🚀 go-pugleaf

In article  <1992Apr13.180334.4861@newshost.lanl.gov> Jim Giles writes:
]In article <15195@optima.cs.arizona.edu>, gudeman@cs.arizona.edu (David Gudeman) writes:
]
]|> [...]                 What JLG describes is usually refered to
]|> as a "static property of an identifier".  The word "type" has several
]|> completely different definitions, but none that I know of refer to
]|> static properties in general.  Furthermore, it is only in low-level
]|> languages that type is an attribute of identifiers at all.  In
]|> high-level languages type is an attribute of a value.
]
]
]It is absolutely false and incorrect that my definition has no application
]to high-level languages.  In fact, I got this definition from the person
]who wrote the abstract semantics definition for Haskell.

What does Haskell have to do with high-level languages?

]It was the
]fact that he stated this particular insight into type (which I agreed
]with) when we disagreed on so much else that particularly impressed it
]into my mind.  His specific statement was, however, that the distinction
]between `type' and `value' is meaningless in languages in which neither 
]is static...

Then his specific statement was wrong.  Given a language with dynamic
types, the expressions "X = Y" and "type(X) = type(Y)" are entirely
different and may have entirely different results.  You might as well
take any function of one argument and claim that there is no
distinction between a value and the result of the function applied to
the value.
-- 
					David Gudeman
gudeman@cs.arizona.edu
noao!arizona!gudeman