RE: [boost] Interval Library and comparison operators

From: David Bergman (davidb_at_[hidden])
Date: 2002-09-06 12:11:33

Doug,

Ok, I see the benefits with tribools in static analysis.

The question is whether this general library should incorporate
tribools. One might argue that other generic libraries should
incorporate tribools, to accommodate for specialized domains with
uncertainties.

If possible, I would prefer to have the general library not dealing with
tribools specifically, but understand that it might make application in
areas (such as static analysis, in your case) a bit more cumbersome.

/David

-----Original Message-----
From: boost-bounces_at_[hidden]
[mailto:boost-bounces_at_[hidden]] On Behalf Of Douglas Gregor
Sent: Friday, September 06, 2002 12:44 PM
To: boost_at_[hidden]
Subject: Re: [boost] Interval Library and comparison operators

On Friday 06 September 2002 12:19 pm, David Bergman wrote:
> You could also look at an earlier post by Joel, where he uses, as an 
> example, "exists"-definitions, which do the contrary, i.e., expand 
> instead of contract.

Wouldn't fit my domain.

> I do object in contracting '==' so that ' A == A ' becomes non-true. 
> IMEHO, that is contracting too far (or, in other words, being too
> cautious)

You can do a _little_ better if you specialize the definitions of x
RelOp y 
when &x == &y, then a == a, a <= a, a >= a, !(a < a), !(a > a) !(a !=
a).

> A naive question: what is the benefit of using tribools in this 
> interval domain?
>
> /David

I just posted a reply to Fernando describing the need for tribools in
static 
analysis.

        Doug
_______________________________________________
Unsubscribe & other changes:
http://lists.boost.org/mailman/listinfo.cgi/boost