A summary of the known orthogonal spaceship speeds is given in the following diagram, using Ford circles to represent rational numbers:

The HighLife replicator after thirty-six generations

Soon after the discovery of the replicator, it was realised that it could be tamed into a c/6 spaceship by pulling a blinker behind it. In 1999, Dean Hickerson proposed the existence of spaceships with much slower velocities, obtained by pushing junk at one end of a replicator track and pulling it at the other end. No explicit examples of spaceships were discovered this way, although Dean found a workable push reaction. This was mentioned on David Eppstein's website and in a chapter he wrote for Game of Life Cellular Automata.

It was pretty much forgotten for 14 years, until Adam P. Goucher wrote a search program to attempt to construct replicator tracks capable of forming spaceships. Initially, he found a c/69 spaceship with over 84 billion replicator units; his results and method of searching are summarised on Complex Projective 4-Space. Due to its immense size, slow movement and general appearance, it was named the Basilisk. Karel Suhajda commented on the post, suggesting trying different speeds. Tweaking the search parameters resulted in a c/63 spaceship with about 2 billion units; however, this was still prohibitively large for Golly.

Matthias Merzenich's (6,6) push reaction

Matthias Merzenich found a vastly more efficient push reaction, which moves a constellation by (6, 6) as opposed to Dean's (8,8) push. Adam P. Goucher did further alterations to incorporate this into the search program, which gave a c/24 Basilisk of just over 15000 replicator units. This is comfortably within the limitations of Golly, and there is an explicit RLE file of it. Using this technology, he proceeded to build c/24 glider rakes and even a gun to periodically emit c/24 Basilisks! A more comprehensive description and links to pattern files were posted on a later cp4space article.

A gun and a series of c/24 Basilisks emerging from it

Helmut Postl subsequently discovered a c/32 Basilisk, which is slightly smaller than its c/24 counterpart. Together with Matthias' c/5 diagonal spaceship from a couple of years ago, there are now twelve known spaceship velocities in HighLife: c/2, c/3, c/4, c/5 and 2c/5 orthogonal; c/4, c/5, c/6, c/24, c/32, c/63, c/69 diagonal. Additionally, Dean found a c/22 quadruple blinker-puffer by composing his (6,6) push with a (6,6) beacon pull, but it is unlikely that it can be exploited to yield further technology.

This all stems from the existence of the replicator in HighLife. If a similar construct were created in ordinary Life (proved to be possible), then similar XOR-extendable spaceships and so forth could be built. Dave Greene has been researching the possibilities in this direction.

The core of the reflector is a staged-recovery mechanism found in an earlier 487-tick reflector. The speed-up is therefore achieved by surrounding the core with a more efficient Herschel track (exploiting the new conduits), enabling the gliders to be delivered to the active site faster than before.

In other news, there is now a continuous version of the Game of Life exhibiting rich behaviour. It cannot be simulated in Golly due to its incompatibility with HashLife, although I believe the next release of Ready will incorporate it.

With Golly, we can run the Mandelbrot set in a cellular automaton. The results are fairly uninteresting with B3/S23, so I simulated the boundary (obtained from the original image by one generation of B3/S23) in HighLife (B36/S23) instead. As with all sufficiently large chaotic HighLife universes, profusions of replicators emerge:

You can download the files from Complex Projective 4-Space yourself if you're interested in running a simulation. For these purposes, you'll want the 262144 by 262144 monochromatic image (25 MB download as .mc.gz), rather than the scaled-down colourful version.

In both HighLife and ordinary Life, we obtain four waves of gliders, a handful of orthogonal spaceships and a few diamonds on the real axis. These diamonds are caused by long unbroken lines expanding. In Life, we get two back-to-back Sierpinski triangles, compared with a uniform pattern of blinkers in HighLife. Further, these blinkers are very vulnerable to chaos, resulting in the destruction apparent in the green-bordered inset image.

The developers of Golly have recently turned their attentions to creating a new piece of software capable of supporting reaction-diffusion systems and cellular automata on arbitrary meshes. This has been discussed on Complex Projective 4-Space and The Aperiodical, amongst other places.

There has also been some work on cellular automata on Penrose tilings. Nick Owens and Susan Stepney investigated B3/S23 a while ago, writing a chapter about the topic in Adamatzky's Game of Life Cellular Automata. This summer, a couple of related, independent and almost simultaneous discoveries were made. One of these was a weakly universal cellular automaton on a Penrose tiling; the other was a glider.

The glider attracted a lot more attention, not least because a $100 prize was offered for its discovery. Again, this was mentioned on Complex Projective 4-Space, New Scientist and The Aperiodical. Even more recently, it was investigated on the kite-and-dart Penrose tiling, where it follows fractal paths and symmetrical loops. An image is displayed below:

The lengths of loops (half the period of the resulting oscillators, since the glider is c/2) are tabulated in the sequence A215878 of the Online Encyclopedia of Integer Sequences.

Smallest GoE.

Marijn Heule, Christiaan Hartman, Kees Kwekkeboom and Alain Noels systematically searched the entire space of 10-by-10 patterns with fourfold rotational symmetry, finding a Garden of Eden with 92 specified cells (56 live, 36 dead). Moreover, they proved the non-existence of Gardens of Eden within a 6-by-6 box.

Flow matrix for five common transient objects.

Firstly, what is so special about the Herschel? Is it really so much more useful than any other transient objects? It appears that the answer is both yes and no: other objects can be used, but they must eventually decay into Herschels. This is illustrated rather eloquently by a simple matrix. The row represents the input; the column represents the output. A red blob indicates if a primary (one-stage) conduit exists to transform the input into the output. Clicking on the matrix will enable you to download a complete collection of primary conduits. (A collection of all conduits, primary and composite, is provided later in this article.)

Guam's 309-generation left-turn conduit.

Some of these conduits are new discoveries. The Pi-to-R converter was discovered by Guam on the conwaylife.com forums, published in the form of a quaternary Herschel conduit: H-Pi-R-B-H. The completed conduit takes 309 generations to turn a Herschel anticlockwise, so is designated L309. In terms of the number of intermediary objects, L309 is the most complex Herschel conduit to date. Indeed, its 309-tick delay is rather rapid for a quaternary conduit.

Isomers of Guam's 266-tick Herschel conduit.

Not content with a single new conduit, Guam proceeded to discover a pi-to-Herschel converter capable of attaching to a handful of 'pre-Herschel' conduits. Moreover, the symmetry of the pi heptomino means that Guam discovered not only one conduit, but two 'isomers'! The conduits are designated F266 and Fx266 for the translation and glide-reflection variants, respectively. However, the restrictions upon which conduits may follow F*266 severely limit its use in practical Herschel circuitry.

Matthias' tertiary Herschel conduit, the Lx496.

From an earlier posting of mine, you may remember the contributions of a certain 'MikeP', again from the conwaylife.com forums. Matthias Merzenich has utilised a particular catalyst of his in a few conduits, including the periodic R135 conduit and a stable Pi-to-century converter. To process the resulting century, Matthias proceeded to find two unique century-to-Herschel converters, one of which is sufficiently compatible to yield a new tertiary Herschel conduit, the Lx496.

Matthias even found a use for this new inundation of Herschel conduits; he has incorporated them into glider guns with smaller dimensions than the current record-holders. Specifically, they are a p421 gun derived from the L309 and p685 gun based on the Lx496.

Guam's 4hd Herschel receiver.

In addition to his spectacular Herschel conduits, Guam also found two reactions in which gliders collide with constellations of still lifes to form extra junk. We already have a glider-to-beehive and glider-to-block converter, virtue of Paul Callahan, but we can now add loaves and bi-blocks to the collection. The latter is especially interesting, as a glider in the same direction as the original can liberate the bi-block in the form of a Herschel. The gliders are separated by 4 half-diagonals, unlike in previous receivers, where the separation must be 2, 5 or 6; hence, this new receiver could function where others fail. Also, it transpires that the bi-block functions as a LWSS eater, which can be toggled by incoming gliders.

Finally, he noted that two gliders separated by 4 or 5 half-diagonals can be reflected into a single glider. Here it is demonstrated as part of a stable reflector and related pulse divider. Alas, this reflector does not break any records, unlike the next subject of discussion -- the rectifier. This fast reflector, subject of a previous article on LifeNews, can be used in various conduits for transforming Herschels into gliders, by either modifying the output or assisting in the cleanup of surplus blocks.

To summarise this article, here is a collection of the 30 distinct Herschel conduits (including four adjustable ones), and a comprehensive collection of every (sufficiently simple) conduit, transceiver and converter known, as of the time of writing.

Matthias' extensible c/7 diagonal spaceship.

Moreover, Matthias has actually discovered an infinite family of such spaceships, as one of the frontal components can support itself to yield an extensible spaceship.

Pulsar-based display device

'Triller' from Nathaniel's forum has engineered an impressive construction: a fully functional display using pulsars to represent individual pixels. The display is continually refreshed at regular intervals, updating it with the data contained within several memory loops.

As he/she has included a comprehensive description of the mechanism, it would violate Occam's razor for me to describe it in great detail. However, there are some interesting features worth mentioning:

• Triller has opted for period-30 technology, as opposed to the more modern option of Herschel tracks. This makes the device significantly more compact than its stable counterparts.
• The image data is contained in data loops, similar to the Golly-ticker in Golly's pattern collection.
• The sample image is the set of hexadecimal symbols, displayed in alternately ascending and descending order.

Matthias Merzenich has finally found this long-awaited anteater, enabling completion of the antstretcher.

Matthias' antstretcher, rendered in Union Flag colours to celebrate the recent Royal Wedding!

Eight copies of this can be combined to create another example of a 'space-nonfiller' (a term coined by Jason Summers), the earliest such example being discovered by him in 1999, which expanded at the vacuum speed limit.

Jason Summers' original nonfiller.

Quadratic growth pattern of width 1.
Stephen Silver, 20 April 2011.
Uses a breeder by Nick Gotts.

At right is a diagram shows what the full pattern looks like, with a sample section of the generating line of cells expanded to explain the mechanism used to construct the breeder. Line sections are arranged to produce exactly-timed two-glider salvos, which collide to produce LWSSes, which in turn collide to build the breeder. A multi-step reaction at the X axis produces the second glider in each pair with an exactly-timed delay relative to the first one.

The width-1 breeder after two million ticks, showing the first six switch engines
heading NW and SW, plus other stable and traveling detritus left over from
the construction process.

The new Turing machine emulator is period-23040, which makes it more HashLife-amenable than the previous UTM and TM emulators (p18960 and p11040, respectively). Additionally, the diagonal stacks run faster in Golly than the oblique analogues. The c/2 version outperforms the c/5 version, apparently, despite having a larger growth coefficient.

This is the first universal computer in Life to emulate a Turing machine in linear time (Chapman's URM takes exponential time; my UCC takes quadratic time), and therefore the most efficient to date.

Exploded diagram of Paul's SSS breeder.

To celebrate Paul Tooke's 50th birthday, this article is dedicated to one of his recent discoveries. Happy birthday, Paul!

Paul has recently been assembling patterns to defy common intuition about breeders, and thus help to determine a valid definition for what constitutes a breeder.

He has used the principles behind Gemini -- glider loops and universal construction -- to build unusual breeders with obscure properties. For example, he has engineered a SSS breeder, which amounts to a slide puffer (slide gun with stationary output) constructing more slide puffers. Moreover, he has designed it to have O(t^1.5) growth, rather than the O(n^2) typical of most breeders.

Paul's breeders, including a related SMS breeder, are available on the relevant forum thread. Rather than using the original Gemini construction arm, he has used an alternative construction arm known as the 'Pianola'. In the SMS version, this lays down block-laying switch engines; in the SSS version, this produces slide puffers instead.

Paul Rendell's c/12 stack constructor.

First: a summary. Last month, LifeNews published an article detailing Paul Rendell's stack constructor, a pattern capable of extending the tape of his Turing machine without limit, thereby endowing his UTM with computation universality. The original version of his stack constructor comprises two orthogonal c/2 rakes, which stretch out an expanding wave of gliders, coalescing to form the tape.

Despite working as planned, Paul was somewhat unsatisfied with his stack constructor. Its population soon explodes due to the waves of gliders, an artefact of using two orthogonal rakes to build a diagonal structure. The natural solution would be to use a diagonal puffer to extend the stack; this is precisely what he has accomplished.

The second version of his stack constructor uses c/12 rakes to provide the influx of gliders necessary for stack construction. However, there is a catch. The tape has a spatial period of 90, so the temporal period of the Cordership fleet must be twelve times that -- or 1080. Corderships, on the other hand, have a temporal period of 96n, so the smallest multiple of the desired period is 4320. This means that four rakes are required to emulate the rakes. This is accomplished in a somewhat similar way to how gliders are interleaved in pseudo-period guns, but with rakes instead.

Additionally, because his rakes use a kickback reaction, the size of the rakes are increased by another factor of two. Here is the RLE file for the completed rake. The resulting stack constructor is truly immense, similar in population count to the Caterpillar, and the corresponding RLE file is 21.8 megabytes in size!

You may be wondering where the second piece of 2010 technology comes into the equation. Recall that four rakes are interleaved for each glider in the stack constructor, which is rather sub-optimal. Clearly, then, there are two options: redesign the tape to have an offset of (say) 96 cells, instead of 90; or forget c/12 technology completely.

Adam P. Goucher suggested the latter, advocating the use of the emerging field of c/5 diagonal technology. He modified Matthias' c/5 rakes to work at p450 instead, resulting in an immense glide-reflective rake, not much smaller than the c/12 pseudo-rakes! However, after several successive optimisations, and a sudden insight from Matthias, he managed to produce a substantially smaller rake:

Adam P. Goucher's p450 rake, based on a design by Matthias Merzenich.

Paul modified the rake yet again, to form a variant capable of inserting gliders using the kickback reaction. Instead of the naïve approach of using two rakes, he used the same engine to produce both incident gliders for the kickback rake. This cuts down on the population count by a factor of two.

The stack constructor based on the c/5 rake is almost five times smaller by population count than its c/12 counterpart (RLE file). What follows is a diagram of the constructor, similar to the previous one, giving an idea of the dimensions of this beast:

Paul Rendell's c/5 stack constructor.