Re: The naive test for equality

"Paul" <paul_at_test.com> wrote in message news:42fa5aeb$0$1215$ed2619ec_at_ptn-nntp-reader01.plus.net...
> vc wrote:
>> " The word "symbols" refers not only to the symbols used to exchange
>> data between people and computers, but also to each of the data items
>> inside the computer "
>>
>> To rephrase, you defined the word "symbol" as:
>>
>> 1. something used for human consumption, presumably a string of
>> characters on paper used to name the thing humans work with ;
>> 2. internal implementation of the apparently same things the computer
>> works with ;
>>
>> Now, you are giving another, even vaguer definition of "symbol",
>> namely, as "numbers managed by a computer". So, which one out of
>> three is it to be ?
>
> I think the word "symbol" has a very well-known English language
> meaning. We're talking here very informally; this isn't a mathematical
> journal so I don't think formal logical definitions are necessary here.

The group is called 'comp.databases.theory', not 'comp.databases.fiction', which presupposes a certain degree of precision in defining terms, not necessarily formally, although some degree of formality would help to avoid misunderstanding. Using words with double or triple meanings hardly helps in clarifying things.

>
> Sorry to stoop to dictionary definitions, but...:
>
> sym·bol n.
>
> # Something that represents something else by association, resemblance,
> or convention, especially a material object used to represent something
> invisible.
>
> # A printed or written sign used to represent an operation, element,
> quantity, quality, or relation, as in mathematics or music.
>
> I think it's clear what we're talking about here.

No, it's not. The first dictionary definition applies to stuff like paintings, photos, sketches, works of art, etc. The second defintion is used to refer to mathematical constants, signs, functions etc. None of those definitions is used to denote "numbers managed by a computer" as David Cressey claimed which is OK since he is fully entitled to invent any private definition at the risk of being misunderstood. So to be clear what one is talking here about, one has to provide an unambiguous definition/specialization of the term, not necesserily a formal definition.

E.g. a FOL language (of which the relational model language is a dialect) may define its vocabulary like this:

variables: x1, x2, ....
constants: a1, ... 123, ..., john, ...
function symbols: f1,...
predicate symbols: P1,....
connectives: &, |,
etc.
......

The FOL vocabulary is used to *name* things in order to be able to talk about them, so the second dictionary definition is correct if it the word "to represent" is a synonym of "to name".

>
>> What's interesting, whilst talking about symbols, representations and
>> such, you've forgot about the real thing, the value, which is of

>> primary interest for modelling, with the value's name and
>> implementation being important but secondary considerations.
>
> We're not talking about modelling here though; isn't the whole point of
> the thread about physical implementation of equality in a DBMS?

If we are talking about physical implementation, then what the word "symbol" has to do with it ? According to your dictionary search results, none of the definitions is applicable to physical implementation.

>
> Paul.
Received on Thu Aug 11 2005 - 00:34:36 CEST