Re: Date, Darwen, Pascal and the alternative to Nulls in the RM

From: mAsterdam <mAsterdam_at_vrijdag.org>
Date: Thu, 23 Mar 2006 19:33:16 +0100
Message-ID: <4422e9df$0$11062$e4fe514c@news.xs4all.nl>

Roy Hann wrote:
> mAsterdam wrote:
>

>>It doesn't make sense to store [soc number, age].

>
> Actually it equally might or might not. What is a sentence into which these
> two values alone can be inserted to make a meaningful statement about the
> enterprise of interest (to "make sense")? If the table is dead_employees
> and the sentence is "the employee with social insurance number <soc number>
> was <age> years old when they died", then it makes sense.

I 'ld have to assume flaky column naming practise here. That doesn't go well with making sense.

> One might not be
> able to conclude very much else without more information, but whether that
> is a problem or not depends on the enterprise of interest. On the other
> hand it is enough to answer questions like: how old was such-and-such an
> employee when they died? Is so-and-so dead? What is the average age of
> death among employees? How many employees have died. And so on; all pretty
> sensible derivations from the limited facts.
>
>

>>You know it, I know it. Why doesn't it?

>
>
> Your premise is false.

How do you know? Received on Thu Mar 23 2006 - 12:33:16 CST