An open problem is a problem for which no solution has been found. An example is "Do oscillators of all periods exist in Conway's Game of Life?".
Unsolved problems can be subdivided into several basic categories:
- Periods: Do oscillators, spaceships, guns or puffers exist of a particular period?
- Unusual-growth patterns: What is the long-term effect of a predefined pattern? For example, it is unknown whether the Fermat prime calculator grows indefinitely.
- Solvable problems: Some problems are known to have a solution, but as yet no pattern has been built. For instance, no Conway's Life quadratic growth replicator has been built to date, but a workable blueprint is available, and no really large new technical problems would have to be overcome to complete the construction.
- Construction and destruction problems: These include finding the smallest Garden of Eden, building a stable eater that can absorb any single glider aimed at it, determining whether a particular object has a glider synthesis, or discovering an unstoppable-growth pattern.
- Spatial minimization problems: Find an object that satisfies some criterion that fits within a certain bounding box. Examples include Mike Playle's prize for a small stable reflector.
- Temporal minimization problems: As above, but concentrating on speed rather than size.
List of problems
Open and previously open problems include:
|Problem||Status||Year posed||Posed by||Year solved||Solved by|
|one cell thick infinite growth||solved||?||Nick Gotts||1998||Stephen Silver|
|The question “is there a one-cell-thick pattern exhibiting infinite growth?” was answered in the affirmative.|
|Coolout Conjecture||solved||<1992||Richard Schroeppel||2001||Richard Schroeppel|
|The question “given a partial Life pattern that's internally consistent with being part of a still life (stable pattern), is there always a way to add a stabilizing boundary?” was answered in the negative.|
|Garden of Eden||solved||?||?||<1970||Edward Moore|
|The existence of a Garden of Eden in Conway's Game of Life was known from the start because of a 1962 theorem by Edward Moore that guarantees their existence in a wide class of cellular automata.|
|Grandfather problem||solved||1972||John Conway||2016||mtve|
|The question “is there a configuration which has a father but no grandfather?” was answered in the affirmative.|
|The question “do oscillators of all periods exist?” remains open for Conway's Game of Life; no oscillators are known for period 19, 38 and 41, and no non-trivial oscillators are known for period 34.|
|Replicators||solved||?||?||1982||Elwyn R. Berlekamp, John Conway, Richard K. Guy|
|The question “do replicators exist in Conway's Game of Life?” was answered in the affirmative.|
|Still life finitization problem||solved||?||Dean Hickerson||2019||Martin Grant|
|The question "given a still life without finite boundaries, can any MxN finite window of it be preserved within a finite still life by adding appropriate unchanging cells around it?" was answered in the negative.|
|Unique father problem||open||1972||John Conway||—||—|
|The question “is there a stable configuration whose only father is itself?” remains open.|
|Universal computer / Universal constructor||solved||?||John Conway||1982||Elwyn R. Berlekamp, John Conway, Richard K. Guy|
|The question “does Conway's Game of Life support universal computation and universal construction?” was answered in the affirmative.|
- Berlekamp, Elwyn R.; Conway, John H.; Guy, Richard K. (2004), Winning Ways for Your Mathematical Plays, 4 (2nd ed.), pp. 927-961
- Martin Grant (December 28, 2019). Re: Unproven conjectures (discussion thread) at the ConwayLife.com forums
- Unproven conjectures (discussion thread) at the ConwayLife.com forums