I don't know if these are stupid questions or not.
1. Does a view require any disk space in the database? Is it stored?
2. If not, does a view have an impact on RAM usage, either on the
server, or on the client's machine (if running Management Studio)?
3. If RAM is significantly affected due to having a large view, does
it impact the speed of unrelated queries (not referencing to the
view)?
4. Is the main benefit of creating a view to abstract the complex SQL
from direct or repeated use (group it how you want it and just access
a single view as if it was a table with a simpler query)?
It seems like having a complex view that needs to be changed during
the development process is unnecessarily cumbersome when, for
instance, you need to modify the view in the database, then modify an
RDLC file that references it, and/or modify code that is looking for
specific columns to be returned from the view. However, I can see the
benefit of having that abstracted once the design is finalized.
JT