Legacy systems and Agile Methodology

Hi
Here is my query.
One of the important charactersitics of Agile methodology is [B]SPEED[/B]. How does Agile handle projects that involve Legacy systems? What techniques does this Methodology adopt for these Legacy systems that involve huge volumes of data? How does Agile handle Relational data issues which may be time consuming?
Does your book address these topics?

Thanks.

I think the book authors have "left the building" but we can still talk about your questions. The biggest problem with legacy systems is a lack of tests. In fact some folks define legacy as anything lacking automated tests.

Michael Feathers' book Working Effectively with Legacy Code has a lot of neat ideas about how to bring this code under test so you can then do refactoring and add new features with confidence.

I have to say a mainframe team I was on in the 80s was pretty darned agile even without automated tests and formal iterations. We broke work down to very small jobs with real customer value, worked from a prioritized backlog, worked closely with the customer, delivered software and received customer feedback very quickly.

You mentioned huge amounts of data. How would that slow down coding and testing and other agile practices? Maybe we can come up with ways to get around it ... like testing only a small number of representative cases instead of a hundred-million-record master file.

BTW: Somebody will be around to talk about ranch name policies. You can go into your profile, select change name and read the policies to beat them to it.