Re: Separation of Relational Engine from Domain Math

topmind schreef:

> This is based on a question I raised on c2.com:
>
> The issue of seperation of relational from the "domain expression
> engine" recently came up in a debate where the somebody claimed that
> finding the nearest N points to a given point in a geographic map
> system would have to traverse every point under relational, and thus be
> O(n). A solution is to move such operations to the domain language
> instead of let the query language directly do it.

Another solution would be to let the optimizer recognize when such questions are formulated using the usual operations.

> I have not seen much info about how and whether the "expression engine"
> can or should be considered orthogonal to the "relational engine", as
> long as a minimum set of requirements are met.

Depends on whether you believe in the nested relational model or not. If you do, and want to allow things like NEST and UNNEST operations, then such a separation doesn't make much sense anyway because there will be too much interaction between the system-defined and user-defined functions.

• Jan Hidders