Standards competition and globalization

I've been thinking a lot more about globalization and the important role standards have played in "version 3.0" of that process due to my reading of Friedman's book "The World is Flat."

In a globalized world, standards are critical, as they are what keeps the digital highways that link all the corners of the Earth open and accessible to all. The Internet, in other words, wouldn't be the Internet without HTML, TCP/IP and HTTP serving as the digital age's lingua franca. Any company with an intention of being an important player in this globalized world - a group that clearly includes Microsoft - needs to understand the importance of truly open standards.

If standards are so good, why aren't governments asked to decide which standard technology to use, thus avoiding the painful process wherein companies reinvent wheels in parallel silos as part of the competitive positioning that goes on in young markets?

That, however, doesn't work, because no person or group can have enough information to make such choices, irrespective of how "smart" the people involved claim to be. You can't declare unilaterally that a particular technology is the standard, nor can you hope to catch the wave early in its run and impose a standard technology ahead of wide adoption in hope of avoiding the fragmentation and competitive positioning stage. The fragmentation stage is important, as that is what helps to delineate the problem domain sufficiently so that potential standards have a decent chance of modeling a large enough swath of real-world usage patterns necessary to ensure staying power in a fast-changing technology marketplace.

That doesn't mean there aren't competitors that are well documented, or that have been ratified by official standards groups in their own right. SIP and RTSP can be considered competitors to HTTP, albeit ones with a slightly different functional orientation. That competition doesn't supplant HTTP status as a core protocol for the Internet, however, any more than Land Rovers (now owned, interestingly enough, by Tata Motors, an Indian company, making it a poster child for globalization) supplants the market for a Mini. Maybe at some point in the future those technologies MIGHT supplant HTTP (unlikely, in my opinion), but if it does, that is part of the process of continual improvement that is essential to the forward momentum of technological evolution.

That's why I oppose those who want to block OOXML standardization not on its technical merits (I have zero difficulty with those who want to quibble with technical details of the OOXML specification), but because ODF managed to cross the official standardization bar first.

Such people miss entirely how standards work in the first place. Nobody mandated that TCP/IP should become the global standard that linked all Internet nodes into a cohesive whole, or required that it be the "first" technical network standard ratified by a standardization body (others got there first). TCP/IP became the standard by general usage consensus, and became the ONE standard because it HAD to (TCP/IP is pretty fundamental as a technology).

Quibble with the technical aspects of OOXML, if you want (though as an aside, recognize that, just as SIP and RTSP have a different functional orientation than HTTP, OOXML has a different one as well: it aims for backwards compatibility with billions of existing documents, something even ODF proponents admit would be hard to convert, whereas ODF was designed to serve as a clean baseline for the future). Don't pretend you are defending standards, however, by closing off avenues for competing ones. In so doing, you are short-circuiting the process by which standards are developed and depriving them of the necessary competition that leads to long-term viability.

Microsoft recently released the specification for XAML, an XML-to-object mapping technology that is most closely identified with WPF. They also released a specification for a XAML vocabularly as applied to WPF, Microsoft's .NET-based user interface technology introduced with .NET 3.0 which serves as the baseline for Silverlight (Silverlight is a subset of WPF).

Some claim that this is Microsoft's attempt to displace HTML as a standard for the Internet, even going so far as make complaints to the EC about it. For a company aiming to displace HTML, Microsoft certainly provides a lot of support for the technology (ASP.NET, an AJAX library in the form of ATLAS, lots of HTML authoring tools, etc.). WPF, however, was clearly designed to make desktop development more attractive by applying the lessons of web site creation to desktop development (markup languages are great for UI layout). Further, they wanted to make desktop applications more network-aware, a detail that makes possible things like Silverlight, to be sure, but also enables users to run full WPF applications in a web browser (think: applications you can access over the web, but then drop onto your desktop for use while offline).

Does that make desktop application development more competitive with web development (and, hopefully, boost the success of the pure-browser variant of that technology, Silverlight)? Clearly, Microsoft hopes it does. That, however, isn't such a bad thing.

Anyone who tells you that AJAX-related technology constitute a clean and elegant development model is lying. HTML, CSS and Javascript were never designed to do all the work we try to make them do in modern web applications, and I say that as someone who has done cross-platform web development work since at least 1994.

AJAX applications are clear attempts to clone desktop applications in a web browser. WPF/Silverlight and its more popular contemporary, Flash, offer a competitive response to that in the form of a development environment that is a cross between desktop application development and web site creation, while offering features (3D animations) that are hard to replicate in pure AJAX.

Flash and Silverlights competition with HTML / CSS / Javascript (which is only partial, as both technologies are designed to be embedded within and controlled by traditional web applications written in HTML / CSS / Javascript) is part of the standards process. It is the process by which technologies get improved (HTML is not static in its definition).

It is, in a nutshell, the de facto standardization process.