Posted Mar 20, 2005 19:27 UTC (Sun) by leandro (guest, #1460)
[Link]
> Why does it need to be in the kernel?
It depends on what's one trying to accomplish. My own view is that we should use something like the Hurd, a multi-server microkernel, and then we could have only basic relational technology in the kernel so that we could have a consistent, relational view of data everywhere; and then features for database applications and the like could be userland. But for now we're stuck with monolithic kernels.
> what in the world is "D compliant"
Like in Date and Darwen's Tutorial D, D being the definition of a relational data access language.