Re: A Normalization Question

> Since you supplant all instances of a specific encoding with references to a
> single character value, wouldn't the references, which are pieces of
> information themselves, be redundant?

Absolutely not, while at the logical level each reference refers to the same thing, the value of references is a physical issue. On some hardware (ie PCs), references to the same thing usually do have the same values. On other hardware (ie brain?), references to the same thing are likely different paths to the same neuron(s) and don't have duplicate "values".

> Thing Person Color Street
> 1 brown brown brown
>
> in Neo's database:
> Person(1,2,3,4,5)
> Color (1,2,3,4,5)
> Street (1,2,3,4,5) 1,2,3,4,5 are repeated.
>
> Value Storage Address (16 bits) Value (8 bits)
> 1 'B'
> 2 'R'
> 3 'O'
> 4 'W'
> 5 'N'

Above is wrong. Below is approximately how XDb1 correctly normalizes the same tuple. Note, the -># is simply a way to express the reference and not indicative of the actual value (if any) at the hardware level. If a web page allowed me to easily draw lines from one point to another, the below would only have ---->'s instead on ->#'s, thus depriving you of your argument about hardware value of refs being the same.

Thing Person Color Street
1 ->2 ->3 ->4

Person Name
2 ->5

Color Name
3 ->5

Street Name
4 ->5

String Sym1 Sym2 Sym3 Sym4 Sym5 ....
5 ->6, ->7, ->8, ->9, ->10

Symbol

6      b
7      r
8      o
9      w
10     n

> If you really want to pursue this form of generalization,
> and you are really honest about finding the purest and most
> generalized form, then you better start with the bit and make your
> references to single bit values from there.

If one really wants to pursue generalization then one best not start with any particular hardware (ie bits). My data model can be implemented on hardware with or without bits, bytes or integers; and yet these things can be represented at the logical level (as does the human brain). Received on Tue Jul 13 2004 - 00:01:38 CEST