14:32
Ok so, what I can do is either verify the ByteCode
and leave it as a mark as
being verified or I can alternatively create some extra byte code information.
I think at this point, there would be duplicated code between the interpreter,
the verifier, and anything which would verify the method. So basically, there
is a class which can process the instructions based on an input instruction and
keep everything processed and handle the stack.
14:35
So the question is, does the handler keep state information or does it just run through the instructions. I think it would be best if it kept track of everything. It would not verify anything, but if execution ever does fail then it would just verify correctly. It would check things naturally and then would call elsewhere. So it would be an interpreter for the most part. The verifier and compilers though, would perform state resets and not handle looping at all. So basically it would execute the code, but would rely on an implementation to actually execute those instructions.
14:42
Probably at this point they can use the same class structures and such but not the verified methods. This is due to the method verifier needing those things.
17:39
The instruction parser can also do some common things. It can for the most part hide the somewhat stack nature of the code by doing copy operations potentially automatically. However, that might cause some issues. Well not really as long as those copies are known to wherever copies are stored and such.