Re: What is a database?

> On Thursday, 27 February 2014 07:40:05 UTC+11, Eric wrote:
> > On 2014-02-25, Derek Asirvadem <derek.asirvadem_at_gmail.com> wrote:
>
> > An ordered, structured collection of data.
>
> That could be good, depending on what you mean by "ordered". Or
> "structured", but that seems less ambiguous.

One answer, really, but in two levels, given the non-computer consideration, which I think is good to maintain. It is normal human logic, and whatever we do in IT should be within that, not outside that.

Qualifier. I do not believe in the concept of private definitions. That leads to insanity and creating a reality that is different from everyone else. This is IT and it is 2014, so the definitions those that exist in that context. Definitions from authorities are fine with me. Given the ever-changing garbage on wiki, definitions must exclude that. Hopefully that eliminates the ambiguity. The insane and the frauds introduce ambiguity and doubt, honest professionals destroy it.

> > Database: An ordered, structured collection of data.
>
> Notes:
> - Ordering and structuring is achieved via correct and formal Normalisation
> - Normalisation eliminates duplicates, therefore "unduplicated" need not be stated, but add that term if necessary
> - Ordering and structuring produces *Integration*, so that need not be stated, but add that term if necessary
> - The definition must exclude any application dependency (by virtue of Codd's definition to that effect). It is a data-base, not an application storage facility. Database=application may be common, but it is dead wrong.
> - Card filing systems; binders; spreadsheets qualify, as long *as they are ordered and structured*
> - 95% of the data storage implementations out there do not qualify, because they are unordered and unstructured; formal or correct Normalisation is absent.

1. General Ordered and structured according to the normal (specifically not abnormal or insane) logical (specifically not the arty variety, because this is a logical task) human (specifically not sub-human or damaged) mind.

We do not need to waste time evaluating propositions that, by definition, are excluded.

2. Computer Systems
Since Dr E F Codd wrote his seminal work, the Relational Model, in 1970, since it is transformed the entire database industry (we have had great platforms since 1984), since it is the de facto the standard for database design and the treatment of data, that should be the measure against which any implementation is evaluated. And any definitions relating to data should be taken from that.

Since there has been a mountain of garbage written by others who claim to be "mathematicians" and "experts" on the Relational Model, but in fact they are either grossly incompetent, or purposefully subversive, that mountain of garbage, non-relational because it contradicts the RM, is excluded.

Structured:

- As per the RM.
- That means Normalised
--- That means Normalised (a) according to the Normalisation given in the RM, and (b) Normalisation as a principle (which existed before the RM, and which the RM did not have to explain in detail).
--- That therefore excludes the abnormal, frgamented pseudo-scientific nonsense  marketed as "normal forms"

Ordered:
- As per the RM.
- Look up his reference to Order, and to Ordinals

> It is normal human logic, and whatever we do in IT should be within that, not outside that.

The deranged pseudo-scientists create "definitions" that are fragmented, isolated, located in outer space. They are irrelevant to undamaged humans, to the real world.

Cheers
Derek Received on Thu Feb 27 2014 - 02:44:55 CET