Higher Standards for Web Standards
Since the emergence of web services in the 90s, we’ve seen an explosion of standards and standards bodies. Sometimes, they emerge based on new innovations, other times they’re created to unblock a stalemate on a similar standard or organization. Occasionally they are created simply to change the technology landscape in a way that is more favorable for certain vendors.
A question that I am asked over and over is – “Does Microsoft support standard X?” or “Is Microsoft going to join standards body Y?” The question we should spend more time debating is “What are the key technology or interoperability gaps and how do we fill them?” As new initiatives emerge, we research the business and technical need before taking action. We do this by putting ourselves in the shoes of actual developers and IT Pros and asking “what are the barriers I’m facing today, and what do I need to solve them?” Pragmatism over theory, always.
In many cases, the right answer isn’t necessarily to define something new, but to instead carefully consider whether technology or initiatives already exist to solve the problem. In the end, we should judge the strength of standards on industry and customer adoption alone. As an example, IBM recently announced a consortium called “WSTF”: Web Services Test Forum which leaves us a tad puzzled.
As of today, the WS-* standards are largely complete within W3C, OASIS, WS-I, DMTF, etc. and are widely implemented in infrastructure products and used by organizations all over the world. We were thrilled to participate in the Oasis announcement just last week on WS-RX, WS-TX and WS-SX. With regard to testing, we think it is critical that customers be able to propose scenarios that match their real-world interoperability needs. Equally important - both successes and failures must be made public. This is why we’re still evaluating our participation in WSTF.
Microsoft and other vendors have been participating in a variety of forums for quite some time to help crack the interoperability code. A few examples of forums that have yielded real world results for developers over the years are:
- WS-* specification development at W3C and OASIS. This formal process defined the protocol specifications for enabling service composition through addressable, secure, transacted, reliable, policy-based, end-to-end messaging.
- WS-I is the base layer process for integration and interoperability, upon which other, more domain-specific or scenario-specific tests, profiles, and guidance are built and the primary WS-* interoperability testing focus for Microsoft.
- Interoperability Plug-fests are more informal events at which multiple vendors get together to test interoperability against all other interested attendees, using agreed-upon scenarios for current and forthcoming products. The test tools that Microsoft developed remain available at https://131.107.72.15/endpoints/Default.aspx. These endpoints (and similar endpoints from other vendors) implement dozens of scenarios that customers and vendors can use to validate interoperability.
- Greg Leake runs one of the largest interoperability labs in the world and publishes results and guidance on WS* / WebSpehere / .NET interop. Stay tuned for more here – Greg is just completing his work on WebSphere 7.
Separately, but potentially equally interesting… An interoperability project that Microsoft recently joined is the Apache Stonehenge incubator effort. We look forward to expanding our efforts in partnership with other vendors on this front. Here’s the latest.
Do we need additional standards? The answer is almost certainly yes. But before touting a new standard or standards organization, vendors need to be clear about what specific issue is being solved and hold all parties accountable for doing so. Public access is a key criterion we have in mind as we think about WSTF. So what can you do? Continue to contribute at all levels; standards are only as good as the community formed around them. As always, let us know what you think.