RE: [boost] 3-state boolean type

From: David Bergman (davidb_at_[hidden])
Date: 2002-08-04 22:32:33

The difference is actually not that subtle, especially in consideration
of the lattice theories, mentioned earlier in this thread.

    "unknown" is the bottom element
    "true_or_false" is an inner element, being "bigger" than both
"unkown" and smaller than "true" and "false", i.e., information has been
added to the "unknown".

In some semilattices, "true_or_false" would actually be considered to
not be less than "true" and "false", i.e., further information cannot be
added to the value (without forcing it to be "overdefined").

A similar distinction is encountered in unification as well as extended
type theories. Also note that the "true_or_false" here is quite
different from the "overdefined" mentioned earlier. That "overdefined"
would actually be "true_and_false" forcing the boolean structure into a
complete lattice, which is not appropriate...

I do not claim to know exactly what (semi-)lattice we would want here,
just wanted to enforce the view that "unknown" and "true_or_false"
indeed are different... maybe this is my clumsy way to say "Hi" as a new
member of this group.

Hi, and thanks for a great library!

/David

-----Original Message-----
From: boost-bounces_at_[hidden]
[mailto:boost-bounces_at_[hidden]] On Behalf Of Matthew Hurd
Sent: Sunday, August 04, 2002 10:52 PM
To: boost_at_[hidden]
Subject: Re: [boost] 3-state boolean type

In the GUI check box example it mean true_and_false.  Other times it may
mean unknown rather than true_or_false, there is a subtle difference.

matt.

----- Original Message -----
From: "Douglas Gregor" <gregod_at_[hidden]>
To: <boost_at_[hidden]>
Sent: Monday, August 05, 2002 12:03 PM
Subject: Re: [boost] 3-state boolean type

> On Sunday 04 August 2002 09:22 pm, Andrew J Bromage wrote:
> > The problem is that while tribools have a well-understood calculus,
> > they are suited to many varied interpretations.  Calling the third
> > value "unknown", "undefined", "maybe", "bottom" or whatever sounds
> > to me like calling your floating point type "length" or
"probability".
> >
> > Unfortunately, I don't think there is a better name, so I'm not
> > going to argue with whatever is chosen.
> >
> > Cheers,
> > Andrew Bromage
>
> Here's my last potential suggestion: "true_or_false". Otherwise, I'm
going
for
> "unknown" :)
>
> Doug
> _______________________________________________
> Unsubscribe & other changes:
http://lists.boost.org/mailman/listinfo.cgi/boost
>
>

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