>Yes, I remember. However, as long as initializers are executed before >anything else, I don't think it makes a difference. Continuous assignments could be considered to be waiting for input changes from the very start of simulation. They are not defined in terms of procedural code executing up to an event control and then waiting for an event. This would mean that an event caused by an initializer would cause them to evaluate, even though initializers are executed before initial and always blocks. So removing the statement about initializers not creating events would make a difference for continuous assignments. If continuous assignments are required to evaluate unconditionally at time zero, then this makes less of a difference. It could still make a difference in whether the time-zero evaluation of continuous assignments has to happen at a particular point. If initializers create an event, then it wouldn't matter whether the time-zero evaluation of continuous assignments happened before or after initializers. They would work properly either way. But I agree that it does not make a difference for always blocks. Steven Sharp sharp@cadence.comReceived on Tue Apr 19 15:05:24 2005
This archive was generated by hypermail 2.1.8 : Tue Apr 19 2005 - 15:05:27 PDT