Re: [boost] boost::nil

From: Douglas Gregor (gregod_at_[hidden])
Date: 2001-06-07 22:36:09

On Thursday 07 June 2001 09:21 pm, you wrote:
> A quick search for "nil" turns up one more use in
> boost::graph::connected_componets. And I think this is more of a case where
> the comment should read, (c_count initialized to "uninitialized"
> represented by max() )rather than set to "nil" which like Doug I tend to
> think of as an alternative value to "zero."
>
> But I could be persuaded that "nil" has a value other than zero. But for
> Doug's case, zero works best. (tuples don't care, it just needs a type to
> use as a marker.)
>
>   Yours,
>  -gary-
>
>
> gary.powell_at_[hidden]

The only reason for nil to have a value is for pointer conversions. Otherwise 
it can just be considered a symbol representing "no value" or "empty" 
(depending on context). 

If it is worth it, I can whip up some simple documentation and testcases.
Why does lambda use both "nil" and "const_nil"?

        Doug