chris_c wrote:There's a pair of eaters that can be welded together closely. Also changing a snake into an eater gives an extra two lanes of clearance...
Come to think of it, the size can be reduced a bit more by using an Fx77 instead of an F117 for the first Herschel-conduit extension, which lets the trombone-slide Snarks move five cells closer to the circuit and improves the repeat time by 40 ticks:
Code: Select all
x = 243, y = 167, rule = LifeHistory
Thinking a little further outside the box -- or rather, throwing away the box entirely and replacing it with a handbasket -- here's a somewhat larger demultiplexer with adjustable width, and also adjustable placement of the boat turner.
As shown below the circuit has 55 transparent lanes. To adjust the width in two-lane increments, move the southeast part of the circuit southeast (to increase the width) or northwest (to decrease it).
The boat is currently on the SEmost of the transparent lanes. To move the boat N cells (2N lanes) northwest, move the two Snarks in the northwest away from the circuit by N cells.
Code: Select all
x = 312, y = 239, rule = LifeHistory
There might be an Oe2 splitter available that's a little smaller and/or can connect directly to the syringe without an intermediate R64, if anyone wants to have a look
Any of the Oe2s is guaranteed to swap in cleanly for the one I used, unless you rotate or reflect the H-to-2G... otherwise you might need an Eo2 instead. Or, to put it another way, a rotated or reflected Eo2 splitter might work and have better geometry.
So the synchronization problem is a little bit complicated -- but only half as bad as the Oo1/Ee1/Oe3/Eo3 group (!). I hope someday somebody comes up with a splitter rating system that's a little less awkwardnessful.