Sudoku Example

[Acentelles]

The goal of this application is to demo how Taiga works. Pending work:

• Define Token VP
• Define Intent VP
• Define Sudoku App VP
• Reward design and implementation
• Write documentation

[XuyangSong]

doc: https://hackmd.io/@yulia/sudoku

Circuit details:

• Dealer Intent App VP

1. check app_data = hash(encoded_puzzle || sudoku_app_vk)
2. if it is an output note, there are no constraints
3. if it is a spend note, check that there is a puzzle note with zero value, the state of the puzzle note is equal to encoded_puzzle, and sudoku_app_vk is equal to the app_vk of the puzzle note.

• Sudoku App VP

1. check the encoding of the current_state
2. check app_data = hash(init_state || current_state)
3. check current_state is a valid sudoku
4. if it is the init sudoku note(output note), check init_state = current_state
5. if it is an output note(the puzzle state is updated from the spend note), check spend_note_init_state = output_note_init_note and spend_note_vk = output_note_vk
6. if it is a spend note, the output_note_current_state is a better solution than the spend_note_current_state. If the output_note_current_state is the final solution, check the value of output_note is zero; else, check the value of output_note is one.

Merged in #127

[XuyangSong]

Some minor TODO lists

• Sudoku state encoding constraints in application vp
• Add the constraints for non_zero_sudoku_cells assignment in check_puzzle
• Consider if we need to make the app_data_static of the application vp unique. A potential solution is to embed the spend note rho into app_data_static when creating the note.