>Business requirements should be what drives I/T. Unless an I/T solution is directly addressing a specific business challenge, it should not be undertaken. Also, ideal I/T solution specifications are generated independent of a specific technology.
>
>The sooner technology folks come to understand this reality, the incidence failed/death march projects will diminish.

My experience has been that failed projects are more often the result of non-technology folks making business decisions that a pre-chosen technology won't support. If an IT manager doesn't choose the right technology to meet the business requirement, is that manager making a good business decision? I say no. I would add that there is no ideal IT solution for anything. All you can do is come up with a good specification with reasonable expectations and a plan to get it done on time and within budget. Business folks can't do this by themselves. I've seen them scrape their faces on the road just about every time a Senior developer was not consulted before deciding on how to best meet "business requirements".

I just left a place last year where the vice president in charge of IT technology decided to move the next framework to Java and the entire shop of 85 developers were hired to code in VFP.
Because he read in a magazine that Java was "The Best". Two years later we had no framework.

To think that IT solution specs should ideally be developed independent of a specific technology is fool hardy at best. Sounds to me like a line from a university text book.