In terms of performance, I would say that in fact there is no negative effect. If the underlying request is inefficient and can cause unwanted blocking, the view inherits these problems, although it is somewhat mitigated by the fact that it is more likely to be cached than an arbitrary statement. I would say where there is a risk of negative effects, so that by creating a presentation and abstracting the complexity of the request, you risk using it without a clear understanding of what the characteristics of the basic request are. Then, on the positive side views, some various optimization possibilities (indexed views, etc.) also open up. And there are times when the ability to de-normalize the data set can significantly reduce the need to record redundant queries.
, , .