Re: [scrumdevelopment] IBM's Trilogy Program
> It reads like a textbook of all the dysfunctions agile tries toBig surprise, hmm ?
> address, however in the end it would appear that the concensus among
> those investigating is that the problems could have been avoided if
> they had simply just got the requirements right at the beginning
A book I read a while ago made a big impression on me and is quite
relevant to the method of investigating such large, complex projects:
"Aramis or the love of technology", by Bruno Latour. I recommend it
If you're investigating a large, complex, costly, failed project and
looking for a simple, tidy explanation of why it failed - chances are
you won't find one. If you do find one, I'd tend to suspect you
filtered and tailored the facts to fit your favorite simple, tidy
Large, complex projects are risky because they are large and complex.
So, no simple tidy explanation will do - whether "They should have
done their BDUF properly" or, for that matter, "They should have used
Instead, I think one learns most by gleaning lots of tiny insights,
lessons and conclusions from post-portems of this kind of project,
about various aspects of it (technical, human, political, etc.), and
how this body of small explanations relates together. (Latour's book
does just that.)
If a program is useless, it will have to be documented.