Re: "thou shalt not conflate meta-data with data"

From: Neo <neo55592_at_hotmail.com>
Date: 3 Mar 2005 11:16:11 -0800
Message-ID: <1109877371.029641.6450_at_z14g2000cwz.googlegroups.com>


> 'John' is data (it's the name of a person),
> but John (the actual person) is a person, not data

Even the actual person John is data within a larger system, the universe. In a RMDB, the person is represented/modelled by a tuple in a table such as T_Person. I consider a tuple to be data. The person's name is represented by the value at the intersection of the tuple and the attribute Name in the relation's header. It seems your understanding of data includes values but not tuples (which is fine by me). According to your understanding, what is data (not examples of data)?

> Names are not male or female.

I don't think anyone in this discussion meant otherwise.

> It's the person John who is male, not the name 'John'.

I don't think anyone in this discussion meant otherwise.

> Hence, male is data about a person,
> not metadata (at least not in this context).

I don't think anyone in this discussion meant otherwise.

> Metadata is generally not about individual occurences, but about
collections (whether called sets, classes, domains, or anything else in the model of your choice). I wouldn't consider "the length of the name 'John'" to be metadata (though one might argue the opposite just as well). But "the maximum length of names for persons is 25 characters" is definitely metadata.

While the above possibly provides yet more examples of meta data, it evades the fundamental definition or providing a step-by-step method of determining if X is meta data. What is meta data (not what are examples of meta data)? Received on Thu Mar 03 2005 - 20:16:11 CET

Original text of this message