Hmm, this brings up some interesting possibilities for constructible circuitry. The smallest known G-to-H converter is a p8 device -- nine still lifes and a figure eight -- much cheaper to construct than a Silver-reflector-based stable G-to-H, and really just as good as a still life from Golly's point of view, since it's a power-of-two period.
Here's an old p8N regulator that uses figure-eight reflectors, for example. If regulator functionality were actually needed in self-constructing circuitry for some reason, it could pretty easily be adjusted to be more Spartan. In its current form it converts arbitrary inputs (in this example from a p1419 gun on the right) into output gliders timed in multiples of 712 (the period of the gun at the bottom with the two Herschel transceivers).
Code: Select all
#C p8N regulator: minimum driving period is 712 generations,
#C minimum safe spacing between input gliders is 1414 generations
#C Dave Greene 2 May 2003
x = 373, y = 254, rule = B3/S23
P8 circuitry could manage slow-salvo construction of still lifes with no problem, since intermediate targets are at most P2 -- you just need four sets of ways to make gliders, for even- and odd-parity gliders on black and white squares, with at least one op in each set.
The problem would be with using p8 circuits to construct more figure eights. To be able to slow-build multiple figure eights synchronized with each other in a Herschel circuit, it might end up being necessary to have sixteen different glider-output elbow ops, eight for each color.
That sounds a bit troublesome -- not as bad as the p30 case would be, but definitely a headache... it's exactly what originally pushed me strongly toward stable (and Spartan) self-constructing circuitry.
It's certainly possible to work around this problem using just the four [even/odd, black/white] op sets, though. Just collect small one-time turner / rephaser constellations to get the seven required timing changes for the figure-eight seed's final trigger glider. Just a question of whether the smaller more efficient p8 circuitry is worth all the extra construction complexity.
codeholic wrote:16.117...16.118... 20.3424
You seem to have switched from hunting eater2 seeds to looking for dead-spark-coil seeds for constructible Herschel transmitters. The first one in particular looks very usable. I'll have to look again for places where Herschel transmitters and receivers have a clear advantage over plain old Silver reflectors.