3 ms·
But if you're writing an interpreter, you can serialize the result of 'evaluating' the structure of your 'function' made out of components that are being interp
by kian 7y ago
But if you're writing an interpreter, you can serialize the result of 'evaluating' the structure of your 'function' made out of components that are being interpreted by your code, thus turning your 'function' (when evaluated with one set of code) into a data structure (when evaluated by something that serializes each of the components making up your function)
- lmm 7y agoAt that point you've already evaluated it and can't recover the original program - you're serializing the result of interpretation, not the computation to be interpreted. Since that computation might contain arbitrary functions, the only way to serialize that would be if you had a way of serializing arbitrary functions.