That is a fair point. I don't think a protocol would necessarily replace photographs of puzzles, just allow for an alternative when a puzzle cannot be described with a photograph. As well I could not imagine the task of converting the puzzle library to anything but what it is now strictly from a scope of work perspective, so wouldn't suggest any necessary changes to existing puzzles.
With that said, the depiction of that puzzle digitally in a playable form would be more up to whoever decided to make a tool for playing such puzzles. Which, to your point I believe would be an interesting task.
To standardize the storage of data, all that would need to happen is to make it broad enough to encompass anything that can be drawn (Which you touched on). That puzzle is still just points, lines, and characters, and I imagine the data would be very similar to a vector graphic simply depicting what goes where in space (even if part of a theoretically infinite set).
I do admit however, (and this is a big however) that a universal way to order puzzle data in a file is made less universal by the need to define which tools are capable of parsing each type of puzzle. It sort of defeats the point in theory, but functionally may not be so much of an issue, I'm not sure.
I guess my thoughts are these: Allowing pictures only restricts the creativity of puzzles that can only be expressed on a computer. Allowing computer puzzles only would restrict (At least functionally) the creativity of puzzles like the example you posted. Whether or not to hybridize in in some form is just a cost benefit analysis, coupled with the same philosophical discussion I promised not to get tangled in earlier.
And I will also not weigh in on cost or benefit, in part because the benefit (although subjective) seems well enough defined, the cost is more intangible unless there are further ideas for solutions.
TL,DR: My suggestion was not perfect, and may not be worth considering further.
With that said, the depiction of that puzzle digitally in a playable form would be more up to whoever decided to make a tool for playing such puzzles. Which, to your point I believe would be an interesting task.
To standardize the storage of data, all that would need to happen is to make it broad enough to encompass anything that can be drawn (Which you touched on). That puzzle is still just points, lines, and characters, and I imagine the data would be very similar to a vector graphic simply depicting what goes where in space (even if part of a theoretically infinite set).
I do admit however, (and this is a big however) that a universal way to order puzzle data in a file is made less universal by the need to define which tools are capable of parsing each type of puzzle. It sort of defeats the point in theory, but functionally may not be so much of an issue, I'm not sure.
I guess my thoughts are these: Allowing pictures only restricts the creativity of puzzles that can only be expressed on a computer. Allowing computer puzzles only would restrict (At least functionally) the creativity of puzzles like the example you posted. Whether or not to hybridize in in some form is just a cost benefit analysis, coupled with the same philosophical discussion I promised not to get tangled in earlier.
And I will also not weigh in on cost or benefit, in part because the benefit (although subjective) seems well enough defined, the cost is more intangible unless there are further ideas for solutions.
TL,DR: My suggestion was not perfect, and may not be worth considering further.