The Open XML Vote (OOXML) and why I hate politics

re: The Open XML Vote (OOXML) and why I hate politics

David Overton — Sun, 09 Sep 2007 12:25:16 GMT

Lee,

thanks as always for your candid comments. I have heard much bull from both sides, but I think it comes down to this:

1) People want to enable interoperability with Microsoft documents

2) MS could just publish it, but be able to change it at a whim, which would mean everyone would have to tow a MS line

3) Standards, IMHO, should never be user to block other standards - look at networks, we have many, many standards - not a problem. Look at TCP/IP - loads of changes to enable progress in the world. Now look at poor implementations - how many routers can handle larger packets, even though it is possible in the standard? Just because something becomes a standard you don't have to use it - it is just another option, so why do people fear a standard happening? In the network world people have released WiFi products before the standard was completed, in the sure knowledge that further tweaks will need to be made and the standard was ratified with that knowledge that as a standard it will also need further enhancements.

4) Finally, everyone should be held to the same bar - I think every criticism of Open XML could be levelled at ODF, except possibly the size one, but ODF does much less and has many undocumented details, so it would obviously be smaller.

I don't think number 4 is happening and that is what irks me. Since ODF is so underpowered vs Office, perhaps MS should take ODF, extend it with custom extensions so that all the office features can be accommodated and release that - I suspect that would not be desired either. Or perhaps we should just scrap a good chunk of what is good in Office so it can just write ODF style files. personally I go for the lets get Open XML made as a standard - just because it is a standard does not mean people have to use it, they simply get a choice - I like choice personally.

ok, another rant over with - thanks for the comments again - I'd rather understand if I am off base, in one corner, or making sense.

ttfn

David

re: The Open XML Vote (OOXML) and why I hate politics

lee_evans — Fri, 07 Sep 2007 10:29:29 GMT

Hi David,

I'm still on the fence a little on this one. I am involved with various groups amongst which are some who are vehemently situated in both camps and I've heard plenty of arguments - for, against, and at times completely irrelevant.

Whilst for my part at least the jury is still out, this is a well written post which raises some interesting points.

Lee

re: The Open XML Vote (OOXML) and why I hate politics

David Overton — Fri, 07 Sep 2007 06:31:28 GMT

Andrew,

I disagree with you in many way. You have to remember that the purpose of making it a standard is to enable people to ineroperate. If you look at most standards there is some form of ambiguity - look at ODF, it does NOT allow 100% implementations - if you are going to hold ISO standards to that bar, perhaps we should remove ODF as a standard.

I'm not an ISO expert, but I will give you some responses

1) Documents are complex - to discount them on size is very silly. My University project was consider too large by one of the markers and he refused to mark it. The University found someone else to mark it - it got a 1st. You have to remember that I used to write these interoperability tools - it is a million times easier to wade through the 6000 pages than it is to guess a file format.

2) I need to check some details here - if there is a need for more documentation or fixing of errors then I 100% agree they need work, however as I have said, the same is also true of ODF, so on the same tack this is also not worth of being an ODF standard. However, purity is not the requirement - who cares is some things are described with binary if that really is the right thing to do. As I say, I need to dig a bit deeper here to find out when and why. Remember, integration is yet, not XML pureness vs XML for 99.99% and documented binary for 0.01%.

3) Microsoft implemented Open XML in their products as was required - as it is going through the standard process changes have been made as required to make the standard more appropriate to more than just Microsoft - until these are finalised you can't expect Microsoft to constantly tweak the file format. One nice thing is that the standard includes version numbers so being able to deal with this is actually easier than you might suspect. Again, remember that I used to do this for a job. Then Open XML solution is much better than ANYONE elses documentation.

So, I disagree with you - it surpasses the bar of ODF by a long way, yet is held to a different set of standards and I suspect that once documentation errors are removed, people would be able to create tools that are 99.99% compatible. Have a look at the history of TCP/IP or HTTP and tell me that that V1 standards are held to a higher bar normally.

thanks

David

re: The Open XML Vote (OOXML) and why I hate politics

Andrew — Fri, 07 Sep 2007 03:19:31 GMT

David, I understand your frustration. You're an expert in these products and you want them to succeed.

But frankly, the euphemistically named "Open XML" was simply not qualified to be an ISO standard. The essential quality of an ISO standard is that it freely allows multiple 100%-compatible implementations of the standard.

OOXML is unqualified for three reasons:

(1) OOXML was way too huge for other people to read and understand, let alone implement: 6000 pages is not going to be implemented by anyone else soon. It's not going to be *read* by anyone else soon!

(2) OOXML was not fully documented, in three ways. First, the XML just wasn't fully documented, second, there were apparent errors in the documentation (not surprising with 6000+ pages, but unacceptable all the same), and third, anonymous binary blobs implementing specific functionality were allowed in the XML.

(3) Microsoft Office 2007 doesn't actually implement OOXML as submitted to ISO. This means anyone who actually managed to overcome (1) and (2) still wouldn't have a product that was 100% compatible with Office 2007.

Again, I sense your frustration, but an ISO standard needs to be *implementable*.

You and I disagree, so how about this. In addition to Microsoft, a couple of companies have promised to implement OOXML... Let's have a look around in a years time and see how many 100%-compatible implementations there are.

I wager it will be a big fat zero... not even Microsoft. :)