Gotcha! No Standards for Developing StandardsBy David F. Carr | Posted 2004-09-02 Email Print
Re-Thinking HR: What Every CIO Needs to Know About Tomorrow's Workforce
It's easy for software developers to trip up with nascent standardswhich aren't always fully flushed out. Here's how to avoid the pitfalls.The Common Picture Exchange Environment (CPXe) specification is intended to create opportunities for a world full of businesses, from Eastman Kodak to Fuji Photo Film to computer makers to the family camera store on Main Street. But as with many nascent standard-setting efforts, it also presents hurdles. Software developers, for instance, must work out the details of how each party will exchange and handle digital files. Each such effort presents new challenges to participants, by the very fact that it's trying to combine multiple approaches into a single standard.
PROBLEM: Early versions of the standard may not include key technical or business processes.
RESOLUTION: Add to the "hit list" of issues to resolve. For instance, CPXe 1.0 included the basic mechanism a digital photo-album software maker would need to add an Order Prints Online button to that product. But there was no standard way for photo software vendors and other intermediaries to record the number of transactions steered toward a particular print fulfillment service. Without this feature, there could be no accounting for distribution fees. And if these brokers of demand weren't able to get their cut, the business would not surface.
One of the first requirements the architects of CPXe 2.0 are tackling is a transaction logging standard.
PROBLEM: If you can't find services, they don't exist. One of CPXe's shortcomings is that it lacks an open, vendor-neutral and easy-to-search directory of photo printing services.
RESOLUTION: Set up a public directory of services. Through the International Imaging Industry Association, CPXe's backers have created the Picture Services Network (PSN) as a directory of print, editing and other services available nationwide.
Other CPXe directories have been set up by individual vendors such as Kodak. The existence of the PSN directory opens the possibility of a marketplace where photo printing services could compete for consumers' attention on a more level playing field. So far, however, PSN is barely more than an experiment, containing just a handful of listings.
CPXe's PSN specification incorporates Universal Discovery, Description, and Integration (UDDI), the established standard for directories of organizations offering Web services, and adds specific rules for listings of photo services. But by itself the UDDI directory is not enough, supporting only basic searches such as the countries in which a service provider does business. Like a code repository for software reuse, a UDDI directory is primarily for software developers looking for Web services to plug into their applications.
What CPXe needs now is for individual providers to create more elaborate directory services, called Service Locator Services, for consumer-oriented search functions such as finding the retail photofinishing locations in a given ZIP code. So far, the only such locator service available on the PSN directory is Kodak's, which directs people back to Kodak's own printing subsidiaries and business partners.
PROBLEM: Incompatibilities crop up as new versions of the standard are created.
RESOLUTION: Minimize potential problems by coming up with guidelines that narrow the possible interpretations of key pieces of the standard.
CPXe standards writers did this by adopting guidelines from the Web Services Interoperability Organization (WS-I), which specifically aims to weed out ambiguities in the specifications established by other groups and knit them into a more coherent whole. For example, the UDDI specifications allow two ways of specifying the Internet address of a service, but WS-I directs developers to use the one-the [ital]uddi:accessPoint[end ital] parameter-that is more consistent with the way such addresses are expressed in the Web Services Description Language, another key specification.
PROBLEM: Fear of flaws hinders adoption of the standard.
RESOLUTION: Identify shortcomings of the first standard and rapidly churn out revisions.
While some standards efforts collapse or slowly fade away, others just take time. Bluetooth, the short-range radio networking technology that's now gaining popularity as a means of wirelessly linking devices such as cell phones, handheld computers and headsets, has been generating hype since version 1.0 was released in 1999. But that first release turned out to leave too many important details to interpretation, particularly in authentication and encryption. That led to incompatibility between products from different vendors. Version 1.1 of Bluetooth addressed most compatibility issues, and version 1.2 in 2003 introduced "adaptive frequency hopping" to minimize radio interference from other devices. Fix fast, and fix until flaws are few and features abundant.