Ambler on the risks of traditional development methods.

http://www.ddj.com/architect/202101049?cid=Ambysoft.

The purpose of the article is to reset the debate on the risks of adopting agile methods; sure he says, Agile has risks, but define them against the risks of our current model and they don’t stand out so sharply.

The key to the discussion seems to centre around BRUF – big requirements up front and the costs associated with locking down your requirements early.

What of things change? What if your assumptions were wrong?  You are now committed.

What Ambler writes is all true, but there are limits to his argument.

The first is an assumption that the cost+risk associated with just-in-time design are lower than the costs associated with rework and I am just  not sold on this principle.

This article challenges the way analysis do business, and possibly the fundamental value of business analysts.  It’s written from a developer centric view.  Any comments?

Nice points.

The agile evangelists put a lot of people off.  The truth is tht there is room for several methoologies, and some suit particular cirumstances better than others.  Of course good requirements analysis and management are critical to success in any complex project.