Normal sudoku rules apply. Identify two "indexer" cells in each row, column, and box, with one indexing in the row and the other indexing in the column. The digit in an indexer cell indicates the position of the other indexer cell in its row or column. Positions are counted left to right in rows and top to bottom in columns. The value of an indexer cell is the digit in the cell it indexes. For all other cells, their value is their own digit. All lines are region-subset lines: ie 3x3 box boundaries divide each line into segments. For any pair of segments on the same line, the values on one of those segments must all appear on the other segment. Examples: i) If () represents an indexer cell, then 12(6)34(7)598 would be a valid row of digits, because there are exactly two indexer cells, and one of the cells indexes the other: the (7) is in position 6, and the other indexer cell contains a 6. The value of the (6) cell would be 7, because that is the digit in the cell it indexes. Given that this is a row, we would also know that r7c6 must be an indexer cell, because the (7) cell indexes it. ii) A region-subset line could contain the values 189-8-98-918, where a dash represents a box boundary. This is valid because (8) is a subset of (89) and (189), (89) is a subset of (189), and (189) is a subset of itself.