Re: A new proof of the superiority of set oriented approaches: numerical/time serie linear interpolation

From: Bob Badour <bbadour_at_pei.sympatico.ca>
Date: Mon, 30 Apr 2007 15:52:37 GMT
Message-ID: <9hoZh.29498$PV3.315689@ursa-nb00s0.nbnet.nb.ca>

David Cressey wrote:
> "Cimode" <cimode_at_hotmail.com> wrote in message
> news:1177938789.949723.62480_at_h2g2000hsg.googlegroups.com...
>

>>On Apr 30, 1:56 pm, "Brian Selzer" <b..._at_selzer-software.com> wrote:
>>
>>>"Cimode" <cim..._at_hotmail.com> wrote in message
>>
>>[Snipped]
>>
>>>I'm not sure I understand you.  Missing information is a problem

>
> independent
>

>>>of any implementation.
>>
>>Sure but NULL is only one way to handle missing information and it is
>>one that does not fall into 2VL required by RM.

>
> Why does NULL not fit into 2VL? I've never propoerly understood this.
>
> In practice, I've been willing to use NULLS in databases rather than
> decompose tables to eliminate the need for NULLS. And I've tended to avoid
> SQL 3VL, by forumlating my queries carefully.
>
> But I've never fully studied the question. Once you accept NULLS as (one
> way) of dealing with missing data, are you then forced, willy nilly, into
> 3VL for your DML? It's not obvious to me, even though that's what SQL did.

One can use something like NULL with 2vl by using union types. But then NULL will be a value just like any other.

As soon as one uses a marker of any kind, one deviates from using values and 2vl. Even the 'Default' markers Date recommends/ed require special predicates. Received on Mon Apr 30 2007 - 10:52:37 CDT