Re: Declaration design preference: type or name first

Richard Harter wrote:
> On 1 Nov 2007 14:50:16 GMT, ram@[EMAIL PROTECTED]
 (Stefan Ram)
> wrote:
> 
>> cri@[EMAIL PROTECTED]
 (Richard Harter) writes:
>>> the code.  In natural languages a symbol (word) has the same
>>> definition(s) everywhere.  This is one reason why natural
>>  In English, the meaning of a name (like »James«) depends
>>  on the context, too.
>>
>>  Even terms have different meanings, for example, an
>>  »object« in Java (JLS3) is different from an »object«
>>  in C++ (ISO/IEC 14882:2003(E)).
>>
>>  Many other words do not have any clear meaning at all,
>>  but are still used widely.

> Did you see the "(s)" after the word "definition"?  

Sure.  If you want to stick the subset of natural language
where people only use words as defined in the dictionary, that
works.  But in the real world, people can and do give new
definitions to words based on context.  This is in fact, how
the definitions got into the dictionary in the first place:
words are constantly being formed and being evolved and being
given new meanings in certain contexts, and when the context
becomes large and a few other things happen, it is sometimes
accepted and canonized into the official dictionary or accepted
and canonized into the popular (global) lexicon.

For example, take the word "chip".  It was at first just a
piece of something, like a wood chip.  Then someone invented
potato chips and eventually the word "chip" became short for
that.  Then integrated circuits and someone called them "chip"
and that definition got adopted too.  You can bet that only
a small number of people used "chip" to mean integrated circuit
at first, and only in a certain context.  (Probably a cor****ate
context; I don't know the exact history.)  During that time,
"chip" meant "potato chip" or the original "chip" (as in wood
chip) meaning outside that context, and it meant IC within that
context.  Eventually, the IC meaning got accepted into the
industry scope, and now it's in the formal dictionary scope
as its own sense of the word "chip".

In reality, natural language is a bit more complicated because
there are overlapping contexts and pragmatic concerns are used
to disambiguate things.  But the point is there are definitely
local lexicons.  That's why we have the jargon file, after all.

   - Logan