Nearly right. It's rather the case that when x is null we have
x - x
null - null
Hence it is not the case that x-x=0 if x may be null. Similarly, the law of the excluded middleIt's true, it doesn't contribute anything to this discussion, just like the Gauss law of normal distribution doesn't contribute anything to this discussion, but it is perfectly valid. That's all. I did't bring that law in this discussion. Andrew Max did. I've just said - yes, that's it, and shown that it's perfectly valid law in our 3VL world staying at the logical level of releational theory where NULL means that a value is missing, not that "proposition can be neither true nor false" (fuzzy-logic).
does not hold for values that may be null -- apart from the operators IS NULL and IS NOT NULL.
From this we conclude that the law of the excluded middle doesn't contribute anything
meaningful to a discussion of the implementation details of NULL in one RDBMS or another.
SQL> CREATE TABLE fuzzy (Believe me, there are many "fools" on the earth, who would argue that (X - X), when X is defined as NUMBER, must always be equal to 0 (zero), regardless of whether X is NULL (UNKNOWN) or NOT NULL (KNOWN).
2 X NUMBER NULL
SQL> INSERT INTO fuzzy (X) VALUES (NULL);
1 row created.
SQL> SET NULL NULL
SQL> SELECT X FROM fuzzy;
SQL> SELECT (X - X) FROM fuzzy;
I think Albert doesn't understand the state of his own mind let alone the state of data.You said:
What is domain of attribute (column) "AGE" in an employees table?
I'd say 18-65 ages. If a scalar is UNKNOWN, is it "totally unknown" or
just "partially unknown"? I'd say partially, since we exactly know that
250 years old man cannot be an employee in our company.
I'd say that (x - x) is always zero regardless of whether x is KNOWN or UNKNOWNSuppose x is ∞, Would you say infinity(∞) - ∞ is zero?
of course, in levels of their intelligence quotients (IQs). However, the
perilous difference between them, with fatal consequences, is in the fact
that Mr/Mrs "IQ" DOES KNOW he/she is the "IQ", whereas Mr/Mrs "non-IQ" DOES
NOT KNOW he/she is the "non-IQ". It's by definition.
I don't want to suggest anything here.
My "special thanks" go to those who were constantly obstructing anyThat's like saying "I'm going to take my subjective view (or should that be judgement) of people and state it as fact. Because I'm right!"
decent discussion by posting only petties and nonsenses.
Whenever they said the truth, it was something trivial; whenever they
said something that was not trival, it was wrong
Wouldn't your time be better spent finding a databaseThe issue is not about being capable of understanding something. There is null to understand. It is about believing weather something is implemented the best way. You seem to like throwing that one at people.
you are capable of understanding and learning to use
without whining and leave Oracle alone?
I've just said - yes, that's it, and shown thatYou didn't show anything. You just randomly claimed it. There is a difference.
it's perfectly valid law in our 3VL world
We didn't talk here about 3VL in the context of3VL has nothing to do with fuzzy logic.
fuzzy-logic: "a proposition is neither true nor false
at the same time, but may have a degree of truth
between 0 and 1 (eg 0.1, 0.6 etc), so truth values
are possible: 1) true, 2) false, 3) undecided degree
Actually it was me that said that but our "friend" is not very good at quoting things so it's obvious someone else said it.is determined by business logic that is applied at aThe domain of reasonable values is something that
level above the database level.
Aside from which, this is completely incorrect. The
correct place to restrict the domain of a type is in
Which would just be unrealistic to implement.
CREATE TABLE mytable (person_id sequence, name name_domain, age age_domain, gender sex_domain . . )