a testcase where we had existentials in the conclusion

Dan, while testing the case mentioned in
http://lists.w3.org/Archives/Public/www-archive/2003Mar/0001.html
i.e. wether

:a :b :c.
{?x :b ?y. ?u = [:sf (?x ?y)]} => {?x :k ?u}.
{?x :b ?y. ?u = [:sf (?x ?y)]} => {?u :m ?y}.

entails

:a :k _:v.
_:v :m :c.

I found that cwm wasn't --think'ing that

:a :k [...


I then added (the tautological)

[:sf (:a :c)] = [:sf (:a :c)].  ## needed for cwm

and then it was OK
    :a     :b :c;
         :k  [
             :m :c;
             :sf  (
            :a
            :c  );
             =  [
                 :sf  (
                :a
                :c  ) ] ] .


(maybe I'm working in the wrong direction
but I'm desperately looking to get rid of
my messing with functional terms but keep
the benefit of single triple conclusions)


-- ,
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/

PS tested with a previous cwm as the latest cwm on my laptop gives
Traceback (most recent call last):
  File "/w3ccvs/WWW/2000/10/swap/cwm.py", line 58, in ?
    import LX.engine.llynInterface
  File "/w3ccvs/WWW/2000/10/swap\LX\engine\llynInterface.py", line 18, in ?
    import LX.uri
ImportError: No module named uri