Re: boolean datatype ... wtf?

From: Hugo Kornelis <hugo_at_perFact.REMOVETHIS.info.INVALID>
Date: Thu, 07 Oct 2010 01:16:00 +0200
Message-ID: <lbvpa6pn6lab018627r3c3vd5j8ikap6r9_at_4ax.com>


On Wed, 06 Oct 2010 10:51:18 -0300, Bob Badour wrote:

>Hugo, you misunderstand the simplest of things. I added you to my killfile.

Either you debate with me (present your arguments AND read mine), or you don't. But killfiling me (i.e. refusing to hear my arguments) and then continuing to present your arguments is very similar to what my children did when they were six - yell their opinion, then stick fingers in their ears so that they can't hear any counterarguments.

>Null is an indicator that evaluates to NULL or NOT NULL. ie. it is a
>boolean.

How can Null ever evaluate to NOT NULL?
The predicates "SomeColumn IS NULL" and "SomeColumn IS NOT NULL" however do evaluate to a boolean. As do the predicates "SomeColumn = 40" or "SomeColumn < 5" - so this is not really a feature unique to nulls.

(snip)
>By "NULL hater", do you perhaps mean "intelligent and informed data
>manager" ? After all, the intelligent and informed try to avoid NULLs
>like the plague.

Call it however you like. Fact is that there are people who are convinced that NULLs should never be used in a database, and there are those who believe that NULLs are an adequate tool to handle the problem of missing information in a relational database. Rather than sticking my head in that wasps nest, I tried to stay out of it by giving both a design with a nullable column and an alternative without one.

When I used the term "NULL hater", I intended it as a short and jocular reference to the people who believe that NULLs should be avoided; I did not intend it to be derogatory. If you, or anyone else, felt offended by the term, please accept my apologies. I have avoided using this term in the replies I wrote today.

>>>>That's the logic of the first choice.
>>>
>>>I think you missed the point about 'logic'.
>>
>> Well, then I'm sure you can explain exactly when and where I missed the
>> point about logic, how you think I went wrong and what I should have
>> written.
>
>When you wrote "logical" you meant something else where logic had no
>bearing whatsoever. I am not sure whether the options you gave were the
>first 2 things that popped into your head or whether they best matched
>your limited preconceptions or whether you simply found them
>aesthetically pleasing; however, logic had nothing to do with your
>decision process.

Well, thanks for this fascinating insight in my mental processes. But I am sorry to have to disappoint you - your mindreading skills are even less than your social interaction skills, as you are completely wrong on how my decision process took place.

But I have never been too modest to admit that I am fallable, and I am always willing to learn. So on the question originally raised by Paul Mansour, how would you answer? And why is that answer better than mine?

>> For if you make only this remark without further arguments, it is just
>> an unfounded ad-hom attack that adds nothing to the discussion.
>
>I asked my question in the hope you would learn to distinguish logic,
>the foundation of data management, from internal bias or intuition. I
>agree with Erwin that you completely missed the point.

Obivously. I hope your answer to Pauls question will help me understand the points I am still missing. For so far, you are wasting lots of words to tell me how miserably I fail, but you don't spend half as much effort in explaining the better ways.

Best, Hugo Received on Thu Oct 07 2010 - 01:16:00 CEST

Original text of this message