Re: Impossible Database Design?

-CELKO- wrote:
> >> Are you referring to 'Databases, Types and the Relational Model' by C.
> J. Date, Hugh Darwen? <<
>
> No, the book Bob refers to is from Morgan-Kaufmann and has gone out of
> print.

Amazon can ship it within 24 hours. You can get used ones too.

> It is not really very useful for a working programmer, since
> Date invented a language that not implemented for the book. It also
> uses a "Chronons" (discrete points of time) model for temporal data,
> which is rejected by virtually all of the temporal database academics
> because it fails to model time as a continuum.
>

Actually as a working programmer I found the work of Lorentzos, Date et al useful, yes it is theoretical but by knowing the limits of the RM I was able to deal with the situation better and more quickly.

As for the "rejected by virtually all academics" remark I can't accept it unless you have researched the topic? All the academics I worked with held Lorentzos'/Date/Darwen's work as the shinning light. Maybe I was just sucked into a cult ;-)

As for the continuum of time, I'm only aware of the physics definitions which are based on discreet sequences of events, like tick tock or so many cesium atom vibrations. I'm interested to know who uses a continuum?

As for contraversy, of course, I've never encountered one theory of man which is free from contention or opposing views. Especially in relation to time we've been debating it in philosophy, religion, physics, math ... well dare I say, since the beginning of time (whatever that means).

> A better and cheaper cholice would be to download the PDF book by Rick
> Snodgrass at the University of Arizona website. Rick has been doing
> temporal RDBMS work for over 20+ years and has actual SQL code in the
> major dialects in his book, as well as accepted theoretical basis.

Dialects of SQL don't do a lot for me. Especially since the problems are at the model level. I can't imagine any theory of time, discreet or continuum based, that does not require ordering, sets are unordered.

But thanks for the reference I'll check it out. Cheers ;-) Received on Thu May 18 2006 - 06:06:53 CEST