One of the LinkedIn Groups I subscribe to is the "Informal Network of Standards Professionals". There have been a few interesting threads over there, including one about how standardization processes have changed -- or need to change.
In the case of information and software-related standards, the issue is simply that standards development methodologies haven't kept up with the pace of change in software development methodologies.
Software development methodologies have undergone revolutionary change in just the past decade. By comparison, the development methodologies of most SDOs and consortia have remained largely the same. SDOs and consortia tend to be the last great bastion for "death march" projects and "waterfall design". Under traditional standards development approaches, the balance of development time is spent in upfront design and drafting of specification documents. Key quality assurance processes, such as testing and the development of reference implementations tend to be activities that happen at the very end of the standards-approval project (if they occur at all).
What is needed?
- Realistic planning and estimation of what it will take to get a standard to a reasonable state of maturity.
- Some notion of iterative and test-driven development.
- Better delination of roles and responsibilities among a broader, more inclusive set of standards stakeholders.
- Better specification of requirements and interim testing and retrospectives as means to ensure requirements are effectively dealt with.
There are only a few consortia that have started to adapt their methodologies. David Carver of Starstandards.org has been an early and consistent champion for the application of agile methodologies to standards development. Making the transition to agile development is challenging within a single organization, so it is especially challenging within an organization of organizations. However, it is difficult to see how consortia can remain relevant and viable without some fundamental changes to development processes.
