[Termtools] competition

[Zantema, H.]

Dear all,
 
I see your problem. I agree with giving Salvador the opportunity to withdraw: I can imagine that he prefers that over the situation where every one in the world can see him being disqualified. I would agree with this for every disqualified tool (formally or not), just like for a conference rejected papers are not made public.
 
I do not feel comfortable with the present situation where Claude introduced disqualifying colors for his own tool......
 
    And, finally, a meta-rule:

   4) the rules of the competition are decided during a meeting at the
       end of the previous competition. This includes new categories or
       syntax extensions.

I think this is too strict. If two months before the competition the idea for a new category comes up, this should be possible. On the other hand, it is a good idea to prepare a proposal for syntax adjustments and rule modifications for the meeting at WST.
 
Best regards, Hans Zantema
 
PS This was the first time that I wrote an email in my garden.

	-----Oorspronkelijk bericht----- 
	Van: termtools-bounces at lists.lri.fr namens Prof. Geser 
	Verzonden: ma 19-6-2006 18:02 
	Aan: termtools at lri.fr 
	CC: 
	Onderwerp: Re: [Termtools] competition
	
	

	Dear all,
	
	I am strictly against applying rules that have not been stated when
	the competition started. There is a principle in law: you cannot be
	convicted for your deed perpetrated when there was no law for
	it. According to our current competition rules, Mu-Term cannot be
	disqualified.
	
	I suggest to give Salvador the chance to withdraw Mu-Term from this
	competition. Of course I am looking forward to see Mu-Term competing
	next time.
	
	We can, however, and we should, change the competition rules for next
	time.  Here are a few suggestions for future rules:
	
	1) a tool author may choose to withdraw his/her tool. Then the results
	will not be shown.
	
	2) if anyone can demonstrate that a tool yields, under competition
	conditions, a false answer for some system then that tool will be
	disqualified. The example system need not be in the database.
	
	A false answer also includes a false justification of an otherwise
	correct answer.
	
	Point 1 gives the authors the possibility to withdraw when they
	suspect that their tool is not reliable. Point 2 extends the
	disqualification to false answers outside the database.
	
	 > I know at least one other buggy tool:
	 > matchbox/satelite gives a wrong answer for
	 > (VAR x)(STRATEGY INNERMOST)(RULES  a -> x)
	 > because the *-bound computer does not handle
	 > new-born variables correctly. For standard rewriting this bug
	 > is hidden because the loop checker finds a loop earlier.
	 > But the loop checker is turned off with strategies.
	 > (I subscribe to Claude's viewpoint that this really
	 > *is* a rewriting system and this really *is* a bug.)
	
	Wait a minute. What a term rewriting system is, is not stated in the
	rules of the competition. So one could construe that a term rewriting
	system is whatever definition (s)he could find in the literature. If
	we want a consistent definition, then we have to cast it into a
	competition rule, which will be applicable next time:
	
	3) a term rewriting system, for the purposes of this competition, may
	introduce fresh variables at right hand sides, and may have variables
	as left hand sides. In the ordinary case (without strategies, types,
	or conditions), this will imply that the system is non-terminating.
	
	And, finally, a meta-rule:
	
	4) the rules of the competition are decided during a meeting at the
	end of the previous competition. This includes new categories or
	syntax extensions.
	
	Kind Regards,
	Alfons
	
	_______________________________________________
	Termtools mailing list
	Termtools at lists.lri.fr
	http://lists.lri.fr/mailman/listinfo/termtools
	

-------------- next part --------------
_______________________________________________
Termtools mailing list
Termtools at lists.lri.fr
http://lists.lri.fr/mailman/listinfo/termtools