Re: Shared game-data

>> That statement should have been more like: In general, normalization is a very important technique that allows efficient data mgmt; however, the time it takes to normalize data and retreive normalized data (relatively small amounts of data) reduces performance and as such may not be appropriate in apps where maximum performance is required.

> "efficient data mgmt" What is data mgmt and how does one measure its efficiency?

One place to find these answers is in Date's book. The meaning of efficiency might vary in context to specific apps/requirements. I don't spend a lot of time defining words. I am still waiting on c.d.t. to define data and fact.

> "time it takes to normalize data" What time? When is it spent?

The best way to find out is to code a database and count the cpu cycles.

>"time it takes to... retrieve normalized data" This is slower than... what?

Data that hasn't been normalized.

> "maximum performance" Do you mean minimum time?

For some apps, yes.

>> For example, in dbd, where performance is not a primary requirement and data mgmt is, all represented things are auto-normalized (no redundant representations). .
> Could you explain a bit more of this "dbd" and the differences between performance and data mgmt?

In context to dbd, it don't give a hoot about performance. In context to dbd, representing things such that there are no redundant representations, provides the foundation for systematic data mgmt.

> And I'd be interested in hearing more about how "things are auto-normalized."

Ancient chinese secret, seriously :) Received on Mon May 01 2006 - 01:43:23 CEST