diff options
Diffstat (limited to 'TODO.md')
| -rw-r--r-- | TODO.md | 17 |
1 files changed, 8 insertions, 9 deletions
| @@ -14,13 +14,11 @@ It's more of a personal reminder than anything else. | |||
| 14 | * invert an alg | 14 | * invert an alg |
| 15 | 15 | ||
| 16 | ### More steps for `solve` | 16 | ### More steps for `solve` |
| 17 | * QTM optimal solving | 17 | * QTM optimal solving (important: fix possible_next, which works only for HTM now) |
| 18 | * Block-building steps (cross, roux blocks, ...) | 18 | * Block-building steps (cross, roux blocks, ...) |
| 19 | * Other common steps (LSE, ...) | 19 | * Other common steps (LSE, ...) |
| 20 | 20 | ||
| 21 | ### Improvements to currently implemented commands | 21 | ### Improvements to currently implemented commands |
| 22 | * batch mode: add separator / info on which command it is executing | ||
| 23 | (also change man page for this) | ||
| 24 | * solve should re-orient first if needed and not just give up if centers are off | 22 | * solve should re-orient first if needed and not just give up if centers are off |
| 25 | * solve should try up to a small bound without loading the large pruning table | 23 | * solve should try up to a small bound without loading the large pruning table |
| 26 | * drfin for HTR scrambles should try all 3 axis and pick the best solutions; | 24 | * drfin for HTR scrambles should try all 3 axis and pick the best solutions; |
| @@ -42,7 +40,12 @@ It's more of a personal reminder than anything else. | |||
| 42 | 40 | ||
| 43 | ## Technical stuff | 41 | ## Technical stuff |
| 44 | 42 | ||
| 45 | ## Performance | 43 | ### Memory management |
| 44 | * Check if memory is enough for loading pruning tables; if not, abort | ||
| 45 | * For optimal solver: choose largest that fits in memory between | ||
| 46 | khuge, shug6 and light | ||
| 47 | |||
| 48 | ### Performance | ||
| 46 | * solve (allow_next): filter out based on base_move; only check once for each | 49 | * solve (allow_next): filter out based on base_move; only check once for each |
| 47 | triple of moves; how to deal with different movesets? | 50 | triple of moves; how to deal with different movesets? |
| 48 | * try htr corners + edges in slice but not oriented (300Mb table); | 51 | * try htr corners + edges in slice but not oriented (300Mb table); |
| @@ -51,12 +54,8 @@ It's more of a personal reminder than anything else. | |||
| 51 | (like in light optimal solver) | 54 | (like in light optimal solver) |
| 52 | * Another idea: DR + cornershtr (5Gb table); same as above, de Bondt's trick | 55 | * Another idea: DR + cornershtr (5Gb table); same as above, de Bondt's trick |
| 53 | does not work but I can use half-turn trick | 56 | does not work but I can use half-turn trick |
| 54 | * On the contrary: DR + separate UD corners allow dB's trick, but no ht-trick | ||
| 55 | 57 | ||
| 56 | ## Coordinates, symmetries, pruning tables | 58 | ### Coordinates, symmetries, pruning tables |
| 57 | * Cleanup symcoord.c: some coordinates and symdata are never actually used; | ||
| 58 | remove also sd_eofbepos and just use sd_coud for khuge (this changes the | ||
| 59 | coordinate so the whole table must be generated again!) or viceversa | ||
| 60 | * Use pruning values mod 4 instead of mod 16 (or maybe not, I like the | 59 | * Use pruning values mod 4 instead of mod 16 (or maybe not, I like the |
| 61 | current system) | 60 | current system) |
| 62 | 61 | ||
