Is Red Hat Opposing Document Standard to Prepare Ban on Windows?

Add Red Hat (RHAT) investors to the list of shareholder communities being poorly served by their corporate management in the matter of opposing a group of public companies and well-respected organizations trying to get a new document format standard approved by the International Standards Organization [ISO] this month. The group that Red Hat opposes includes Apple (AAPL), Barclays Capital, BP, The British Library, Essilor, The Gnome Foundation, Intel (INTC), Microsoft (MSFT), NextPage, Novell (NOVL), Statoil (STO), Toshiba (TOSBF.PK), the United States Library of Congress, and ECMA International. The standard under review is Office Open XML [OOXML], the most popular implementation of which is Microsoft Office 2007.

In a standards process like this, a regional or functional organization such as ECMA International (formerly called the European Computer Manufacturers Association) takes the lead when going to ISO. By way of comparison with OOXML, AiiM carried the water to ISO for Adobe (ADBE) for .pdf and the OASIS Group (whose foundational sponsors are BEA (BEAS), IBM (IBM), Primeton, SAP (SAP) and Sun (JAVA)) was the conduit for Sun's Open Document Format [ODF]. Red Hat says its primary objection to the ECMA standard is that there already is an ISO standard, Sun's ODF, for document formatting. But of course there are already multiple ISO document format standards such as Adobe's--with others coming.

So what does it matter if there are three or four or five? Red Hat's position is nonsensical unless you read between the lines. From an IT investment research point of view, you would think an infrastructure software supplier like Red Hat (Linux and JBoss middleware, which runs on top of both Linux and Windows) would want multiple suppliers offering application products in the "international" marketplace in order to sell more infrastructure software. This is especially true for JBoss, whose sales--as opposed to "free" open source distributions--have taken off slower than Red Hat had expected.

But apparently Red Hat is taking the restrictive position of trying to close off one market to Microsoft--even one that Red Hat itself does not compete in--on the hopes that eventually it can also convince governments around the world to close off the operating software market to Microsoft's Windows. Although Red Hat currently sells the Linux operating software primarily to replace 20- and 30-year old Unix operating software, it has said it plans on taking 50% of the infrastructure software market by 2015. Red Hat apparently believes to reach that goal it will also have to compete against Microsoft Windows server operating software at some point.

Red Hat also objects to the ECMA standard because Microsoft supposedly hasn't released some obscure specifications from the 1990s. Microsoft has released 30,000 pages of specifications and would certainly release 30,001 if need be.

Finally Red Hat criticized ISO's fast-track process. But of course, that's ISO's process--one it uses to standardize household items, airplanes, bowling balls, you name it--so to criticize Apple, Barclays Capital, BP, The British Library, ECMA, Essilor, The Gnome Foundation, Intel, Microsoft, NextPage, Novell, Statoil, Toshiba, and the United States Library of Congress because of the ISO rules is an incredible stretch. As Dr. Istvan Sebestyen, secretary general of ECMA explains via email: "In case of OOXML we tried to be very open, we have put up the first public draft in Summer of 2006. At the same time we intensified liaison with (ISO's) JTC 1 Sc34, and invited them to participate in our meetings. They also had our interim drafts. We have also installed on the internet a public response channel, on which interested people could submit their reactions to the draft.

"... (and) the standardization of PDF 1.7 fast track at ISO TC171 has been actually faster."

Comments

[Doode]

Are you on crack? The OOXML "standard" is Microsoft's attempt to get their proprietary document format the label of a "standard". They already have a ready made implementation in the form of Office 2007 as you yourself pointed out. I guess Microsoft has never done anything to protect their turf, oh, wait, the anti-trust actions were in another universe where Gore became president.

As for the other parties interested in OOXML, the US library of Congress is the most significant public voice. Their interest is quite simple - its better to have a Microsoft document format standardized that not. People are already using OOXML as part of Office 2007 and they would really like these documents created today to at least be readable in a few decades.

RedHat is an open source company. They back open source technologies and are doing the right thing (TM) in supporting ODF, a true open source format that anyone can use, including Microsoft. The company should be commended for sticking to their principles and business philosophy. Yeah, now go rant about open format doublespeak to Clippy. Remember to turn on the office spellcheck first.

[ThomasCameron]

A colleague of mine named Bryan Smith here at Red Hat posted this, I thought it was a really great argument about the OOXML "standard:"

"Top 10 Technical Reasons why [Office] OpenXML (OOXML) is NOT a Standard
at All."

10. Marketing: XML is not an application standard

Uses XML in name [Office] OpenXML (OOXML) to prey upon ignorance of what
XML is. XML is a template standards for writing vendor standards, not
an end-application usable standard. It still requires a full set of
support specifications, and even, that does not guarantee
interoperability at all. OOXML not only does not have a full set of
support specifications, but purposely avoids them (see #5-7 below).

9. Poor History: Lack of standards in even RTF

Microsoft has a horrendous history on any alleged standard. Most of its
products do not even qualify as "proprietary"... standards, because that
would require it to assign value and maintain compatibility for many
versions. It, instead, chooses an approach of purposeful
"hostageware"... (data is hostage after a few versions) aka "abandonware.&quo...

A great example of this is Rich Text Format (RTF). Not only does RTF
undergo revisions with every MS Office (specifically Word) release, but
Microsoft exports RTF from MS Office with embedded DOC (Word) attributes
that are not even in the RTF version standard. I.e., all office suites
(even Microsoft Office itself) use the Microsoft DOC (Word) import
filter to read RTF, because even a latest RTF filter does not work at
all.

8. Poor History: The abandoned XML "standard" of Office 11 (2003)

This is not Microsoft's first attempt to use XML. Microsoft Office 11
(2003, 2004 on Mac) offered allegedly XML support. The XML support was
limited to two things: 1) third party content import (so third party
applications could use MS Office as a "fat" content system/service) and
2) a "dumb content" XML export (no style, no support specifications at
all). This "standard" has been utterly abandoned. Will Office 12
(2007, 2008 for Mac) be yet another repeat once Office 13 arrives?

7. Core Issue: Designed to encapsulate binary data formats

OOXML as implemented in Office 12 (2007, 2008 for Mac) is designed to
largely address document compatibility by encapsulating older,
"abandonware"... (not even "proprietary"... standard, but time-limited formats
that should be considered "abandonware"... binary data formats in between
XML tags. These include the regularly conflicting attributes in Office
8 (97), 9 (2000), 10 (XP) and 11 (2003) for Windows formats, and let let
alone made worst in their Mac equivalents (98, 2001, X and 2004,
respectively).

6. Core Issue: Designed for undocumented, binary data formats

The regularly conflicting attributes for style (and even content) in
Microsoft Office 8, 9, 10 and 11 binary data standards are poorly
documented (if at all) in the OOXML specification. In fact, in several
areas (especially in areas of locale, date/time, etc...), the Open
Document Format (ODF) specification does a far better job. And even
worse is the fact that Office 12 (2007) for Windows implementation seems
to introduce some new undocumented, but encapsulated binary data
formats. I.e., this trend of using undocumented, binary data formats --
only now encapsulated between XML tags -- will yet still continue going
forward in Microsoft Office implementations.

5. Core Issue: Differing interpretations of binary data formats

The Office 12 (2007, 2008 on Mac) implementation can only run into a
"compatibility mode" to get the same interpretation of Office 11 (2003,
2004 on Mac). Office 8, 9 and 10 still differ from that, especially
when more revision apart. It is purposeful hostageware (your data
becomes unable), commonly known as abandonware. Furthermore, Office
12's native mode implementation still cannot read some of these binary
formats, while having no equivalent or import. The hope is that by
encapsulating some of these formats, Microsoft itself hopes (again,
hopes) to be able to read them at some point (with added patches).
Again, none of these are documented in the OOXML specification to a
point of reproducible context+style (if at all).

4. Core Purpose: Designed to solve data alignment issues (portability)

The primary purpose of [Office] OpenXML (OOXML), as technically detailed
in #5-7 above, is to solve an engineering nightmare in the Office
division -- namely and currently, the Mac and Win64 ports -- by and for
Microsoft itself. Win32 and the entire legacy of Office 8-11 for
Windows, as well as the Office 11 (2003) compatibility mode of Office 12
(2007) for Windows, is not portable to either Mac or Win64. Mac ports
are done by another division, and binary write compatibility is limited
-- e.g., Mac can "read" Windows files (with an intelligent, byte-by-byte
import), but cannot "send them back" to Windows (which has exacting data
alignment assumptions, which also causes compatibility issues between
even Office for Windows versions itself).

I.e., there are data alignment assumptions in Office for Windows (Win32)
that break in Office for Mac, and any planned, native Win64 port. By
breaking down binary formats into bytes, encoding and encapsulating them
in between XML tags, non-Win32 implementations of Office 12 (2007) --
e.g., Office 2008 on Mac and future, planned, native Win64 relesaes --
can maintain better 2-way data exchange of binary formats, as they are
encapsulated/encoded at the byte level (eliminates alignment issues)
between XML tags.

Again, these encapsulated binary formats are poorly documented (if at
all) in most cases in the OOXML specification.

3. Avoids standardization: Does not implement support standards

Office 12 (2007) makes no effort to incorporate other, standardized
(e.g., ISO, OASIS, W3C) support formats -- e.g., MathML. This is unlike
virtually every standard documentation and publication language, even
pre-Web. It continues to rely on legacy, regularly conflicting, Office
8/9/10/11-based attributes, some not available as full context+style in
the OOXML document. This contributes heavily to the bloat of the OOXML
specification, while not being remotely as reproducible as ODF.

Further Comparison/Perspective...

ODF is based on the legacy of StarOffice, which actually pre-dates
Microsoft Office (and has always had better integration and feature
support, especially for the Internet). Given that StarOffice "made the
switch," Microsoft's decision not to switch to supporting other,
existing standards is purposeful in comparison. MathML is probably the
biggest example, which leverages broad compatibility (e.g., even TeX
interchange) and should have been a mandatory move from an engineering
perspective.

2. Avoids standardization: Lack of scrutiny in process

Per page, specification has been scrutinized 1/100th less than Open
Document Format (ODF). It is 10 larger and has undergone 1/10th the
time period for review by ISO. Ironic that a large organization can
spend less and prove far less than a coalition of not just open source
developers, but sister standardization organizations to the ISO (e.g.,
OASIS, W3C) and industry leaders in large, major documentation.

1. Biggest Issue: Implementation does not match specification

Office 12 (2007) implementation does _not_ match Office OpenXML
specification. A standard that presented, but not implemented, let
alone not even intended and never will be implemented, is not a
standard. It is XML, which is standard for creating standards, not any
indicator of an actual, open standard. As an additional note, even
Microsoft Office 12 (2008) for Mac lacks the VBScript, so its
compatibility, "as implemented," is even worsened.

As I always say, "Microsoft Office for Linux would be as incompatible as
Microsoft Office for Mac, especially when sending documents back to
Microsoft Office for Windows users." Microsoft itself has proven it
cannot even offer a compatible implementation between several versions
on Windows (Win32), let alone to other platforms such as Mac and its
own, inevitable Win64 API.

[ThomasCameron]

Should have mentioned that I did clear it with Bryan before I used his text.

Also - obviously I am not representing this as Red Hat's position, just mentioned I work there for disclosure purposes.

[steveballmer]

OOXML will pass! We shall prevail no matter who we must pay off , i mean lobby!

[Mr.Penguin]

Unbelievable. A puff piece to promote OOXML, and an obvious one at that! This URL should be flagged as a prime example of the FUD machine in action. It doesn't get any more blatant than this.

I really have to wonder about the author's motivation to write this article. People who have read the standard and understand what it means would have a difficult time reaching these conclusions.

This article trashes the credibility of everything on the entire website. Very scary to think that investors read this stuff and then make big decisions.

[Peter Green]

I also work for Red Hat and used to work for Sun 1995 - 2002 and Novell 1988 - 1992. I have plenty of experience competing in a fair and legal way.

I have studied MS behaviour for 20+ years and as is well documented elsewhere this issue is continuing evidence of 'bad' possibly illegal behaviour.

I absolutely agree with the sentiments expressed paticularly Thomas & Mr. Penguin.

I have been a fan of seekingalpha for a while but this damages your credibility with me.

[Dennis Byron]

In answer to Mr. Peter Green, thanks for your rational comment.

Just so you understand how seekingalpha works, I am an independent analyst/consultant that blogs on my own web site and at other sites with which I am affiliated. Seekingalpha picks up my posts when it wants and when it feels the content is germane to its readers. Please do not stop reading Seekingalpha because you do not agree with my opinion.

Since you say you agree with Mr. Penguin, the Linux devotee who commented before you, please understand that all my posts--the ones that SA picks up and those it chooses not to--are written from the perspective of the investor, primarily the institutional investor. My blog post says nothing pro or con about any particular standard but simply asks the question: Why is a publicly traded U.S. company with very specific legal governance obligations to the shareholder, wasting its time and the shareholders' money on an issue related to an IT market (document processing software) in which it does not even compete? A rational investor can only conclude that Red Hat wants to further an agenda that will in fact help it in a market in which it does compete at some point in the future.

(Therefore, I have no idea if Thomas'/Red Hat's 10 points are correct nor do I care. I do know that at least one of them is misleading. It is correct to say that Microsoft has itself not implemented the standard in question. But that is because the ECMA has changed the standard in the last two years based on input from something like 87 national ISO standards bodies. Microsoft cannot implement the standard until ECMA finishes the standards process. Red Hat's press statement, which I reference at the beginning of my blog post, includes similar red herrings.)

At this point, the investors in Red Hat (RHAT) are ill served by trying to get governments around the world to "legislate" (or even worse-"edictize&q... the brands of IT products their citizens will pay for and use. Despite Mr. Whitehurst's postion as reported in Infoworld on 3/27 that "George Bush is good for open source," it is an unbelievably short-sighted position. One American company that is very well respected in the open source community has already thought through the consequences of this position on its blog (see hyperic.com/blog/hyper.../).

How soon before governments will legislate against companies based on the geographic location of their headquarters? Or the religion of their owners. Or the color of their owners' skin? Red Hat is on a slippery slope and its investors need to understand that.

-- Dennis Byron

[Peter Green]

Dennis Thank you for your answer, But I think you miss my point. MS has forced this through in a way that is typical of it's previous behaviour. This threatens many competitors both directly and indirectly. To suggest that a company is not serving it's shareholders by spending time on the issue if only indirectly threatened is wrong in my opinion.

Red Hat's position is very clear on the support of Open Standards. Microsoft's patent threats are also very clear.

Microsoft's behaviour may have been, in a narrow sense, good for it's shareholders but they have been bad for the rest of us to such a degree that I believe competing companies are often, if not always, serving their shareholders best interests by opposing Microsoft.

Regards - Peter Green


On Mar 29 03:23 PM Dennis Byron wrote:

> In answer to Mr. Peter Green, thanks for your rational comment.
>
>
> Just so you understand how seekingalpha works, I am an independent
> analyst/consultant that blogs on my own web site and at other sites
> with which I am affiliated. Seekingalpha picks up my posts when
> it wants and when it feels the content is germane to its readers.
> Please do not stop reading Seekingalpha because you do not agree
> with my opinion.
>
> Since you say you agree with Mr. Penguin, the Linux devotee who commented
> before you, please understand that all my posts--the ones that SA
> picks up and those it chooses not to--are written from the perspective
> of the investor, primarily the institutional investor. My blog post
> says nothing pro or con about any particular standard but simply
> asks the question: Why is a publicly traded U.S. company with very
> specific legal governance obligations to the shareholder, wasting
> its time and the shareholders' money on an issue related to an IT
> market (document processing software) in which it does not even compete?
> A rational investor can only conclude that Red Hat wants to further
> an agenda that will in fact help it in a market in which it does
> compete at some point in the future.
>
> (Therefore, I have no idea if Thomas'/Red Hat's 10 points are correct
> nor do I care. I do know that at least one of them is misleading.
> It is correct to say that Microsoft has itself not implemented the
> standard in question. But that is because the ECMA has changed the
> standard in the last two years based on input from something like
> 87 national ISO standards bodies. Microsoft cannot implement the
> standard until ECMA finishes the standards process. Red Hat's press
> statement, which I reference at the beginning of my blog post, includes
> similar red herrings.)
>
> At this point, the investors in Red Hat (RHAT) are ill served by
> trying to get governments around the world to "legislate" (or even
> worse-"edictize&q... the brands of IT products their citizens will pay
> for and use. Despite Mr. Whitehurst's postion as reported in Infoworld
> on 3/27 that "George Bush is good for open source," it is an unbelievably
> short-sighted position. One American company that is very well respected
> in the open source community has already thought through the consequences
> of this position on its blog (see hyperic.com/blog/hyper...
> /).
>
> How soon before governments will legislate against companies based
> on the geographic location of their headquarters? Or the religion
> of their owners. Or the color of their owners' skin? Red Hat is
> on a slippery slope and its investors need to understand that.

>
>
> -- Dennis Byron

[Dennis Byron]

Peter,

No--as noted elsewhere on SA multiple times--I don't think participation in the Open Standards movement is a good use of Microsoft shareholder value either.

My point is that I think Red Hat's participation is not the "for the common good" activity you think or hope that is. Red Hat seems to want governments to pass laws or issue edicts that favor one brand of software over another.

(I think you are adding patents to your objection for the first time, no? That's a separate discussion and it is ongoing here on SA on a different thread. But it has two forks: current patent portfolios and whether the U.S. will continue to grant them liberally in the future. The current SA discussion considers only the latter patent-issue fork)

- Dennis