What is Invariant Good For?

[Walter Bright]

Manfred_Nowak wrote:
> Walter Bright wrote:
> 
>> I might as well have used [...]
> 
> A problem is, that your article refers to "existing terminology 
> confusion"---then more of the same should be avoided.

The thing is, nobody has actually been confused about the meaning of 
invariant as applied to data, as opposed to class invariants. In 
contrast, what 'const' means when applied to data has plenty of existing 
confusion.