When you create an integrator project, you ALWAYS do first steps over and over again: create connection, create extract, create field transform, create load, create job.
With a very simlar approcach to DDM you could limit this to a few clicks by selecting a connectoin, database, cube, maybe even some settings for the extract, select load of data and/or dimension, parallel/standard job, done. 15 minutes and a lot of annoying work saved.
Also in this way we could get people to streamlined usage of naming or other processes in the integrator projects.
In the and it is sth like the "Pate View" for integartor project.
