Issue 81 in teyjus: Propagation of disagreement pairs

[tey...@googlecode.com]

Status: Accepted
Owner: fafo...@gmail.com
Labels: Priority-Medium

New issue 81 by fafo...@gmail.com: Propagation of disagreement pairs
http://code.google.com/p/teyjus/issues/detail?id=81

The information that there exists some disagreements pairs is lost if the
logic variables for which there is a disagreement do not occur in the query.

For instance, given the following program:
p X Y :- X = Y X.
the query p X Y warns about disagreement pairs:
?- p X Y.

The answer substitution:
Y = Y
X = _T1

The remaining disagreement pairs list:
<Y _T1, _T1>


Whereas with the following code:
p :- X = Y X.

simply says Yes.


I think that there is no reason to forget those pairs. In particular, this
strongly assumes that every disagreement pair has a solution.
Even if this the case in theory, bugs already appeared in the past where
this was not the case. This kind of problem seem difficult to spot and
that's why I think all pairs should be propagated to the toplevel.


--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings