> There is the "CRUD" concern for server databases. That is, you wanted consistent access to creation,retrieval,updating,and deletion of data. Placing these functions in a data services tier comprised of SPs guarantees consistency and greatly reduces the chance of deadlocks.
How does writing SPs guarantee consistency? Each SP could be written differently, accept different parameters, etc. OTOH a view uses the same mechanism every time.
> Theer are other advantages as well: Stored procs are precomped, re-entrant, and if your data access bogs down, you have access to server management tools for both the database server's hardware and the server database.
I'm not arguing against using SPs or SPT, I'm just not willing to say that RVs should be completely removed from my bag of tricks, particularly without any concrete evidence that demonstrates tangible and significant performance or scalability benefits. I'm not saying this evidence doesn't exist either, I'd just like to see it.