Of disk swapping, database structure & project management..

[Chris Gray]

:Has anyone experimented with swapping unused parts of the mud database to
:disk, reading it back in when needed? What sort of format is it stored in
:(eg binary, ASCII, etc)?

Several of the systems represented here are disk based, i.e. information
moves back and forth between memory and disk. In mine, there are two
files used for the database. The index file contains a few header values,
then is simply an array of descriptors, which grows as needed. Each
descriptor describes a contiguous space in the main data file, either
in-use for something, or available for use. All MUD entities are stored
as a set of contiguous bytes in the main file, and are represented by
their 'db-id', which is just the number of the descriptor that points
at them. The data can move to a new slot, without the db-id changing.

In-memory, there is a cache of some subset of the on-disk database, store
in a somewhat similar fashion. The cache is mostly flushed on an LRU basis.

--
Chris Gray   cg at ami-cg.GraySage.Edmonton.AB.CA