Re: the relational model of data objects *and* program objects

erk wrote:
> Frank_Hamersley wrote:
><snip>
> Temporal aspects of data are important, no?

Very.

> Then why are such treated differently than "non-temporal" data?

I guess because the current crop of main stream DBMS packages do not provide an integral temporal nature. You can construct it of course.

IMO there is no such thing (whoop whoop whoop - dogma alert!) as non- temporal data. Of course you can choose to discard the temporal nature if it is not relevant to you interests.

> <snip>
> If you think "theory" is somehow inhibiting any of the above, you're
> wrong - in fact, those you'd probably poo-poo (Date and Pascal, to name
> to) repeatedly state how much more DBMS vendors could and should do to
> make the lives of developers and users easier.

No, I am not lambasting the theorists about their theoretical postulates. What I was alluding to was ill-considered application of a theory to a problem that doesn't need it. Inappropriate use is just as bad IMO a sin as an oversight caused by ignorance.

> Instead, the DBMS
> industry spews products that make it only slightly easier than before
> for its customers to do what the vendor should be offering as part of
> the engine.
>
The cynic in me says that is either a vendor eking out the life cycle or simply the incremental capacity we have to absorb change attenuating our appetite...for the latter just look at Neo's struggle to garner interest not withstanding the missing kilobuck! BTW that is not an endorsement of xDb at all!

Cheers, Frank. Received on Thu Apr 14 2005 - 17:34:56 CEST