How does automation in the game really work?

[BLACKBERREST3]

So I was practicing with automation permutations and I came across some really weird behavior that seems utterly random. I'm trying to see what makes a "super Oscillator" tick on a memory gate. I'm hoping someone can shed some light on this. I tried testing different lengths of wire and was getting nowhere fast because one minute, a design will oscillate and then the next time I deconstruct and build it again, it won't oscillate. this is incredibly confusing, mainly because I do not understand what automation does for a single tick. the same thing can be done with a not gate if you connect the output to its input and also some permutations with xor gates. The switches that are turned on represent the ones that oscillate.

Another weird instance which I do not know if it is reliable or not is that the longer wire will oscilate, but the shorter one will not even though they are coming from the signals that are going through the same places. this makes me think that the game calculates the total length of wire in a circuit to perform calculations.

This is the one that decided to stop oscillating after rebuilding it.

 

I'm uploading a vid, stand by

 

 

[Risu]

Probably seeing a quirk of build order influencing the logic propagation.
 

[BLACKBERREST3]

I know that build order can affect a shutoff getting stuck, but I didn't know it accounted for automation wires and gates as well. I think that does play a part.

But then I don't understand this.

I can't even replicate the build order bug with the shutoff, but I know it happens in game. I've had to reset my batteries because of this.

[Steve Raptor]

The automation in the game doesn't account for constant propagation delay like in real logic gates.

So the behavior of connecting the same wire to two ports, for example, will result in random results every time, there will always be a "Race" between the two with no consistent "winner" between the ports.

The length of the wire has no effect on this, and every time you re-connect a circuit, it will generate a different outcome.

[Nate82]

I really jumped into the area after working at HP labs on software verification.

[Tonyroid]

Wow.

Some of this stuff is crazy. I haven't tried any of it out but my intuition is that you are exploring the realm of "undefined behavior". The differences could be caused by literally anything. Maybe build order, but also maybe whether the third length of a wire is referenced in an odd numbered position of an array that is referred to by a subroutine that supports the processing of automation gates. Nobody knows because the designers didn't care to explicitly control it.

I think it's a realm that's worth exploring because you sometimes find neat behaviors. But you also shouldn't hang your hat on anything because it can be randomly changed just as easily as it was randomly created.

[BLACKBERREST3]

6 hours ago, Tonyroid said:

I think it's a realm that's worth exploring because you sometimes find neat behaviors

That's the goal. I haven't found anything super useful yet out of this.

15 hours ago, Steve Raptor said:

The automation in the game doesn't account for constant propagation delay like in real logic gates.

That's interesting. I'm hoping to make use of that somehow.

I'm not great with logic and automation, but this game makes learning it a little easier and I'm going about it using a different learning method. I'm not done with it yet, but I'll post about it later with a save file and blueprint. I've found RNG gens, oscillators, latches, flip flops, edge pulses, etc. from combinations of only 2 gates, this can go a long way in removing excess automation from my builds.