Like the 2003 examples, this pattern consists of a Herschel/glider loop (period 1536 this time, instead of period 1450) attached to a series of thirteen 'quadrupler' conduits of two different types, L-shaped and straight. Each quadrupler allows only one Herschel signal in four to get through to the next conduit.
In the starting configuration, the first three quadruplers are set to absorb three signals each: an extra block has been added to the first two L112 quadruplers and the first Fx70 quadrupler, which has a standard F166 dependent conduit appended to suppress the following Herschel's first glider. So the circuit initially counts down from 63. When all circuits are empty, a signal gets through to activate the 263-step glider advancer -- after which the circuit counts down again, this time from 4^13-1.
As usual with this kind of Herschel-track construction, Karel Suhajda's 'Hersrch' search program was used to design the base loop, and to locate an efficient connecting circuit between the end of the quadrupler chain and the glider advancer. The continuing challenge, of course, is to fit an oscillator with a higher prime period into a bounding box with a smaller number of cells.
Here's another batch of puffer orbits found by Jason Summers.
In the following tables, the percentages are approximate, intended to illustrate how common or rare a particular puffer might be. The bit patterns are not the actual puffer, but a simpler pattern which will eventually evolve into the puffer.
Thanks to John Green for pointing out that a number of the RLE files in the Puffer Orbits would not download properly. In fixing the problem and reviewing other possible problems, I found several other bad links. Please let us know when those things happen. If you can't see the images, or get and run the RLE files, then a mistake was made and it needs to be fixed. Or you may be using a combination of browsers and operating systems with which we are unfamiliar and may not realize we are breaking.
However, since the stripes are not diagonally symmetrical, each of the edges would need a different type of support. Whether such a construction is possible is currently an open question. The main problem is that diagonal spaceships could not repeat at period 4, because each stripe is two cells away from the previous one -- so p8 spaceships would be needed.
Unfortunately p8 is not really within reach of currently available Life search programs (mainly David Bell's 'lifesrc' and Jason Summers' ' WinLifeSearch' port). Some fairly large leaps in efficiency and/or CPU speed would be needed to make this a likely goal. [Readers are, of course, invited to prove this statement wrong -- or to supply the necessary advances in search technology.]
Here are separate commented files for Beluchenko's previous wave-based spaceships in this series, from 7 April 2006, posted last month in a combined pattern:spaceship made with wave ends
It was shown in 1993 that signals travelling parallel to the striped medium must move at the speed of light -- but it turns out that non-destructive signals perpendicular to the stripes are also possible, and these follow different rules. In particular, they can move at a speed of 2c/3 orthogonally, slower than lightspeed but faster than a spaceship can travel through vacuum.