi Phil,

Thanks for making me laugh on this gray day in Denmark.

You are hilarious!

:o)

Phil wrote:

"If you claim that you are unbiased, please prove it."

Good point, Phil. I've strongly suspected Rick of being a Martian for some time now - since you're getting him to prove stuff anyway, would you mind taking this one up while you're at it? Thanks in advance.

By the way, the letter refers to "Open XML". Is this a new standard or has the official name changed?

I believe that Phil and Don are the only ones who are saying anything sensible in this thread. Rick, based on your past reputation, I'm surprised and saddened by your decision to buy into Microsoft's propaganda.

Oh, and yes, it's not "OpenXML" as all the Microsofties seem bent on calling it ("If you tell a lie big enough and keep repeating it, people will eventually come to believe it" -- Goebbels). Rick, please don't contribute to Microsoft's clear attempts to pollute the name space. It's "Microsoft's Office Open XML" or MSOOXML - clearly, in itself, an attempt at subvert OpenOffice.org which is perhaps the leading implementation of ODF.

The discussion of MSOOXML's technical viability or its suitability for ISO standardisation shouldn't even be taking place - that's not the issue (despite the fact that it's not technically viable anyway). ODF is the standard in the space. It's simply, it acknowledges the place of ancilliary standards, and it's got multiple implementations. ODF has already won the race. Microsoft didn't even turn up at the starting line (despite being invited).

All Microsoft is doing with its MSOOXML gambit is show the rest of us how desperate they are to avoid having to compete on a level playing field.

It would be a shame if Microsoft's attempts (using every external influence it can must - financial, political, PR, etc.) to redefine the game so that can win, too, meant that ISO lost the respect of the NBs and the governments they represent... I understand that's a very real danger from some of the people involved now.

Regards,

Dave

Apologies - a few typos in my previous post:
"It's simply," should be "It's more straightforward,"
"it can must" should be "it can muster"

Oh, and to clarify - my "namespace pollution" accusation isn't for the use of "OpenXML" - I meant to say that calling the format "MS Office Open XML" is a rather clever but most likely disingenuous attempt to cloud the OpenOffice.org waters for those only peripherally aware of the issues involved (e.g. most users, government officials, and CIOs). The "OpenXML" religion that the Microsofties have adopted is an attempt to further overload the term "open" to have Microsoft's connotations... i.e. not very open...

Hope that clarifies things.

Phil: You say Bringing another standard only makes it harder on every developer out of Microsoft that will have to implement two formats, spending double the time but why wouldn't people just implement one format and use a converter, if they are only interested in the common subset or more interested in one format than the other?

As long as converters are available, I don't see that OOXML adds to the complexity of at all, given that there are so many formats already out there that need supporting. In fact, the reverse: by providing a standard or common pivot, it can allow multiple existing legacy import systems to be consolidated.

I don't think it is as simple or one-sided as you are presenting.

Don: It is commonly used. I tend to use OOXML because it is too easy to make mistakes with Office Open XML: people read or may write Open Office XML. But OpenXML does not have that problem either, so it is OK.

One of the comments I submitted to Standards Australia was AU-4, that the standard should have a different name than just its current name. I think clarifying all these "scope" things would settle some of the concern that people have that ISO OOXML would be adopted where ISO ODF was more appropriate. (However, I admit that this issue is hardly unique to OOXML/ODF, and that with all voluntary standards the user needs to be informed and responsible about whether to use the standard.)

One names, I have a pet issue of my own: I think it is important for people to try to use "DIS29500" when they mean the text of the draft rather than OOXML or OpenXML or Office Open XML, etc. That opens the door for a lot of wooly and bad logic, where comments about a particular draft get attributed to the possibilities of the finalized technology.

@Phil:
Apple also supports OOXML on several of their devices. You almost certainly can't argue that they have a product based on Microsoft's codebase. You can look at docx files on an iPhone today, so I think you should give up your wishful fantasies about the world abandoning OOXML. Even IBM will come around too in the end, when they realize that continuing this battle will net them less profit than simply making tools that work with OOXML.

Your argument that there IS a standard does not hold much ground either. As has been stated ad nauseam, there can be more than one voluntary standard for the same thing (and ISO standards are voluntary). And standards can be superseded when the technology changes.

Hi, Rick!

Thanks to slashdot I found where your money's coming from:
http://slashdot.org/comments.pl?sid=451660&cid=22401102

You're an Ecma employee. Ok, so as Microsoft bought Ecma and that's the money that's paying your salary, you're excused to say whatever they tell you to and make an ass out of yourself. After all, that's your job.

Having said that. You said: "why wouldn't people just implement one format and use a converter, if they are only interested in the common subset or more interested in one format than the other?"

Well, I return the question: Why wouldn't Microsoft just implement the existing standard format, if they're only interested in interoperability?

Well, the only possible answer for that is that they don't want interoperability! Well, it's their right to not want it, but it's wrong to subvert standard bodies to get their beloved profit.

The excuse that ODF doesn't do all that Office needs doesn't convince me at all, because if that's the case, Microsoft could join the ODF TC to discuss what how they would like ODF to be. Microsoft was even invited to that, but of course they didn't have interest. They only showed interest now, because they see it as a threat to their lock-in business model.

You say: "I don't think it is as simple or one-sided as you are presenting."

Well, not as one-sided as you are presenting. You focus on the technical aspects, but the bigger issue here is a political one. Would you care to comment on it? Or, well, they only pay you to talk about the technical ones?

@phil

It is in Rick's interest if both ODF and OOXML exist.
That way there is much more work for Schematron to be
disentangling.
And Rick can always consult on that basis.

@Phil

I'm just curious, who are you Phil? Your comments are everywhere on the web regarding ODF and OpenXML, most of which are half-truths with a good chunk of them completely incorrect.

Your criticisms for ECMA especially are laughable considering it is the standards body that defined the Javascript/ECMAscript, the CD-Rom as well as many popular data compression schemes.

The sad reality is people don't want to be burdened with actually reading the ISO comments or objectively assessing the opinions the technical committee closest to the standardization processs. It is much easier to read Rob Weir's blog and a bunch of slashdot comments to form an "educated" opinion that to understand the meat of the argument. It is a disservice to the software engineering profession to relegate both ODF and OpenXML specifications, both great achievements in their own right, to mere hearsay and shameless punditry.

10 years from now, people are going to look back at this debate and say, what a waste of braincells and perfectly good HTML.

That link to slashdot is only a rumour. My understanding was that Standards Australia told the Cyberlaw symposium held in Sydney December last year that their delegation would consist of 2 SA employees. The head of delegation was to be Alistair Taggart, not sure of the other name but Rick's was not one of them. (Actually, I think the slashdot post is incorrect in another way as Rick has been perfectly open about MS paying him to help them with the OOXML standard.)

Give the EC investigation into the alleged OOXML process irregularities I would be surprised if SA changed this position.

Is this something you can clear up, Rick?

Icebugg: What a strange comment. To my knowledge no-one is using Schematron for ODF and Open-XML at the current time: it certainly could be used though. The question is, given that there will be OOXML files to be disentangled no matter what happens at the ISO vote, would it be better to have to be disentangling files according to the existing OOXML or to be working on files with all these improvements, both now and through an ISO maintenance process? The view that if OOXML is not standardized it will disappear is such rubbish, at least for people in my industry (industrial publishing and markup.)

Phil: No I am not an Ecma employee. On your question of why MS doesn't implement ODF, well for a start they *are* doing that(in their open source plugins, for which I hope there will be enough government pressure to get put into the main release when it comes out again.) You have to realize that MS' has a long development cycle: they had to decide in about 2004 the basic parameters of Office 2007: even though they are rich, they don't have a time machine that lets them implement a spec before it is written. So the game, for people who are seriously interested in standards, is first how to get everyone to the table with their ya-yas exposed. Then we can see what the gaps are, and work out strategies to fix it. And it involves making sure that people don't have illusions about where ODF, OOXML, PDF etc are appropriate.

I have never swallowed the idea that there could be a quick fix in this. Anything that happens will happen slowly, with continued pressure on the parties, and in forums that are fair: and fairness means that things will be judged on their technical merits which sometimes MS will prove their point and and sometimes they will have to adjust. ISO is good at that.

The approach of trying to handle Microsoft's market domination by blind opposition has comprehensively failed. Largely due to government pressure, we have a historic chance to steer them towards being acceptable corporate citizens. People whose livelihoods and world view are based on reaction and opposition being the only strategies really don't get the idea of constructive engagement and containment. But given that we don't live in an age where governments are willing to split MS up, and given that we are seeing more new market dominators arising in their own areas, not less, one practical thing we can do is make sure that all market dominating technologies (and for now ODF is a smokescreen, though not forever) should be RAND-z standards.

I hope people picked up what Patrick Durusau was saying in his letter: as well as saying "Look, Blind Freddy can see there is a measure of openness going on here", he is also pointing out to MS that they are clearly gaining by this little adventure openness. They are getting a much better standard, and they don't need to fear true openness, which is not the same thing as participating in a majority vote where their competitors outnumber them! My boyfriend often says "Honey is sweeter than vinegar" (he insists this is traditional and exclusive Asian wisdom).

Don: The other was John Snair, who has been involved in many standards including at editor level and was a really good choice even if it meant our delegation would not contain subject-matter experts.

Actually Standards Australia people talked to me about it last year first, and I said I thought it would be better if I didn't go. I didn't want to expose them to the nastiness that I thought would inevitably be directed at them. Their number 1 and 2 choices dropped out (who were both very good people) the day before names had to be finalized, they told me, so they put in #3 (Panjan from Standards Australia) and had to fill #4 fast before the end of day, so they asked me again. I had to sign a form saying that my company was paying for me for the duration not any other (e.g. Microsoft, Ecma),etc and I don't think I am a particular shrinking violet at standards meetings.

As I understand it, they picked me exactly on the reasons they pick anyone to be the technical half in a delegation: seniority, experience, subject competence, availability, professionality (e.g. trying to be scrupulous in making them aware of possible conflicts of interest.) Political issues one way or another don't come into it.

I am not head of delegation so I have no vote, of course. But I expect some people will not be happy that I am going at all. That I submitted most of the Australian issues (20 or 22 out of 30), that no-one on the committee has identified any outcomes from our committee that I actually disagree, that I have a kind recommendation from the highest level of ODFdom, and that in this area I think I have a track record of being able to get a good outcome, does not seem to calm them.

Anyway, the BRM is a technical meeting to get a better text, not a political meeting. In fact I have told Standards Australia I would be perfectly happy to withdraw if they asked me, if it I was causing them unnecessary trouble, but they gave me the idea that their method of dealing with contentious issues is just to play it by the book and not take sides. Almost every important standards has people with definite views.

I am quite reconciled that no matter what I say on technical issues, extremists on one side will claim it is biased (unless I say something that is on "their" side, in which case they will quote it like the ODF Alliance.) For example, today I discovered that there are dozens of issues from NBs where they say "ODF can do this, OOXML must do it too to help compatability": for example like absolute positioning of a graphic on a page. Now, is this good news for ODF (one less barrier to ODF adoption) or insulting (how dare I point out that multiple NBs didn't get that issue right, how dare I imply that the reviews were carried out by humans not gods!) or diligent (that I am trying to do independent research not just parrot the Editor's disposition) or bad for MS (see OOXML isn't as different as you are claiming) or bad news for ODF (one less feature that ODF has that it can use to say it is better than OOXML) or what? Answer: none, it is just a technical fact that can go into the mix at the BRM.

By the way, in what way is this an irregularity?

"By the way, in what way is this an irregularity?"

I don't know. I didn't say it was.

I think you would be a reasonable choice for any BRM delegation *if* there were balancing views represented on that delegation. My concern is this, you have been an unflinching supporter for having OOXML accepted as a standard by ISO. The fact that you have worked with Microsoft on this is an indication of the strength of your views.

Many people have strong, pragmatic technical reasons for taking a different view. In reading this blog and listening to you I can see you have had scant regard for their concerns. I think SA would have been better either with a completely neutral delegation (as originally indicated) or a third member who represented those dissenting perspectives. I know there were people in Australia who would have been perfectly capable of representing those perspectives eloquently and relevantly.

I mentioned the extended scope of the EC investigation because I would have thought that NBs and ISO would make every effort to be seen to take a consensual and balanced approach. As you yourself point out above, SA will not now be seen to have taken that sensible course of action.

Don: The purpose of the BRM is to prepare a better text for NB consideration. It will largely be a boring technical meeting based on "What is the best thing we can agree on to reduce problem X or fix problem Y or improve feature Z?" It will be so large that most delegates probably won't get much of a chance to speak, and I expect a lot of the ideas work will be done by delegates interested in a particular issue brainstorming in the corridors, in midnight drafting sessions and over friendly drinks, as usual.

So frankly the kinds of issues where I and some of the Standards Australia technical committee differ either are not on the items that the BRM will deal with, or are in the kind of normal variation that always exists. I don't think this is first time that Standards Australia has had to deal with commercial/technological rivals and their shenanigans.

As I have said on multiple occasions, the ISO standards process is not geared to win/win rather than win/lose. People trying to finesse the process in a different win/lose model are going to be frustrated, and the strategy of slandering the process and disrespecting individuals is the Plan B as this appalling win/win looms larger. Patrick Durusau's open letter is directly a response to the excesses of Plan B.

The Ballot Resolution Meeting is *technical* and *editorial*. I have asked for actual examples of what issues on the table for the BRM that my general pro-OOXML stance would impact (unless something has come in in the last couple of days-I have been offline no-one has provided any) and I am genuinely interested in knowing these technical and editorial issues and working through them.

So is the idea that I won't go the BRM and sneer enough at people who want to have and use this standard, or to inform them that they don't matter and they are puppets of MS, or that they should not be trying to negotiate the best possible editor's instructions? I respect the BRM process, and I won't be trying to sabotage it, but neither will I be trying to sabotage getting good resolution of all the technical and editorial issues that came from Standards Australia: the issues where I personally differ from several other members of the Standards Australia technical advisory committee are either outside the scope of the BRM or within the normal range of differences that occur with *any* delegates.

In this particular area *only*, I think I have pretty good technical credentials (and I certainly acknowledge that there are several Australians with really good credentials, though none with exactly the same mix: Peter Sefton, Charles McCathey-Neville, SVG's Dean Jackson, to name a few.) DIS 29500 uses XML: I was in the larger group that created XML. XML is based on SGML: I was on the group that enhanced SGML to cope with XML. It uses XML Namespaces: I was involved with that. It uses XML Schemas: I was on the XML Schema Working Group. Internationalization is important: I was on the W3C I18n Working group with a particular interest in understanding, defining and supporting CJK requirements. It uses RELAX NG: I was on the SC34 WG that standardized that and did the XSD to RELAX NG conversion for DIS29500. It uses ISO NVRL: I was on the SC34 WG that standardized that. It uses MIME types: I was on the IETF list that worked through the issues of XML MIME types. It is an ISO standard: I have been the editor of a standard. It is in the area of markup and publishing: I have been working in the area of industrial markup, conversion and publishing for most of the last 20 years. It uses a particular range of document structures: I wrote a 600 page book largely on document structures. It is involved in typesetting: I have typeset and designed several books. It is involved in fonts: I have revised fonts for publications. It is involved in Unicode: I worked on the early ISO-entity-to-Unicode mappings and have a very good understanding of character encodings. It is involved in ZIP: SC34 tasked me to investigate an ISO standard for ZIP (and the only reason I expect that we may get it is because with MS at the table we have a critical mass to make a ZIP standard fruitful.) Blah blah blah. That Jon Bosak kindly invited me to join the original ODF TC at OASIS (I couldn't due to lack of time) I hope shows that some people think I am not a complete idiot all the time. That the ODF Alliance quoted some of my criticism of the Editor's Disposition of Comment I hope shows that I am not entirely a puppet.

I am not downplaying that I have a different view than representatives of MS' business/technology rivals Google, IBM and the Gnome Foundation. However, the BRM is a technical and editorial meeting whose aim is a better text. It might be tempting to rejoinder "Well, obviously a person who wants a standard will do a better job than people who want to stymie it, therefore the anti- mob are clearly unsuitable" but I think Google, IBM and the Gnome Foundation's positions are a lot more complex than that: that they all have software that to some extent accepts or produces OOXML shows that. I think even the most touchy of them will be challenged by Patrick's open letter, especially his view that a standard for OOXML is actually positively useful for ODF to achieve its objectives.

It would be an odd world if only people who were against any standard were allowed to participate in its development (I know that is not what Don is saying!) The thing is that the standards development process is an alternation of SC standards committee work by fixed numbers which then gets submitted to NB discussion and ballot; the SC committee work is concerned with getting a better text (what counts is a can-do attitude and co-operativeness), the NB ballots concerned with judging if they are good enough (what counts is good judgment and respect for the requirements of others.)

Okay. It is nice to argue that OOXML is progressively getting more open. Although I do disagree with how this apparent openness is coming to fruition. Everything is being _rushed_ through. A huge number of pages, so little time to review. If we have to live and work with OOXML as a standard for the next 10+ years then DO NOT RUSH IT! OOXML should have taken the long road to become a standard or appropriate modifications made to ODF. Then again, when all is said and done, it is not about our best interests or even about us, it is about the solidification of the Microsoft Office monopoly. This I cannot agree to under any circumstances. ODF = Open Document Format = Open Document Freedom!

"The Ballot Resolution Meeting is *technical* and *editorial*. I have asked for actual examples of what issues on the table for the BRM that my general pro-OOXML stance would impact (unless something has come in in the last couple of days-I have been offline no-one has provided any) and I am genuinely interested in knowing these technical and editorial issues and working through them."

So you mean the random, inconsistent, tag names and bitmask attribute values are not on the table?

I am not very well informed into the individual problems that remained after the September vote, but I vaguely remember that some NBs complained that tag names in OOXML were inconsistent. Furthermore, some attributes had bit-mask values.

Another question was why OOXML did not seem to confirm to XSL.

You are the specialist on this, but I have never been able to find your opinion on these matters (even after I did ask for it elsewhere). If I missed it, I would love to see the place where you worte it.

Winter

Winter: I have had various comments on bitfields on the Grokdoc pages for most of 2007 and I think I have mentioned them in this blog a few times. This is my opinion:

1) Sometimes they seem justified as a good design choice
2) Sometimes they don't seem justified as a good design choice
3) Sometimes it is a toss up

For example, in the first case are the use of bitfields for the sig element, where it is a magic number used to match with ISO Open Font fonts and not of particular interest as individual booleans.

In the third case I would put the repeating tables information. There is a reason for it. However, I think reifying the value with a name and making it a style is the better approach.

In the second category I would put most of the other cases. Explicit attributes are better. I'd expect that a decent maintenance program (or the BRM if it gets around to this) would get things improved.

However, having said that, the premise of most of the NB comments on this issue, which is that schema languages cannot handle bitmasks and that languages (in particular XSLT) cannot either, is wildly, incompetently wrong. Schema languages are my area: I guess the issue comes up because there is no datatype in XSD or XSLT for bitmasks, but the various string, bitwise and numeric operators in languages can cope fine. (XSLT was not designed as a general purpose programming language, nor the ultimate data processing language: indeed it was a stripped down version of SC34's ISO DSSSL which is the LISP (scheme) dialect that has been used to produce Linux documentation using SC34's James Clark's Jade program. When James moved over to the XML work at W3C, stripping down DSSSL and giving it XML syntax, the aim was simplicity not completeness, and it seems a little over-enthusiastic if XSLT is now the bar to judge markup.)

So it comes down to best practise and aesthetics and inconvenience, not the disaster that people say. IIRC I was susprised to see that one well-known company's "independent review" (which just consisted of Googling for a couple of points from Grokdoc AFAICS) mentioned hidden binary fields, which turned out to mean the bitfields.

On the issue of names, the preference for long or short names is aesthetics and habit not holy writ. On the issue of consistency, there are two aspects that have been raised: the first is that all attributes and elements with a certain name should have exactly the same definition: however XSD (and even DTDs) allow local definitions (i.e. context dependent ones) and RELAX NG and Schematron go crazy with allowing context-dependent typing. I think it is matter of aesthetics, whether you prefer to use generic names to match the function regardless of the type or whether you want the name to reflect the type. The second aspect of consistency is that sometimes different names are used for the same thing, and I agree that would be better fixed, but it is not the end of the world. It is not as if anyone will memorize these things.

And that leaves the naming issue that they are mnemonic and therefore confusing and culturally dependent. I have a lot of sympathy for this view. In SC34 WG1 we have been working solutions to this class of i18n problem: the Document Schema Renaming Language (part of ISO DSDL) which allows you to remap element and attribute names into your own conventions (not structures): Martin Bryan has been working on the open source software for this (not a reference implementation but certainly a proof-of-concept and working implementation) just converting DSRL to XSLT. For very distant languages (I am thinking here of Chinese in particular) it may be easier to rename the schemas, which indeed just what they did when they forked ODF into UOF. For very close languages, I don't know that DSRL is the answer.

In a sense, if the infrastructure does not support what some person wants to do in markup well, SC34 peoples' response is more likely to be "Oh, how can we improve the infrastructure" rather than "Oh, silly them".

Once you are aware of three things, then OOXML and its names make a lot more sense:

1) References have one level of indirection, through the various .rels files.

2) OOXML is generally systematic in having a property element with a Pr suffix rather than attributes (so the p element has a pPr subelement which contains properties): this came from an early naughties markup fashion where people wanted to get rid of attribtues and have "structured attributes". I was not keen on it then or now (I like attributes) but it is usually a matter of aesthetics. Knowing when an element is a property makes the various elements much more sensible: for example, you look at and you think "Oh, of course, that would be bold". Not knowing the Pr convention renders some of the names pretty cryptic.

3) WordprocessingML is not structured but linear. SpreadsheetML uses shared strings tables. If you don't know the concepts then even long names will not make sense.

you work ( not formally of course ) for Microsoft: you are helping the current Microsoft office team/ECMA/Jaeschke work to put DIS 29500 in well-enough form to be at least not a laughable fast-tracked ISO standard

so, if you participate in a NB who has to judge if DIS 29500 has the technical merits to be ISO-stamped, then i believe there is a conflict of interest here

you became judge and part.

do you know the word "ethics"?

Marc: What you are saying has never been the way that standards bodies operate. A standard normally happens because the people who want it come to the standards body and champion it through; other people who also want it join in, and frequently commercial rivals join to make sure that their needs are catered to. The standards organizations are set up to reduce the chances that rival organizations can stymie things. In some countries, they even call their standards "Industrial standards" because of this tight connection to serving the needs of industry in a community.

It goes to the idea that I have been harping on about at regular intervals, that a voluntary technical standard is an agreement (i.e. between its parties and whoever wants to use the thing) not a regulation (i.e. for which there is hopefully democratic oversight.)

I suggest you read the material at Consortiuminfo.org on the Allied Tubemakers case. You will certainly see a conflict of interest there.

The initial way to correct way to handle a potential conflict of interest is to declare it. Cards on the table. Then the standard body can figure out if there is any conflict that is different from everyone else's in scale or kind.

But mere professional participation is not a barring conflict of interest. Look at Dr Durusau: he is passionately keen on ODF, involved in INCITS V1, and Sun have sponsored him to be editor this year (full props to Sun).

You are not actually concerned with conflict of interest, otherwise you would be demanding Dr Durusau's head (and the heads of most of the editors of W3C and ISO and OASIS standards, who are all in the same boat, by and large.) You are concerned that someone with a different opinion to you has not been excluded from the conversation.

It is only a conflict of interest if it involves someone on the opposite side to you? What about someone, say, who would get a big fat bonus if their company's corporate objective of blocking OOXML was achieved. If they took the scrupulous line, they would say "I have a potential conflict of interest that may influence me to play up technical problems and cause mischief: a big fat bonus!" ;-)

rick

as usual i don't agree with you... i'm not talking here about OOXML, Open Office, OpenXML, OXML, ClosedXML, ODF, Office Open XML, MSOffice or what ever ...

my point is simple:

. let rj be some person
. let suppose that ISO national bodies represent "national" interests ( including citizens ) and not only the economic interest of corporations

i) rj is paid to help some DIS proposed by a corporation to "pass" ISO technical-merits-judgement

ii) rj participates in the delegation of a national body and then rj must cast his opinion if the DIS has the technical-merits to pass ISO judgement

then, there is a conflict there[1], you are not "free" to do your ii) work because you are against yourself

--marc

[1] http://en.wikipedia.org/wiki/Conflict_of_interest

"Winter: I have had various comments on bitfields on the Grokdoc pages for most of 2007 and I think I have mentioned them in this blog a few times. This is my opinion:"

1) Sometimes they seem justified as a good design choice
2) Sometimes they don't seem justified as a good design choice
3) Sometimes it is a toss up
4) Sometimes, the DIS proposer owns the only useful implementation of the format, called Office 2007 and it happens that this implementation was designed and coded to handle bitmasks and they don't want to burden with change the implementation code base nor have time to do so, because the beast (DIS 29500) is being fast-tracked, so there is no time to fixes like this, so, they prefer to keep things complicated, going against XML goals[1]:

"...
It shall be easy to write programs which process XML documents.
..."
[ Extensible Markup Language (XML) 1.1 (Second Edition) ]

Bet on 4) ;-)

--marc

[1] http://www.w3.org/TR/2006/REC-xml11-20060816/#sec-origin-goals

Marc: See The
Design Goals of XML for why Goal 4 does not mean what you want it to mean. It is related to parser complexity, not application complexity.

"Marc: See The
Design Goals of XML for why Goal 4 does not mean what you want it to mean. It is related to parser complexity, not application complexity.
"

mmmm... i'm talking about "processing" complexity

did you forgot why was XML invented? to be *processed*

i.e: http://www.xml.com/pub/a/2004/02/04/tr-xml.html

do you believe that ECMA examples ( pages 722-732 of [1] ) of bitmasks handling is a model of "easy" processing of an XML document?

( since it seems that you are defending ECMA/Microsoft justification, i'm starting to believe that you like to code object oriented programs using "fast-track" assembler ... i prefer "right-track" Smalltalk ;-)

--marc

[1] http://www.itn.liu.se/~stegu/OOXML/DIS29500-2008-002.pdf

Marc: So under the same logic you would of course say:

. let rw be some person
. let suppose that ISO national bodies represent "national" interests ( including citizens ) and not only the economic interest of corporations

i) rw is paid to help some DIS proposed by a corporation to "fail" ISO technical-merits-judgement

ii) rw participates in the delegation of a national body and then rw must cast his opinion if the DIS has the technical-merits to pass ISO judgement

So I presume you are having the same conversation with people with greater conflicts than me on the anti-OOXML wing?

However, if you read what I have wrote The initial way to correct way to handle a potential conflict of interest is to declare it. Cards on the table. Then the standard body can figure out if there is any conflict that is different from everyone else's in scale or kind. you will see I am not denying that there are potentials for conflicts of interest, and that these should be declared and handled appropriately.

In the Wikipedia article, it mentions that conflict of interest relates to people in a position of trust. So it does not apply to normal committee work at Standards Australia, because there is no position of trust: you represent any legitimate interest you like, openly and honestly.

So what about the BRM? The purpose of the BRM is purely technical and editorial, so there can be no conflict of interest relating to non-technical and non-editorial issues.

In particular, there is no conflict of interest from me wanting the standards process to success by an ultimate yes vote with a good draft, just as as there is no conflict of interest from people who want the standards process to succeed by an ultimate no vote if the draft is inadequate.

So what about a conflict of interest over technical and editorial issues at the BRM? First I would note that it is the delegation which has to represent Standards Australia's position, not each delegate: formally, it is the Head of Delegation's responsibility. (Other delegates responsibility is to support the Head of Delegation and support the BRM process to the best of their ability and their particular skillset and viewpoint. A delegate only needs to accede to his NB's position, not assent, even though I actually do agree with our NBs positions as I understand them.)

The primary position of trust for a delegate is from their NB and revolves around making sure that the BRM is successful. That is the purpose that BRMs are held, and a standards body should not send people who have any interest in not making the BRM succeed. JTC1 would undoubtedly take a very dim view of NBs who do not act according to the rules. (In this, it would seem far more likely that someone working for a company wanting to stop the standard has a conflict of interest at the BRM than someone wanting to help the standard, actually?)

So the conflict of interest you are talking about (applying to voting) cannot exist. The conflict of interest that relates to the technical experts at the BRM is severely limited in scope, and applies to lots of people on both sides. It is not special or unusual or big, is diluted by having 120 of us, and is managed by Standards Australia et al conducting themselves in the usual way they conduct contentious standards discussions: without favour and with transparency to them.

So how does this get handled? 1) I made sure that I declared potential conflicts of interest, even thought it makes me a sitting target for nongs and trolls and the reserve army of the self-righteous. 2) Standards Australia had me sign in writing a declaration that I would not be being paid by anyone but my company for attendance at the meeting, which I most certainly would not, that I would be on top of our committee's positions and so on. 3) Standards Australia then judged something like that there was no conflict that is different in scale or kind than is usual and supervisable.

People participate in national standards bodies committees because they are aware of and represent their particular interests, not because they are free of involvement. These national committees then provide the delegations, mainly from these same involved stakeholders. The representation and mediation of different interests and stakeholders is the bread and butter of standards work, not some blue-sky idea that standards magically appear by disinterested people having a collective lightbulb over their heads.

Marc: You are just being a troll now. The "XML Processor" is the name given to the parser, as distinct from the application.

"i) rw is paid to help some DIS proposed by a corporation to "fail" ISO technical-merits-judgement

ii) rw participates in the delegation of a national body and then rw must cast his opinion if the DIS has the technical-merits to pass ISO judgement"

there is a conflict of interest there, too

your response seems to be: other people are evil,,,so i can be evil. Strange argumentation..

"So what about the BRM? The purpose of the BRM is purely technical and editorial, so there can be no conflict of interest relating to non-technical and non-editorial issues."

BRM is *part* of a *process*. The next stage in this process will be a virtual vote re-casting where national bodies will decide if DIS 29500 has the technical merits to become a brand new ISO standard. This is a kind of judgement... call it technical , editorial or whatever; now you are judge and part, this is not ethical, like it or not ( IMHO ).

But your employer isn't a champion in ethics... two examples:

1998: vendor lock-in strategies [1]
2008: customer confusion playing with ISO standards: google for ODF[2] -> you get an sponsored link for OOXML.
( many national bodies have warned Microsoft that they don't like standards names playing, but Microsoft keep playing ... child attitude )

--marc

[1] http://antitrust.slated.org/www.iowaconsumercase.org/011607/2000/PX02991.pdf
[2] http://www.google.com/search?hl=en&q=odf&btnG=Google+Search

Marc: I am very happy if I can help Microsoft make the changes they need to make in order to become good corporate citizens and competitors-on-features-and-quality in a level-playing-field market.

The trouble with your POV is that in effect you are perpetuating the same game, with the same arguments that have failed to make much of a difference in the last decade. You think you are anti-MS, but actually your actions are trying to prevent the most promising correction of this decade. Now it is quite possible that MS will continue its domination for a long while yet, in particular for non-free software applications; time will tell how good the free implementations get when they have this documentation in a stable, RAND-z, QAed and open-maintained form, and what happens as OOXML-isms creep into ODF and vice versa. That will be a matter for competition; competition and plurality is good.

You need to sever this mental block you have, that it is impossible for MS to participate in anything that ameliorates the effect of their market domination: OOXML is happening because of user and regulator demand, and MS are acting like they are the mental switch to seeing standards as part of the cost of doing business and a way to gain back their system integrators.

There is no conflict of interest in me participating in my national body's technical committees, because there is no position of trust. There can be a conflict of interest in *almost anyone* from industry (and other stakeholders too...) participating in a delegation from a national standards body to the BRM, but for a technical expert it is severely limited, limited both because of the confined scope of the BRM to technical and editorial issues only and limited because it is the head of delegations job to vote and to check for balance. However, as standards bodies are set up to make industrial standards, the participation of people with domain expertise from industry goes to the heart of the purpose of the bodies, so any conflicts of interest have to be declared, managed and monitored by the standards body. It is no bar and not remotely unusual. Sorry if the world is not the way you think it is.

@marc:
What's your problem?

Rick is a technical expert in this field with decades of experience. If there's a point of contention remaining during the BRM and a solution needs to be found, Rick's experience and hard-earned wisdom has a chance of helping to guide the parties to a mutually acceptable solution. There are other experts too, of course (as well as plenty of people, I'm sure who are not actually experts and who will slow down the meeting pointlessly).

Now Rob Weir is going to this too. He seems like a pretty smart guy and appears to have some experience in document processing. But until this whole OOXML mess, he was a "Performance Architect" at IBM, not necessarily a document person. And he has demonstrated a desire to push issues that are divisive and ultimately minor just to stymie the spec. Rob's role is useful too, though, because in attacking the spec he points out areas for improvement that more pro-OOXML people can fix to everyone's benefit.

Essentially, the conflicts of interest will probably be managed by the moderator (Alex Brown)'s actions in keeping the meeting moving in the face of the more mendacious criticisms of the Rob Weirs in the room while ensuring that ECMA reps (and Microsofties) do not sweep legitimate issues under the rug. Nowhere in this process does the OpenForum Europe event fit in, and hopefully the National Bodies will police their own members to ensure that no one goes on nightly excursions sponsored by either Microsoft or its opponents.

nksingh: Thanks for that.

But what is wrong with OpenForum Europe that people shouldn't go? I don't know anything about it, but usually it is good for delegates to talk and brainstorm with as many other delegates as possible about the issues of the day. I might go myself, if there are other delegates going. I don't know how people can have confidence in their views unless they get exposed to a mix of viewpoints. Will they be having fondue?

@Rick:
The OFE is a pretty transparent lobbying organization against OOXML. Based on their public statements, they seem to intend to give talks to the delegates about IPR issues, anti-trust, and general competitive marketing of technologies. None of these categories have much to do with the stated purpose of the BRM to resolve technical issues in the spec, and they don't strike me as legitimate reasons to reject a Standard proposal. For fairness and integrity of a standards process, the party proposing the standard should not matter so long as they are acting in good faith.

The OFE event looks like an attempt to picket the BRM. It seems like a rather silly thing to do at a technical (and by design non-political) conference.

@nksingh
"None of these categories have much to do with the stated purpose of the BRM to resolve technical issues in the spec"
which is by design. The OFE conference is not directly going to address technical issues because that is what the BRM is for.
"and they don't strike me as legitimate reasons to reject a Standard proposal"
yes, they are. In Alex Brown's FAQ you will note that point 6.7 reads "No constraints are placed upon the criteria NBs may use for deciding their voting position."

@Rick,
great question about Fondue, I will try to find out.
There are indeed several people with different points of view sailing close to the wind on the 'conflict of interest' issue. I fully expect everyone to put everyone else under close scrutiny and we appreciate the clarification of your position as a technical expert alongside the head of delegation and that you have been working for Microsoft and not ECMA.

C'mon Rick, a blind dingo can see the MS-OOXML standardisation push for what it is: a way for MS to continue to sell Office to governments with ISO document format requirements without having to surrender their lock-in by providing first class support for ODF.

If MS were genuinely serious about high fidelity access to existing documents, they would be submitting the legacy binary formats to ISO, not a new-fangled spec with zero working implementations (no, not even MS Office).

Great post in Groklaw from grokker59!!!
http://www.groklaw.net/comment.php?mode=display&sid=20071023002351958&title=Rick+Jelliffe&type=article&order=&hideanonymous=0&pid=673248#c673386

I'll include his post here:

Rick is good with the MS propaganda - perhaps that's why he seems to be in
popular demand by Microsoft....

I've pointed out some examples on Andy Updegrove's blog of Rick's more subtle
FUD.

Example #1: Rick asks Andy "as a lawyer" to identify which laws were
broken by MS' OOXML activities...

This is FUD because it is a misleading question designed to mis-direct the
reader. By using the pre-condition 'as a lawyer', then specifically asking for
'laws', Rick tries to limit the discussion and the scope of MS' OOXML activities
to either "broke the law" or "did not break the law", hoping
the reader does not engage in critical thinking as follows:
A. ISO/IEC is an international committee. As such, it does not fall under
the legal system of any one nation.
B. Asking a US lawyer to identify any international laws that MS may have
violated is like asking a farmer for the 5-day weather forecast on Jupiter.
Andy's law training is not international in scope, nor does it cover the entire
world.
C. By specifying 'laws' that may have been broken, Rick distracts from -
and puts out of play - any rules, regulations, documented policies, bylaws,
committee best practices, historical precedent, expectations of civilized
behavior (comity), etc on the part of either JTC1 or NB committees around the
world, or procedures for document handling, etc. It also puts 'beyond the
scope' any deviations MS may have made concerning openness of process or
irregular practices or unethical behavior - what many call 'stacking the
committees'...

Example #2:
Rick uses the techniques of 'faint praise' and 'partial truth' to selectively
portray the OOXML as a net gain for the world because he sees it as MS opening
up - not as the legal trap that everyone else sees.

For example, when telling a story, it sounds good to the average layman but
it's hard to tell why it sounds 'sour' to someone familiar with the topic. One
of the things I picked up on was his diction (choice of adjectives, verbs,
nouns, and adverbs) - disparaging when describing FOSS, positive when describing
MS products. For example (my own words): A beleaguered Microsoft tries to
salvage it OOXML specification during this month's BRM in Geneva dispite claims
from a rival clique consisting of IBM and other radical open-source elements
that a second document standard is not necessary. Yet we have multiple, often
overlapping standards today - HTML, PDF, RTF, etc - even before the addition of
Sun's ODF format. Why not allow Microsoft to standardize their formats as well
?"

Note the use of adjectives and choice of nouns that develop a bias in the
reader against IBM & FOSS. Note the one-sided approach to this statement.
Note the partial truths or characterizations ('clique' instead of 'consorium',
'radical' associated with 'open-source', the term 'open-source elements' used as
it would be in 'criminal elements'. By describing Microsoft as 'beleaguered',
he makes Microsoft sound like the victim of the story. By having Microsoft
'salvage' their OOXML spec, he tries to make the reader feel that Microsoft is
making an attempt to save something that is good and valuable under very trying
conditions or at great expense instead of describing Microsoft as the aggressor
trying to push their standard on a reluctant standards committee.

I suspect that Rick is also being paid to use his word-smithing skills and his
analyst credentials and credibility to visit and play "devil's
advocate" on various open-source blogs where his reputation may convince
new or less-informed readers that there really is room for two document
standards or that Microsoft has really done nothing wrong because they've broken
no laws and because their offerings comply to the minimum legal letter of the
law as set forth by the ISO committees. This statement invites the readers to
ignore the other issues (rules, procedures, customs, expectations of fair play,
etc) that Microsoft can be said to have either violated or ignored in their zeal
to get their standard approved.

I most certainly do *NOT* expect Rick Jelliffe to admit or publicize whether or
not he is actually being paid by Microsoft to 'seed' blogs or to post
OOXML-positive articles on various blogs with the weight of his reputation
behind his postings. I can only say that I've now seen Rick Jelliffe posting on
4 or 5 blogs and his postings are always pro-Microsoft and pro-OOXML despite all
evidence to the contrary - more than sufficient evidence than should be required
to convince a technical person like Rick that foul play is (and has been) afoot.

@Nick:
Even a deaf walrus can gather that the ISO version of ODF is so under-specified that Microsoft or any other vendor could implement more or less whatever they want under extensions. You'd end up with a worse morass than you see now. Be careful of what you wish for. N.B. Office has a strong set of other compelling features that the product would not be particularly compelled to follow any of the non-standard ODF conventions employed in ooo.

Writing a competitive office processing system is no small endeavor (even simple plain text editors like VIM or Emacs have been refined over decades and couldn't be replicated all that quickly). If the file format is the thing tripping you up when the format is in XML with a more-or-less complete set of documentation, then perhaps you're not ready to implement the higher-level features that actually make Office worthwhile.

@NSingh,

What are you talking about? I don't think anyone has issues with MS Office 2007 using Microsoft's convoluted file format (which is vaguely similar to MS's OOXML spec, but is not compliant, if I understand correctly) for storing files if users want to.

What we object to is making it an ISO Standard. There's already an ISO Standard for office documents: ODF. It was there first, and MS supported it. They're just being a petulant child now, trying to rewrite the rules now that they've realised that they've created a situation in which they might - horrors - have to compete on a level playing field.

Here's the thing: it's not unusual for an office package to support multiple file formats, even those that cannot reflect the full richness of the content the package is capable of expressing. You may recall that before MS Office had its monopoly position, it was very good at reading in WordPerfect files (as well as other competing formats - the fact that it didn't write them was a tactical decision, not a technical one).

MS Office has almost always supported writing files in RTF format as well as older MS Office binary formats, all of which are (in theory at least) less expressive than those specified in the MS OOXML spec, right? So how can MS, with a straight face, insist that it cannot provide ODF compatibility on the same basis as it does for those other "lesser" formats in its MS Office package? Microsoft most certainly can, but it won't (for tactical reasons, not technical ones) and it doth protest too much.

No one cares whether or not OOXML is a better spec technically (it's not, if only for the fact that it fails to incorporate other established standards, and insists on creating a whole new constellation of parallel standards). Microsoft is trying to sell the concept that specifying distances in inches, yards and leagues and speed in knots is somehow better than using centimeters, meters and km/hr because the former Imperial units are somehow "richer". Bullsh*t. Microsoft and its agents - including Rick (shame on you) - are disingenuous, and should be treated as the shysters they are: with total contempt.

Regards,

Dave

Phil Again: You are getting more and more hilarious.

Are you really making up quotes ("my words"), then parsing them ("note that") to prove bias? When have I used "clique"? Do you really believe that any my readers are dumb enough not to see through it? I think you are just another troll.

As for asking Andy Updegrove about legal issues, that is his expertise and angle, and he had raised issues about a European investigation, which presumably is looking for some illegality or incorrectness. Why isn't it an obvious question. It is like when some people say "X is a criminal" it is obvious to ask "what is the crime" and even "what is the law that makes that a crime?"

Nick: But the trouble is that the ISO standards process has to operate fairly. That means that it allows an open forum and helpful process to anyone, even Microsoft.

Now MS *could* be doing this for all the most evil and noxious purposes, but that is no reason for ISO processes to be acting any different way than they do normally. There is a process, it is being followed in the full glare of public scrutiny, and the neutral custodians of the process (ISO and standards bodies, who are quite used to controversy though not on this scale) are making sure that things are being done by the book.

My analogy is with the Post Office. The mailman should not refuse to deliver mail to prisoners.