From: "Kjeld H. Mortensen" To: designCPN@daimi.au.dk Date: Tue, 1 Jul 2003 17:26:14 +0200 Subject: Re: (dCPN) A question on guard For the initial marking "1`A++2`B" on the left place there are two cases: - The variable in is bound to A: The guard is only true if out is bound to C. - The variable in is bound to B: The guard is true regardless of what out is bound to. There are no restrictions on out except that it has to be bound to some colour from O. Hence, at the end of a simulation run, with the given initial marking, the marking of the right place should have at least one C token and two other tokens from the O colour set. Regards, -- Kjeld H. Mortensen * khm@daimi.au.dk * www.daimi.au.dk/~khm University of Aarhus * Department of Computer Science * Denmark Quoting "Taeho KIM \\(±è ÅÂÈ£\\)" : > I wrote down a guard which seems to be a syntax error, > however the Design/CPN 4.0 simulates unexpected behavior. > > This is an example : > > ( O ) --1`in--> [ (not (in=A)) orelse (out=C) ] ---1`out--> ( O ) > > color O = with A | B | C; > var in : O; > var out : O; > > ( ) : place > [ ] : transition > -- --> : arc > > The input arc to the transition specifies 'in' and doesn't specify 'out', > but the guard on the transition specifies both 'in' and 'out'. > It looks like a syntax error, however the Design/CPN passes the syntax > checking and does unexpected behavior. > > If the initial marking of the left place is 1`A+2`B, then the final marking > of the right place is 2`A+1`C. > > --- > [[ Post messages and summary of replies: designCPN@daimi.au.dk > ]] > [[ To (un)subscribe, send "help" to: Majordomo@daimi.au.dk > ]] > [[ The moderator's address: designCPN-owner@daimi.au.dk > ]] > [[ World Wide Web URL: http://www.daimi.au.dk/designCPN/email/ > ]] --- [[ Post messages and summary of replies: designCPN@daimi.au.dk ]] [[ To (un)subscribe, send "help" to: Majordomo@daimi.au.dk ]] [[ The moderator's address: designCPN-owner@daimi.au.dk ]] [[ World Wide Web URL: http://www.daimi.au.dk/designCPN/email/ ]]