Simon Slavin wrote: > On 2 Jun 2010, at 1:14am, Darren Duncan wrote: > >> What are some examples of the proposed SQL replacements that do this? > > You might be interested in NoSQL, or in databases which have no schema: every > piece of information is a property of an object. Please note: I am not > recommending these systems, or saying that I use them; I'm just talking about > database theory.
Yes, I understand. And similarly, a relational database doesn't have to have a schema, where "schema" means a pre-defined set of relvars/tables with specific attributes/columns that restricts what data the relational database may hold, though it is typically a good idea to use a schema. With a sufficiently smart relational DBMS, users never have to explicitly define indexes and the DBMS can implicitly figure out for itself how to optimize the way the database is used or cached or indexed for best performance. If users ever explicitly define indexes, this is more of a hint to a DBMS to assist it figuring out how to best performance, rather than being necessary to have good performance at all. -- Darren Duncan _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users