Re: In an RDBMS, what does "Data" mean?

mAsterdam wrote:
> Anthony W. Youngman wrote:

>> x writes:

>>> Why you have not answered the question ?

>> And if we haven't got a philosophical definition, we can't compare the 
>> philosophical and theoretical definitions, and therefore we haven't 
>> got a clue as to whether either "the relational model mostly works", 
>> or (and this is important) where its limitations are and where it 
>> breaks down.

It mostly works, but we have some clues where it breaks too: metadata, use patterns...

> I won't answer the original question either (I'll just rephrase it),
> but I will share some thoughts about just what "data" means.
> Just a few associated concepts I have used to have some
> grasp of this - a semantical network, if you will.
> I have no sources or proofs, no famous
> philosofer to refer you to.
>
> The network roughly consists of: sign, media, shape and meaning.
>
> We have signs. They serve to communicate.
> Signs: A handshake, a hieroglyph, an ideogram (e.g. a chinese
> character), a sonogram (roman, arab character), a facial expression,
> a traffic light on red, an alarm - these are elementary, but
> I would also include: the collected works
> of <your favorite moviestar>

I think our aim is to model reality and entertain users by creating nice illusions or giving them competitive advantage by reducing entropy in their work environment or by predicting the future.

There are many realities, but let me mention two; the reality of the current IT with implemented infrastructure and the worldview of a modern intellectual. Our interpretation of what's implemented in our (heads) is what we try to model in our toys. And by what we learnt this is nothing like mechanical wheels of a watch nor computer's random access memory and not even the relational database. The problem is in compressing the representation of data and easing the recall of that data. Here it becomes useful to know what data is, but for the current state of the art that has unfortunately already been settled.

> In order to (just) exist all of these signs have media and shape,
> their pure existence does *not* require human (or just active)
> interpretation. Their function (purpose, ie
> communication), however *does* require some
> interpretation activity to assign meaning to them.

That's what you think and if I'm ever your customer, you won't model it that way :) Seriously, I don't believe in _pure_ shits or that anything exists without being observed/interpreted, but I'll not go deeper as it may look like off-topic religion bashing.

> This combination of sign and meaning we call data.

I'd say fixation of this on a media is called data, couse otherwise you can't recall it later. And there is a very important thing that folks miss: if you vanish and nobody knows the way you fixed that data there's just (series of ones and zeros) without meaning. Thus a fixation can't be generally called data without known way to interpret it.

Regards,
Karel Miklav Received on Tue May 18 2004 - 13:10:19 CEST