Symmetrical Cellular Automata Control Panel V1.1 Changes
1. Addition of Seed Values 4-9 top
Some rules only create Checkerboard patterns when initialised with mirror symmetric seeds. This limits the seed to '0' and '3' with the existing choices. The change adds the ability to toggle the state of more cells on both sides of the boundary. These additional cells are adjacent to the cells currently modified and form a line with them. It is useful to keep numeric values as this is required for a random seed.
| Direct Code | RLE Code | dcba | ca-bd |
| 0 | o | 0000 | 00-00 |
| 1 | a | 0001 | 01-00 |
| 2 | b | 0010 | 00-10 |
| 3 | c | 0011 | 01-10 |
| 4 | d | 0100 | 10-00 |
| 5 | e | 1000 | 00-01 |
| 6 | f | 1100 | 10-01 |
| 7 | g | 0101 | 11-00 |
| 8 | h | 1010 | 00-11 |
| 9 | i | 1111 | 11-11 |
Inconstancies in seeds removed. seed characters '1' and '2' where reversed on diagonal bands to horizontal bands and to the description. This has changed the seeds for some patterns.
Done.
2. Change random seed with leading zeros. top
Leading zeros in random seeds are ignored as the seed is changed to numeric. This is different for other types and as the seeds used in current scripts do not have leading zeros they are not affected by this change. The change is to detect a leading zero and add an extra leading '1'.
Done
3. List of Changes from previous release. top
Done. link to this file added to help
4. Allow scripts to specify the location of Lua modules. top
New command "PATH" added before the segment.
Done.
6. Work through old Scripts and add regression tests. top
Not part of the control panel, an ongoing activity.
7. Diamond Island changed. top
Diamond islands where incorrectly implemented by rotating a square. This looked OK for a random islands but a solid island shows holes as a diamond shape on a grid has more cells than a square of the same length of side. Correctly implemented also allows the inverted pattern added for H010 (B4) to work. However this means that the random diamond islands are now different and the results for H1D8 will change.
Done.
8. Add update available to SymCACP. top
Done, added to about help and help.
9. Correcting Errors. top
- Correct seed implementation on a diagonal boundary.
The modifications to the boundary on diagonal bands was the reverse of the definition which is the consistent with the changes on an horizontal boundary.
To get the same pattern as before on a diagonal boundary swap "1" and "2" values of the seed".
Done.
- View H rule or B/S ends up displaying H000
Done
- Fix Solid Islands. Current workaround is change random ratio. There is no seed on solid islands?
Solid islands and inverted solid islands have seeds.
Done.
- On startup universe is cleared instead of keeping pattern.
Done.
- next seed list not updating label.
Done.
- SCRIPT script2-chaosIsle does not allow restarting.
Done.
- check that universe size is ok for diamond island fails.
Checks removed, Only cells in the universe are displayed.
Done.
- Rebuild from precious session invert island says true but not inverted.
Done.
Wish List for Later Releases top
- Using the Geometry to Determine Seed Symmetry more Accurately
- Allow scripts to call scripts.
Scripts need there own environment so that they do not alter the environment of the calling script.
- Tool for detecting Symmetry
- Exception Handling
Checkerboard patterns in some wave rules do not always create checkerboard patterns.
In order to create good statistics the search code should detect failure to create a checkerboard pattern and have extra seeds and make up the number of samples required.
- Some file handling exceptions are not handled so correcting path names can be a pain.<.li>
- script2 findEnd lua module reports the population of the largest state rather than the population of the state that golly reports. This is confusing.<.li>
| |
© 2024 Paul Rendell. |
Last Update 23/January/24
Comments to Paul Rendell |