I didn’t suggest that. I just noticed that Information Processing Standards also have an important history and, being in the digital domain, along with the kinds of documents and their formats that we are discussing, they are worthy of consideration along with razor blades.

I didn’t mean IT either, but I guess ISO does, huh? Go look at the title page you published.

When I look at your examples, I seems to me that a lot of standardization is customer driven. Customers, which do include retail and producers of other equipment, want standardization. In general, producers do not.

In a demand driven market, we tend to get standardization. In a supply driven market, we don’t.

The razor blade industry tries to keep the market supply-driven by constant change. If they ever slack down, standardization will raise its “ugly” head again.

Given that soft-drink companies don’t want to do the tool-making and distribution, I think they forced standardization onto the can-construction and filling machines and were forced to standardize to some extend by their distribution channel, eg, the vending machines and retail shops. I am not sure what part of the market vending machines take, but that could be an important factor.

This reminds me of a story about the introduction of ATMs in NYC (I believe).

The biggest bank tried to cash in on its market dominance by restricting their ATMs to their customers only. Hoping their larger number of teller machines would draw customers away from the smaller banks, who had less teller machines.

The other banks responded by opening up their teller machines to customers of all compliant banks. In one swoop, the competition had way more teller machines than the biggest bank and now it was losing customers to the smaller banks.

In one go, this seems to me to be the strategy behind ODF. It also illustrates why MS will fight to the end to kill off ODF. Like they did with Open Document.

It also shows a way for ODF to “win” (get accepted). If the world has more developers than MS, the world can construct more ODF applications and services than MS. Given that MS are hampering all efforts from outsiders to build innovative services and applications on top of Office documents, there should be a market here.

One simplistic suggestion from me.
Why are the books in project Gutenberg stored in unprintable formats?

If they were available in ODF, I could buy a portable book reader and have an instant library of classical literature. Or just print and bind them. It cannot be done in PDF because that would prevent text based processing of the books. It cannot be in MS Office format, because each of them have become obsolete in a few years. With Ecma376/OOXML we would have to wait until MS found the time and mitivation to give us the tools. With ODF, the Gutenberg Project could build the transformers themselves easily and we could have printing on demand of all the classics, and more.

Winter

But once again, you bring up engineering style arguments (e.g. saving aluminum costs, even the cupholder is something of a design criteria at this point).

You end your post with a whole laundry list of products that are not standardized, and some questions. Maybe it’s where your articles are headed anyway, but it seems to me that most of these decisions are made by companies trying to make money (and gain market advantage any way possible). Good design is very much an afterthought, and innovation happens by dumb luck (and I don’t mean how MS defines ‘innovation’). Once something approaches infrastructure status, then it’s time to standardize.

Something like razor blades are not an infrastructure item. However, Operating Systems and Word Processors are very much part of the infrastructure of computers. Hopefully we’ll get some sanity in that area in the coming decades.

Well, I’ll go back to lurking.

That leads me to think that the material in the ends of the can is more expensive than the walls of the can. But in any case, my impression is that can companies can only innovate within the narrow range that preserves compatibility with the installed base of cup holders embedded in cars and other places. That was my point.

Whether a uniform standard or proprietary design wins out over the long term depends on many details and competing interests. I find your blog very interesting, but I don’t think that you’re going to settle this question!

I think a document format will not be a very important choice in deciding what Office suite to use. As ISO wants to support standards with market usage it would be weird move letting it pass by.

Also several governments will want to have the MS format ratified by ISO as it gives some kind of control of Microsoft. They could put pressure on ISO which in turn can pressurize Ecma to make changes in the format. This is of particular interest to possibly the EU which has a lot of voting power within ISO.

The moment ODF got ISO ratification, it became illegal to refer to MS Office compatibility in government procurement.

That is why MS desperately needs ISO ratification of Ecma 376. If it fails, MS will be forced to support ODF as, eg, the EU members will not be allowed to buy or require MS Office anymore.

Wesley Parish said:
“Jason Matusow’s already given the farm away, with some recent statements that what is actually essential – for customers – is working and workable file format translators. I happen to agree with him – it just happens to make the Microsoft insistence on official standardization moot.”

Winter

For him to make customers’ needs the basic, essential necessity is indeed praiseworthy, considering the customary Microsoft attitude of what’s good for us is good for the customer, even if the customer doesn’t agree, because we know much more than the customer by definition … ;)

So, I declare the “file format war” over. It’s not really a “war” when one party comes up to you and begs, “Can you please say I won this time? Please?”

In any case, the important part is not demolishing Microsoft – that’s Microsoft’s privilege, and it’s one they fully believe in and pay Steve Ballmer handsomely to carry out – the important thing is making sure that Microsoft products are not the automatic purchase decision, because when that happens, Microsoft goes to sleep and we all suffer.

Wesley Parish

Think of Darwinian speciation, where geographical isolation will cause separate adaptions and genetic drift that lead to the creation of new species.

I think video standards are similar. The majority of users in the U.S. will go their entire lives without needed to deal with a video in another format. I wonder if this is different in Europe? For example, how are broadcast signals handled in the region bordering France and Germany?

In any case, nothing is final. If circumstances change, then a domain which once accepted multiple standards without complaints could suddenly find them unacceptable. Note the example of the creation of Standard Time prompted by the increase in rail traffic.

As for razors, I’m not saying that having multiple standards there is bad. I’m just observing that in this domain multiple standards have shown a longevity. In any case, I use an electric razor.

However, shouldn’t the broadcast TV standards be under the category of multiple standards? NTSC/PAL/SECAM and their numerous, semi-compatible variants do cause lots of problems (e.g. a film festival I was at where we could not show the feature film!) And HDTV is no panacea, as Europe and the US have mandated–rather stultifyingly–totally incompatible digital TV standards.

Also, Gillette’s proprietary razor attachments are less an example of why multiple standards are bad than the downsides of the patent system. (Do we really need patent protection to devise new razors? Are R&D costs for razors so high that we need to incentivize their development with limited monopoly rights?)

Hmmm….

Isn’t that a bit like trying to break through an open door?

Reducing complexity is the bread and butter of CIOs and enterprise IT architects. They all want fewer suppliers, fewer technologies to integrate and the lower TCO that result from lower complexity. Can you imagine half the market using ODF and the other half using OOXML and everybody being forced to use both standards for document exchanges? Having a single document format is a no brainer. It is the case for having _two_ document standards that is hard to make.

When Microsoft argue there is room for two standards, they actually mean that their format must be elevated to standard status. They don’t care about ODF. They believe the market forces (read their monopoly power) will make OOXML the unique standard that matters practically. What they want in the mean time is to make sure governments that mandate ISO approved standards don’t exclude OOXML and don’t force build an installed base for ODF. It is really that simple.

When Microsoft friends argue that ODF and OOXML are designed to meet different requirements, I always wonder what are the requirements that ODF meets that OOXML doesn’t. Could it be vendor independence?

But Microsoft usually doesn’t discuss the advantages of ODF. They prefer to paint OOXML as the standard that provides interoperability with the installed base, implying that ODF doesn’t. They are saying something every CIO in the world will understand. They are saying if the CIOs want to be compatible with their installed base of documents, applications and workflows, they need OOXML. This is the message Microsoft is sending loud and clear.

Microsoft counts on the CIO’s desire to have only one standard and remain compatible with the installed base to kick ODF out of the market. All this coexistence talk is a diversion to make ODF proponents look like hard line extremists and reduce the opposition to the ISO approval. The real plan is to have only one standard left standing. Everybody knows it and wants it that way, suppliers and customers alike.

There shouldn’t be any discussion on the number of document standards we need. The real question is why this standard should be dictated by a monopoly’s desire to retain control over the market.

Today the answer to this question is one of credibility. As long as CIOs don’t believe ODF proponents can displace Microsoft, this is a self fulfilling prophecy. CIOs will adopt OOXML because they believe everybody else will adopt OOXML, but everybody else adopt OOXML for the same reason. This is the catch-22 ODF must break.

The credibility can change overnight if some influential governments mandate ODF and exclude OOXML. Then application developers must support ODF to enter into this market. ODF will have credibility and the catch-22 is broken.

This is why Microsoft works so hard to get ISO approval. If they can make sure OOXML is not excluded, CIOs will be free to stick to OOXML and ignore ODF.

But this ISO approval is not designed to improve vendor independence. If you believe that, I have a bridge in Brooklyn to sell you. This ISO proposed standard is designed to give Microsoft control over the barriers of entries, making sure they are not excluded by government policies while they retain the ability to exclude everybody else on a legal and technology level.

I suspect that this is going to be true as you look at other areas as well. I look forward to the rest of this series.

I think taking a narrow technical view of documents, as just being an information processing standards, is to unappreciated the role documents play today. Who creates and consumes documents today? How are they created, exchanged and published? How has this changed in the last 25 years, or even the last 10 years?

This is not longer a concern merely for the guys in the back room with the mainframe. I’ll argue that point a bit more in Part III, including my thoughts on how historically there has always been but a single document format.

1. It would be interesting to talk about information processing standards, because that is significant in the case of document formats. The development of standards for digital codes has a great history, and LBJ’s Federal ASCII mandate had no direct impact (e.g., EBCDIC continued to be the mainframe code of choice), until the microcomputer and then the PC more-or-less settled the matter (so is EBCDIC dead?). Programming languages are also a great example (why isn’t PL/I the only one?).

2. There’s something off in your observation that “historically there has always been but a single document format.” I don’t think that has ever been true, especially around standards for formats and even with the ebb and flow of popular de facto formats. You might argue that the Word .doc is now a de facto standard, but I don’t think even that establishes your thesis.

I look forward to the continuation and the lively discussion that it may bring to the comment threads.