Complex singularity versus openness?

Thu Jul 3 09:32:28 CEST 2014

Dear Murata-san,

One of the thoughts that came to me while reading (some of) the article was, that up-front, the article seems thorough and fact-based. But if you examine the written text, you quickly realize that the article is one-sided, biased and even though the article is "fact-based", it only deals with a very limited part of the facts available. You really need a lot of knowledge (as Jim said, dating back to 2008 ☺ ) to remove the reality-distortion field the author was clearly affected by when he wrote the article.

So I am not surprised that people everywhere (including the EPUB community) believes this article, it was also thrown in my face at http://www.version2.dk/blog/odf-12-i-iso-pas-submission-67510#comment-280080 … but that doesn't make it any more right.

With regards to your proposal about SC34/WG4 doing something, I don't really see the need to do any more than we already do. And I certainly do not think we should try to do any updates with regards to ODF. That must be the task of the ODF TC.

Med venlig hilsen / Best regards

Jesper Lund Stocholm
Senior Architect

t: +4544662466
m: +4530945570
jesper.stocholm at ciber.com

From: eb2mmrt at gmail.com [mailto:eb2mmrt at gmail.com] On Behalf Of MURATA Makoto
Sent: Thursday, July 03, 2014 5:46 AM
To: SC34
Subject: Re: Complex singularity versus openness?

Folks,

One of my colleagues in the EPUB community believed this article
and twittered about it.  Another colleague (again in the EPUB
community) retweeted.  They are not all committed to ODF.
They are credulous.

I do not think that WG4 should respond to this article, but
it might make sense for SC34 to provide some status
update to the public.  That status update should cover
both ODF and OOXML.  How do people feel?

Regards,
Makoto

2014-07-03 7:02 GMT+09:00 Jim Thatcher <Jim.Thatcher at microsoft.com<mailto:Jim.Thatcher at microsoft.com>>:
Ah, yes, back to the “open” only means what we in the pro-OpenOffice crowd say it means. My, how I miss 2008!

This article is so full of inaccuracies, half-truths, outdated information and outright propaganda that it’s just sad. But as a Microsoft employee there is no benefit from me posting a response for two very different reasons. First, since Microsoft Office now has a very high quality implementation of ODF, it makes no real difference to our business if the pro-ODF cheerleaders whip up a few public sector groups to only use ODF, and second, the audience of this article completely filters out anything that Microsoft has to say because, in their view, Microsoft is just evil. So I’ll just respond here, because I know that those on this mailing list are actually open to things like facts. And this group recognizes the hypocrisy in complaining that Open XML was originally based on Microsoft’s implementation without ever mentioning the origins of the ODF specification.

The lack of integrity in this article begins with the title, and becomes very evident in the subtitle. The subtitle claims “Open source impeded by incompatibilities and inconsistencies in the Office Open XML document format”. Although the author identifies a couple of bugs in Microsoft Office, no attempt is made to actually identify specific incompatibilities or inconsistencies in the Open XML standard. Instead he just asserts that the mere existence of “three different versions” results in insurmountable complexity. And of course the author completely ignores the three different versions of ODF and the “incompatibilities and inconsistencies” between those versions. But those who are willing to be honest and actually read the Open XML standards (including the Ecma 376 First Edition) recognize that coding for the differences between those three is not rocket science, and most university CS majors are capable of handling such “complexity”. As I recall, basic if/then coding principles are taught in pretty much every introduction to programming course. If you can’t figure out how to use existing XML parsing libraries to look at the namespace declarations in a .docx file to determine which edition of the Open XML specification that document is based on then you have no business changing code in any office productivity application. (Did you notice how I used the if/then approach there?)

What the author of this article really means when he says “to date there are no free and open source solutions that fully support OOXML” is that the developers of OpenOffice and LibreOffice either have chosen not to invest or simply don’t have the necessary development expertise to do a professional quality implementation of Open XML. There certainly wouldn’t be any reason to confuse readers with actual facts about open source implementations of Open XML, including the significant improvements in Open XML support in the most recent update to OpenOffice. The author completely ignores another post on the same site that directly refutes his claim. The article at https://joinup.ec.europa.eu/news/three-five-biggest-ooxml-support-issues-fixed-open-source-office heralds the resolution of three of the five biggest Open XML gaps in OpenOffice and LibreOffice. But apparently the OO/LO developers have been completely incapable of resolving the other two issues in the 18 months since that article was posted. And apparently the meaning of open source solutions doesn’t mean what I thought it did, because one of the best third-party implementations of Open XML is the Apache POI project, which is used by at least two really solid office suites for Android devices. Last I checked, Apache projects are still open source.

Perhaps the OO/LO communities should look to China for some new, capable development talent. Over the past few years CS students at Beihang University and BISTU in Beijing have developed a pretty robust translator that converts documents between Open XML and Uniform Office Format (China’s XML-based document format standard). The recently released update to the UOF Translator adds support for translating “Strict” Open XML documents. These university students were able to master the “complexity” that apparently has completely stumped the “experts” Mr. Feilner cites in his article to handle both Strict and Transitional documents in the same code. Oh, and this project is also open source, so maybe the experts can look to that code for some help.

And the claims that ODF is implemented by more implementers and on more platforms than Open XML is clear evidence that the author and his “experts” have no interest in painting an accurate picture. Every major office suite on every major platform has Open XML support, but the same is not true for ODF. They are very generous as they count all of the vendors supporting ODF, but they do not afford the same generosity to implementations of Open XML. They want to have it both ways by counting Microsoft Office as an implementer of ODF (we agree that Microsoft Office versions since Office 2007 SP2 should be counted as ODF implementations even though we have publicly stated that we don’t implement ODF’s change tracking) but not counting OpenOffice or LibreOffice as implementations of Open XML because they aren’t “complete” implementations.

But the reality is that this author, and the “experts” he has consulted, have no real interest in improving the Open XML standard, or even developing enough expertise to understand and implement it well. Their real interest is in attacking Microsoft. The small investments they make to implement Open XML are unnecessarily complicated as they reverse engineer files written by Microsoft Office. Why? Microsoft publishes full details of our implementation of ISO/IEC 29500 (and Ecma 376, First Edition), including detailed specification of each of the extensions we have implemented (using MCE). We have even published complete details of our implementation of ODF. If you want an exercise in futility, try to find published details of OpenOffice’s variations from the ODF standard, or specification of the development choices they made when the standard was not clear?

I could go on and on, as I’m sure any of you could as well, pointing out other errors and efforts to mislead in this article. But for what purpose? This group already knows these details, and the audience of that disingenuous article will not be swayed by something as trivial as reality.

So I’ll join Jesper and take a couple of Tylenol for my trouble.

Best regards,
Jim Thatcher

From: Jesper Lund Stocholm [mailto:Jesper.LundStocholm at ciber.dk<mailto:Jesper.LundStocholm at ciber.dk>]
Sent: Monday, June 30, 2014 11:49 PM
To: MURATA Makoto; SC34
Subject: RE: Complex singularity versus openness?

Dear Murata-san,

I saw the article a couple of days ago, but my head started hurting after the second paragraph, so I didn't make it through the whole article.

☺

Med venlig hilsen / Best regards

Jesper Lund Stocholm
Senior Architect

t: +4544662466<tel:%2B4544662466>
m: +4530945570<tel:%2B4530945570>
jesper.stocholm at ciber.com<http://jesper.stocholm@ciber.com>

From: eb2mmrt at gmail.com<mailto:eb2mmrt at gmail.com> [mailto:eb2mmrt at gmail.com] On Behalf Of MURATA Makoto
Sent: Tuesday, July 01, 2014 4:22 AM
To: SC34
Subject: Complex singularity versus openness?

Dear colleagues,

I find this article from EC.

https://joinup.ec.europa.eu/elibrary/case/complex-singularity-versus-openness

Regards,
Makoto

--

Praying for the victims of the Japan Tohoku earthquake

Makoto
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.vse.cz/pipermail/sc34wg4/attachments/20140703/71ee3338/attachment-0001.html>