Earlier this week I linked to a post on the Engineering Windows 7 blog called "A View from the Bottom" by Larry Osterman, a Microsoft veteran whose own blog is a pretty good read if you are a development type. Osterman's post for the Engineering Windows 7 blog is all about organization, however, and it's a great read for CIOs or anyone overseeing software development projects.

Engineering 7 is a very authentic and unvarnished attempt to drum up some positive buzz for 7, started by Steven Sinofsky and Jon DeVaan in an effort I applaud. While the discipline surrounding the product development has been relatively good for Microsoft, with few leaks and little of the random and contradictory discussion of features which plagued Vista through its extended development process, in recent weeks various developers have been given guest star status on Engineering 7 and revealing exactly the sorts of things we want to hear about the product: less hideous User Access Control implementations and less invasive Indexing among them.

Osterman's entry gives us a perspective on how the dev team operated. I found it especially interesting as I am working my way through Microsoft Secrets, a circa 1995 analysis of Microsoft's management and development practices by Michael Cusumano and Richard Selby. Perhaps the first company to have to develop a systematic approach to massive consumer software product development, things haven't always been pretty inside the dev process, but I find the evolution fascinating. While revealing few specifics of the features involved, Osterman provides mountains of reassurance as to their overall quality and completeness by describing a development process designed to produce winners. Microsoft has been gradually implementing certain agile development methods company-wide for years now, but this may be the first major product developed there which has truly embraced the concept, rather than simply implementing certain practices and missing the overall goal of delivering working software.

I made a joke about Osterman's quote "cutting is shipping" but it's a great philosophy, concordant with the XP practices of refactoring mercilessly and delivering the minimum necessary. Other changes from the Vista model that Osterman describes, particularly in the planning and testing realms, also illustrate an organization that has taken agile development methods to heart. If you believe in the efficacy of such things, as I do, this allows a considerable amount of confidence in the product, even if you haven't seen it yet. Osterman's general tone and apparent high morale are other important factors boosting my general confidence in 7: developers don't lie well, and when they are slogging on a death march on a project that isn't going well, you can often tell. Osterman sounds stoked about how things are going and how they were done, and that speaks volumes about the likely quality of the product.