I classify a bug being major if it can affect game play (making an illegal move or a losing one instead of the winning one, etc.), otherwise it is considered to be a minor problem.
| Problem | Likely / actual cause | Solution | Status |
| Engine did not make the best move in positions with pawn promotion races . | The scoring for a won pawn race is too high. | Reduced the scoring. | Fixed. |
| Does not make a move in 20/40mins games. | The time extension caused an infinite loop. | New code and a flag to only allow one time extension per move. | Fixed. |
| Problem | Likely / actual cause | Solution | Status |
| The program sometimes stops after finished solving a number of testset positions under Fritz GUI in UCI mode. | Unknown. | None. | Open. |
| Failed to use a large opening book in the "pmx" version when no indexing file is present. | The use of opening book look up was incorrectly disabled when the large book could not be totally loaded into memory and no indexing file was found. | Correct the error. Now the large book file will be accessed without an index file, just as intended. | Fixed. |
| Learning did not occur. | The hashing function for probing of learned positions were changed in 2.6.2 but the saving hash function was not - hence probing the learned hash table will almost always fail. | Corrected the problem. | Fixed. |
| The numbers printed when merging book files were large. | The number of positions were counted but for "good" position, this count could be large. | Displays the number of records read and written instead. | Fixed. |
| Outposted pieces gets a wrong bonus. | Wrong pawn attack map for finding outposts. This is minor since the bonus for an outposted piece is small. | Corrected the attack map. | Fixed. |