[PEAK] Re: trellis.Set.discard

Sergey Schetinin maluke at gmail.com
Mon Nov 3 15:18:14 EST 2008


> I think I will not use the _finish() hack.  Since there is no actual use
> case for discrete reset, and I can definitely see use cases for them *not*
> resetting, I think that the "promise" we have is the right one.

The kind of use case for resetting the new cells I was thinking of is
when one of the cells in the new component reacts to the that discrete
cell, a logger for example. So that made at least some sense to me.

I'd like to know the case for not resetting the new discrete cells,
because I don't really have one for that either. But I can think of a
case like this:

@maintain
def rule(self):
    return NewComponent().discrete_rule

which would result in infinite retrigger of this rule. This is bad
code, but still a case to consider.

Overall, I agree with the decision, if the cell initialization
shouldn't be thought of as happening before the current transaction
everything seems to fit nicely.



More information about the PEAK mailing list