Re: Proposal: 6NF

Cimode wrote:
> dawn wrote:
>
>
> > Again, that depends on your definition of NULL. I don't understand how
> > you would not think it depends on your definition. Do you think these
> > letters fell from the sky with a meaning attached?
> Values are universally defined as output of functions and NO science
> does not fall from the sky...
>
> > Not convinced? I dare you to produce/define ANY function that certainly
> > produces NULL as an output. If you can't then just shut the hell
> > up...and stop confusing people...
>
> > I did produce one. If you would like to see it in action, you can use
> > the open source dbms OpenQM. Remember, this is NOT an SQL NULL.
> > OpenQM does not support SQL.
> Stating that f(NULL) = NULL is NOT a proof and F(x) = x NOT a precise
> example of a function...cos(x) is an example of function 2(x) + 3 is a
> function... Do you know what is a function?

I am quite sure that produced a valid function. There are three values in the domain of my function, "M", "F" and NULL (Yes, in this case it is a value and no, it is not an SQL NULL, but it is a NULL none-the-less). If you are looking for a function on an interval domain on the real numbers, then we could define a function g on the intervale [0,1] where

f(0) = "Male"
f(1) = "Female"
f(x) = NULL for all x in the interval (0,1)

This, too, is a function, even if not a function mapping reals to reals. NULL is not, of course, a real number, even if it is a legitimate output value for a function whose domain is the real numbers or a subset thereof.

> You did not demonstrate anything except that you don't understand the
> formal mathematical definition of a function...

If someone will confirm that you are correct in this, I will revisit what I have though to be the definition of a function. I suspect, however, that you are the one who needs to revisit the definition and verify that I have presented you now with two functions where NULL is a value in the range of the function.

> You consider as logical
> *proof* a specific example of implementation (open source bulshit)...

Nope, that was not the logical proof -- the function was. In case you were confused, I suggested you could give it a whirl in an open source dbms.

> > OK, I'm open to correction and would like to understand your point. In
> > practice, functions like this are common in Pick (even if not specified
> > like this). The point was to show that in fact a NULL (not an SQL
> > NULL) can be the output of a function.
> A *correction* is unsufficicent to compensate your ignorance or
> stubborness to ignore the formal definition of a function. You build a
> reasonning based on basically well ...nothing
>
> Any implementation that follows such flawed principles should be
> ignored...

Well that would be another discussion--whether 2VL is more useful for most "data processing" purposes or 3VL is, and whether the use of NULLs in 2VL is more useful than doing fancy footwork to avoid them. I understand that those using SQL with DBMS tools are rather "stuck" right now with 3VL, so there might be a reason for the OP's suggestion.  I would like to see the industry move forward into the past, ditching 3VL and going back to 2VL where we, as an industry, would have been well-served had we remained there.

> > What is an accurate generalization or abstraction from Pick? What are
> > the precise features that permit this example of a function with null
> > in the range of a function within the dbms environment? I thought the
> > 2VL was a key factor in null being a value rather than a non-value as
> > it is in SQL. Thanks. --dawn
> A confused mind...confusing others...

Sometimes confused, but in this case not so much. smiles. --dawn Received on Sun Oct 08 2006 - 00:18:39 CEST