Databasen - NULL dates

Nigel Rantor wiggly at wiggly.org
Wed Oct 18 15:36:25 BST 2006


Paul Golds wrote:
> On 10/18/06, Nigel Rantor <wiggly at wiggly.org> wrote:
>> Well, let's say we have:
>>
>> 1) As yet unknown, will possibly be known in the future (date of death,
>> employee resgination)
>>
>> 2) Unknowable (end of the universe)
>>
>> If something falls into category 2 then why model it at all? It is a
>> constant.
>>
>> If we're going to start talking about things that depend on company
>> policy then you cannot try to model all things that a policy states we
>> cannot know because the list would be endless.
> 
> Things can fall into the two categories in the same column, though.

[perfectly reasonable requirements snipped]

Yes, I agree, there are situations where you may need to know what you 
want the information for, whether or not you want to try and gather it, 
whether you tried and failed.

This data would have to be stored as well.

   n



More information about the london.pm mailing list