Posted on 04/01/2008 7:00:13 PM PDT by Bloody Sam Roberts
SEATTLE (AP) — The format Microsoft Corp.'s Office 2007 programs use to save documents was approved as an international standard Tuesday, a step the company touted as proof it is willing to make once-proprietary technology work openly with competing programs.
But the International Standards Organization vote didn't quiet some opponents, who argued that the Office Open XML standard still locks out competitors and gives Microsoft customers no choice but to keep buying its programs forever.
The decision was made public on the Web site of a European standards organization, Ecma International, on Tuesday. ISO is expected to formally announce the vote Wednesday.
(Excerpt) Read more at ap.google.com ...
I agree that is very disturbing.
However, it's almost impossible for a large company to prevent each and every one of their employees from doing something stupid.
Microsoft said the offer, when discovered, was quickly retracted and that its Sweden managers voluntarily notified the SIS, the national standards body.
"We had a situation where an employee sent a communication via e-mail that was inconsistent with our corporate policy," said Tom Robertson, general manager for interoperability and standards at Microsoft. "That communication had no impact on the final vote."
The offer was rescinded by upper management as soon it was discovered that it was made, and Microsoft informed the standards body about the incident.
From the same article:
But bloggers claiming to have been present at the SIS meeting wrote (here and here) that more than 20 companies showed up in the waning moments of the meeting with the sole intent of voting in favor of Open XML.
While that's being used to make it sound like Microsoft was buying votes, it's not unusual.
Standards meetings are EXTREMELY boring affairs. At any one time usually only about 20% of the people there understand or even care about the particular portion of the standard that is being discussed.
It's like sitting in a really boring class where 90% of the material won't even be on the test. A lot of people cut class.
You go to these meeting every couple of months. You learn who is knowledgeable about what and working on what. If you have a particular issue with something, you can show up at the meeting, sit through a couple hours of things that don't concern you, and 90% of which are editorial issues rather than content issues. If you're lucky the meeting won't get sidetracked into some argument between two opposing groups who want different implementations (the usual result is that both implementation get included), and neither group is completely happy. If you're very lucky they might even get to the topic you are concerned about and you can make your concerns known and try and get something changed.
Or you can have lunch with a guy you know from another company that is deeply involved in that standard, express your concerns and have them bring them up in the meeting, and since they have a better understanding of the standard as a whole they can probably make the point better than you can if it is a valid point.
When important votes come up, attendance swells. A lot of times they need to go round up all the people that are out in the hall talking on their cell phones, trying to get work done.
The vote gets held, then over half of the people in the room either disappear, or go back to working on their laptops.
People think that standards are some big universally accepted thing.
Most of the standards I saw approved cam out of small working groups, and were put forth by one or two companies who were interested in gaining wider acceptance for their products through standardization, and few of they became widely used outside of that small group.
In the main meetings where letter ballots are voted on for standards coming out of committee, most companies vote in favor of whatever the working group decided simply because the issue doesn't concern them. Most of the comments that are generated are editorial to make the standard easier to read and understand, but I never saw a standard shot down because another group didn't want it standardized.
Writing a standard is an effort to put how something is done in clear, concise, and understandable language. It's a very slow process and I've yet to see any significant standard that doesn't end up with some parts being ambiguous regardless of how much effort is put into it. You rarely even see how parts are unclear until someone else tries implementing it.
There's bound to be some flaws in OOXML, doubtless there are flaws in ODM as well. However, at some point you decide it is as good as you can get it at that point, push it out, take a short break, and then you start on the next revision.
The only time the process ends is if people cease to care about that particular standard.
After reading this, I really hope you're lying about being on a standards committee. Unfortunately, I am equally convinced that you are in earnest. The problem is not Microsoft "opening up their own work", it is the fact that they say they are, but they really aren't.
The whole idea of a standard is that I could take a copy of it, sit down at my computer and implement it (whether that means implementing a client/server, or the parser for a file format). With Microsoft's OOXML, that has not been the case so far. The standard has been sprinkled with little nuggets that are closed, that Microsoft alone can implement. References to proprietary work that do not open up that self same work. Moreover, there are other sections that act as little more than XML wrappers around Microsoft's proprietary formats. In other words, I still couldn't parse out a full OOXML document without also reverse engineering a proprietary format. I want you to think for a moment and list any other standardized file formats that allow this kind of nonsense. If Microsoft wants to keep their formats proprietary, that is their own business, but a proprietary format flies in the face of the very idea of a standard (which is interchange; if interchange were unnecessary, standards would be as well) and should not be allowed to become one.
If Microsoft had introduced a truly open standard to the ISO and ECMA, the reaction would have been quite different. There would have been a lot of people who would have been wary of using it due to Microsoft's long standing policy of Embrace, Extend and Extinguish. Others would have considered it superfluous after ODF had already passed. But you would not have seen the outrage against this mock standard.
Yes. However, most large standards do also have some things that are not essential to the implementation of the standard that are in a vendor defined format. For example, the vendor specific sense data in the SCSI Primary Commands standard.
Actually just do a search in any of the SCSI or Fibre Channel standards docs for the term "vendor specific". You'll find a number of things that are not generally critical to implementing the standard that are done in a vendor specific way. They aren't critical, but leaving placeholders for them in the standard allows vendors to implement things that can add value for their customers, that other vendors might not care to implement, without breaking compatibility with the standard.
So there Sense Data types 0x80 to 0xFF are reserved for vendor specific implementations so that vendors can pass additional information about their SCSI target device back to a SCSI initiator without breaking with the standard.
Sometimes vendor specific represents things that different companies have implemented in different manners, and they don't see the need to standardize, and sometimes it represents company's proprietary implementation of features that they use to differentiate themselves in the market.
So what are the things that other companies are complaining about in the standard? Are they really critical parts of the standard, or are they minor things?
Every company has some things they don't open up in the standards process, they just can't be things that prevent the implementation of a compatible implementation within the bounds of what is common, and not listed as vendor specific.
Obviously different companies and organizations are going to disagree on what might be allowed to remain vendor specific, especially when you are dealing with some of the open software groups that basically oppose private ownership of intellectual property.
In reality, few companies want to just give away all the details of software they spent millions if not billions of dollars developing.
For OOXML to be a reasonable standard most companies should be able to save their documents in the format without using any vendor specific parts of the format, or the data in the vendor specific portions should not be critical to reconstructing the data.
I heard some rumors that some of the issues involved compatibility with previous versions of Office. Since this is a file format, I suspect that they are talking about including data or objects created in a previous version of office in a new file.
That's a difficult subject to address. Microsoft probably doesn't want to open up all their previous file formats as well as OOXML, but they do need to maintain backwards compatibility, and backwards compatibility is in customer's best interest.
However, allowing objects from a non-standard format to be included in the standard format, is something that open source people aren't going to like very much. It goes in opposition to the viral approach they use with GPL to try and open up all software.
However, it's not something that hasn't been done in many standards in the past, and it's also not something that consumers can't avoid if they wish to make their documents compatible with other company's office software.
To remain compatible, consumers would have to redo their old documents. By banning the use of objects from old formats, other companies are trying to force consumers to redo old documents.
That's not in the best interest of consumers, but it is in the best interest of companies that are producing competing products.
Not surprisingly, the companies that supported ODF opposed approving OOXML, while the newer members overwhelmingly voted to approve it. OOXML got majority support easily, they just had trouble getting a 2/3 super-majority.
Moreover, there are other sections that act as little more than XML wrappers around Microsoft's proprietary formats. In other words, I still couldn't parse out a full OOXML document without also reverse engineering a proprietary format.
Despite your naive to the contrary, the existence of such is not unusual in itself.
The question is what proprietary formats are they referencing. Is it for backwards compatibility, or are they features used in the current versions of Office as well? Are they crucial to the implementation, or can you ignore that data without losing significant features?
If Microsoft had introduced a truly open standard to the ISO and ECMA, the reaction would have been quite different.
Many of those opposing this would not be happy unless Microsoft released every bit of proprietary information related to every version of Office Microsoft ever developed. However, since some features of Office have been tightly integrated with Windows at least in some versions, they probably wouldn't be happy unless Microsoft completely released all information that has been proprietary in Windows as well.
Standards, only standardize the portions they say they standardize. They only open the formerly proprietary data that they document, which is almost always less than many in the "free software" crowd would like. They are going to be "outraged" about anything short of the end of intellectual property.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.