Fill in a diagonal line in every grid square so that there are no loops in the grid, and so that every numbered point has that many lines meeting at it.
On a touch screen, a tap does what a click does, and a long press what a right-click does (more ways to right-click).
Click a square to cycle it through \, / and empty. Right-click it to cycle it the other way, through /, \ and empty. Keep the button down and drag to do the same to every square you pass over that looked the same as the first.
With the keyboard, the arrow keys move a cursor around the board. Enter does what a click does to the square under it, and Space what a right-click does. Press \ to draw a \ in the square under the cursor; clicking it round gets there too. Press / to draw a / in it; clicking it round gets there too. Press Backspace or Delete to empty it; clicking it round gets there too.
The Mouse button order preference swaps the two buttons, and Enter and Space with them.
The Fade grounded components preference dims every line that is joined, through other lines, to the edge of the grid: such a line can never be part of a loop.
On harder boards you will often know that two squares side by side slant the same way before you know which way that is. Notes lets you write it down. Press the Marks key on the keypad, or P, to turn notes mode on, the same way as in every other puzzle that takes notes; a small pencil shows at the top right of the board while it is on, and pressing it again turns it off.
In notes mode, tap near the side two squares share to put a same-slant mark (=) across it, and tap there again to take it off; drag on across more sides to do the same to each. From the keyboard, press Enter on one square, move to the square beside it and press Enter again; Escape lets go of the first square.
Two squares side by side slant the same way exactly when just one of them touches each end of the side between them. So they do when a clue at that end needs exactly one more line and they are the only empty squares around it, and when they can neither both touch one end nor both touch the other: a 1 there stops them both touching it, a 3 stops them both slanting away from it.
Lines of 2s. When 2s lie in a straight line with a 1 at each end, every pair of squares along the line, one on each side of it, slants the same way; so does every pair when there is a 3 at each end. A 2’s four squares are the pairs on either side of it, and the 1s (or the 3s) leave each pair giving it exactly one of its two lines. A diagonal can stand in for either end: one meeting the end 2 does what a 1 does, and one missing it does what a 3 does.
A mark that turns out to be wrong shows in red when you check the board.
The hint carries on from the diagonals you have drawn and your marks, as long as none of them is wrong. When a step needs to know that two squares slant the same way, it puts the mark down first, as a step of its own that says why.
These parameters are available from the ‘Custom type…’ option on the ‘Type’ menu.