> >By lowering maintenance costs via quality control, > managers can > >avoid (or at least postpone) having to scrap their > entire system > >and rebuild from scratch in India. > > I think this is a bit of a stretch.
I'm in the middle of one such situation. I have inherited a 500k line legacy system that has never been significantly refactored over its 10 year history and has zero tests. It is no longer safe or cost-effective to extend this system to meet future business requirements, so the entire thing is going to be replaced (although the work is being done in-house, not in India). Granted, a lot of other factors and decisions led us to this point. But if more effort had been put into quality (such as creating tests and applying best-practices) then the system might have been viable for several more years. Each situation is unique, but I think the general theory is reasonable. As a system matures, the cost of extending it to meet new business requirements (I'm calling that "maintenance") inevitably gets higher and higher. Eventually, you reach a point when it is cheaper to just replace it. I tend to think that from-scratch projects are the types of things that frequently get outsourced (I could be wrong about that). So if you believe that better quality leads to lower maintenance costs, then improving software quality extends the life of the product, which help the customer get more return on their investment and can help you (the developer) keep your job. Is that still too much of a stretch? Should I forget about the outsourcing stuff and just focus on the ROI aspects? -Jeff __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
