| Age | Commit message (Expand) | Author |
|---|
| 40 hours | Compiler warning | tslil clingman |
| 40 hours | Finally fixed move generation | tslil clingman |
| 40 hours | Can't stop half-way through turn | tslil clingman |
| 40 hours | In-lined as much as a i dare | tslil clingman |
| 40 hours | Finally fixed memory order of weights! | tslil clingman |
| 40 hours | Opps, wrong stone | tslil clingman |
| 40 hours | Inlined placements, should be faster | tslil clingman |
| 40 hours | Memory oob fix | tslil clingman |
| 40 hours | Revert "Starting work on MCU port" | tslil clingman |
| 40 hours | First steps towards character LCD | tslil clingman |
| 40 hours | Leftover from another branch | tslil clingman |
| 40 hours | Starting work on MCU port | tslil clingman |
| 40 hours | New weights from a wider training | tslil clingman |
| 40 hours | Beginning to work on probabilistic depth now that MCU is out | tslil clingman |
| 40 hours | Went back up to 1986 weights | tslil |
| 40 hours | Quick sketch of non-determinism | tslil |
| 40 hours | Another road bug | tslil |
| 40 hours | Dragon is not separate, fixed concession, allowing playing as b/w | tslil |
| 40 hours | Changed function back to macro | tslil |
| 40 hours | There was a bug in the minimax! | tslil |
| 40 hours | Better trained weights, an important oversight in minimax (steps <=) | tslil |
| 40 hours | So after a small tweak, it actually plays well! | tslil |
| 40 hours | Seems tolerable, but not great. Still wont win or lose ??? | tslil |
| 40 hours | Trying alpha-beta pruning. There's a bug somewhere ... | tslil |
| 40 hours | Fixed minimax (!), fixed bugs in tak.c | tslil |
| 40 hours | Trying minimax | tslil |
| 40 hours | Crass error | tslil |
| 40 hours | Weights? | tslil |
| 40 hours | Wider? | tslil |
| 40 hours | Trying various things to teach ti | tslil |
| 40 hours | Have to account for CPS | tslil |
| 40 hours | Well it works, and it's bad. Many bugs fixed in interim | tslil |
| 40 hours | Closing in! | tslil |
| 40 hours | Auto-generating weights | tslil |
| 40 hours | New weights | tslil |
| 40 hours | one output (of course), Pade approximant of logistic for activation | tslil |
| 40 hours | Train on data that is close to the end of the game only | tslil |
| 40 hours | This finally works! | tslil |
| 40 hours | Close in principle, but there;s a bug in ct1997 &/ pptdb! | tslil |
| 40 hours | This is the general idea, but i'm sure order is incorrect somewhere | tslil |
| 40 hours | Starting on importing network! | tslil |
| 40 hours | Fixed a small bug in PTN parsing, other stuff | tslil |
| 40 hours | Check for road wins before flat wins | tslil |
| 40 hours | Tabs for indentation, spaces for alignment | tslil |
| 40 hours | Fixed indentation and some bugs, stats programme | tslil |
| 40 hours | Set dialect to C99 | tslil |
| 40 hours | Incorrect usage of globals in header | tslil |
| 40 hours | In-lined a few things, reworked flat count to iterate only once | tslil |
| 40 hours | tak.h now tracks win state, optimisations, saved a byte on counts | tslil |
| 40 hours | Line-mode mostly complete | tslil |
