"Dean's Eaters" Paul Rendell's universal Turing machine Gabriel Nivasch's Caterpillar Paul Chapman's MRM David Bell's Unit Life Cell Jared Prince's Deep Cell Jason Summers' telegraph David Bell's adjustable-period receding LWSS shuttle, boat-burning Cordership spaceship 29 Feb 2004 Elbow ladders -- Scot Ellison's (?), Nicolay Beluchenko's 21 Feb 2004 Stamp collection of syntheses from Mark Niemiec, e.g., 2 March 2004 Put together some tubstretchers and tubeaters: 4 March, 18 March, 5 April 2004 Paul Chapman's stable SBM 5 May 2004 -- there's a stable MRM using it, too... Paul Tooke's engineless p195 2c/5 rake -- 2 Jan 2003, 14 Jan 2003 updates -- look in rakes collection p38 rakes 29 Oct 2003 p62 rakes 20 Apr 2003 (13,1)c/31 technology 7 May 2003 Nicolay Beluchenko pi reactions, August 2003? Jason Summers O(log t) logarithmic-growth puffer 16 August 2003 c/2 period 38 and 70 rakes, 7 Nov 2003 p168 c/4 orthogonal puffer Paul Tooke, Jason Summers 14 Nov 2003 p52 and p52*N rake 16 Nov 2003 results from Paul Chapman's Glue project: four-eaters pattern, glider turners 29 Jan 2004? add to small guns in lifebc: p36, p156, others? p36 tug-of-war 23 July 2004 Gabriel Nivasch's lightspeed signals in "grey" agar c/4 linestretcher 22 Nov 2004 -- maybe mix with some of David Bell's recent lineship/line-cutting results? p185 engineless 2c/5 rake (update by Jason Summers 15 Dec 2004) David Bell's 'bobsled run' and 'swimmer' (switch engine track) David Bell's period-doubling Cordership reactions, 26 June 2005 Andrew Okrasinski new methuselah 14 July 2005 Nicolay Beluchenko's spaceships? HK database export, maybe? Jagged Ruler Other rules: Stephen Morley, B368S245 David Bell's large oscillators in Just Friends rule ? -- Then there's some stuff that I've helped build... so someone else will have to make the judgment call on whether it's worth including! programmable constructor (memory-tape and glider-loop versions, building an eater and a line of eaters, respectively) sample big guns, from completion of gun collection: maybe even 14-100? Wide variety of Life technology, and some impressive ants' nests and Herschel pipeworks 2c/3 signal: ugly and oversized, but I still haven't rebuilt it yet and it _does_ work... stable pseudo-Heisenburp device? Not quite finished... stable block-pusher? 7 Sept 2003. Probably not interesting enough. universal regulators: Paul Chapman's p30N -- 26 March 2003, my slower p8N -- 2 May 2003. Both on one page with test guns makes a fairly large-period oscillator... Cordership guns: 17 and 19 January 2004; 7-in-a-row V-gun with eater -- 4 July 2003 2c/5 spaceship puffer (Jason's) -- 23 March 2003; 2c/5 gun -- 11 April 2003. unclustered Blockic starseed, and p256 oscillator seed -- examples for Glue results Huge prime periods -- 7 August 2003 Freeze Tag 29 Feb 2004, but use 2005 updated version H2G stamp collection with boojum reflectors 16 April 2003 ----------------------- I was reading over old ideas for extending RLE recently, to try to figure out what would make sense for a new/extended format to support (and how best to make any extensions as backward-compatible as possible.) An idea from Nicolay Beluchenko seemed really interesting: shouldn't it be possible to support the simulation of patterns that extend to infinity in one or more directions? -- Meaning regular patterns like wicks or agars that repeat out to infinity, of course; random soups need not apply. Of course, people tend to like to make soup out of random agars by perturbing them, and no doubt Golly would have a hard time keeping up with the usual lightspeed explosion after a while... Anyway, obviously existing formats don't have the flexibility. But if at some point we get around to defining a multi-compressed-chunk format, eventually it might be possible to support parameters for one or more chunks saying "add copies at (n*dx,n*dy) for all n". However, I could imagine that these repeating units might easily become a bookkeeping nightmare, especially let's say at high zoom levels. To avoid this kind of problem and keep patterns strictly finite, maybe a halfway solution would be to support (dx, dy, dt) parameters -- a copy of the subpattern would be added at the given offset every dt generations. This would allow for a good representation of patterns that need an infinite support system (fenceposts for wicks with no known stretchers, reactions that need a stream of gliders for support, lightspeed signals in a striped medium, etc.) I'm thinking about this because both on gameoflife-news.blogspot.com and in the prospective Golly pattern collection, I generally prefer to include only patterns that can be run indefinitely without blowing up... but occasionally things show up that are definitely worth seeing, but that are only supposed to be run up to generation N -- because if you run them any farther, you'll get an uninteresting mess that is not part of the intended exhibit. So that's another possible parameter to add to a pattern format: "stop after generation N". Or even "auto-replay gens 0-N after t seconds." Or else for things like guns, mark the unit of repetition [the output glider] somehow, and optionally suppress the output without having to add an eater? Thinking a little more along these lines: how about a hint for "should be run with hlife / qlife only"? What about "suggested speed = t ticks/second"? All of these could help improve the default presentation of many patterns. For example, most of the above make sense for Dean Hickerson's block-and-beehive puller -- "vacuum.lif" in Jason's main collection -- where the action is all over in a fraction of a second at warp speed and then nothing interesting happens after that: the pattern just starts expanding to infinity so that an autofit operation shows increasingly uninteresting satellite views of long lines of gliders. -- Not that any of these hints should be binding or anything; I'd visualize something like a "padlock" icon in Golly, where if you click to unlock the padlock, Golly allows the current pattern to run without any further interference.