Re: boolean datatype ... wtf?

On 06/10/2010 6:51 AM, Bob Badour wrote:
> Hugo Kornelis wrote:
>
>> On Wed, 6 Oct 2010 02:44:08 -0700 (PDT), Erwin wrote:
>>
>>> On 5 okt, 00:53, Hugo Kornelis <h..._at_perFact.REMOVETHIS.info.INVALID>
>>> wrote:
>>>
>>>> On Sat, 02 Oct 2010 10:42:50 -0300, Bob Badour wrote:
>>>>
>>>>> What makes either of those choices logical?
>>>>
>>>> Hi Bob,
>>>>
>>>> My first observation is that these three possibilities do not come from
>>>> a single domain. Male and female are values from the domain of genders;
>>>> unknown is not. So it would be wrong to define a domain with three
>>>> values to represent the three possibilities.
>>>>
>>>> But if unknown is not in the same domain as male and female, then what
>>>> is it?
>>>
>>> From domain boolean ?
>>
>> Eh? I guess I don't understand what you mean. Why bring up the domain
>> boolean at this point in my reasoning? Either I misunderstand what you
>> try to say, or you misunderstood what I tried to say.
>
> Hugo, you misunderstand the simplest of things.

I don't think it's so simple. I know people who think there is a third sex, maybe even a fourth, and other people who don't. They would all come up with different domains if they designed a db. Personally, in my own db, I'd just use {male, notmale}, but I wouldn't call those two values Boolean. Some people I know would be angry with me for choosing that domain but it wouldn't bother me, just as it wouldn't bother me if they chose a domain values of {female, notfemale}. If I assigned sex to a Boolean 'domain', my db could record only one of the sexes. Received on Wed Oct 06 2010 - 17:31:57 CEST