In the last 10 years the Aspect-Oriented Software Development (AOSD) has gradually become a concern stone in Software Engineering as an engine to reduce complexity and increase reuse by providing modularization of concerns that tend to crosscut. Nevertheless, its use in certain situations can presents some problems that can not only discourage its mainstream adoption, but also hinder the realization of software quality goals. The first problem, the AOSD-Evolution paradox, encompasses the difficulties with evolving software developed using AOSD. The second arises as a result of the invasive nature of aspects. The use of aspects without any control can result in a harmful practice. This work describes these problems and exposes the strength and limitations of the current approaches to solve them. Thus allowing us to reason in a clear fashion about the problems and their solutions, then justifying a contract base approach, which aims to control the usage of aspect without constraining the power of AOSD.