The first week is up!
I can say that I'm very happy about the results so far. I've learned a ton about the internal workings of the Ethereum protocol in general – not only the EVM. I've had a chance to look at client software, to read the yellow paper, and to brainstorm my way through some of the more important parts of the execution model.
Enough talk, let's get to what I did today: it was finally time to implement the Memory structure! I had a draft model of this which I did a few days ago (on Day 4 to be precise). It turns out what I did was way too simple and bugged, so I had to basically rewrite it completely.
I learned how to manipulate Buffers in javascript - and I'm surprised as to how flexible and reliable they are. I also learned how MSTORE and MLOAD work. In particular I created a generic write function in the memory, which takes a size argument that can be either 1 or 32 depending on how many bytes I want to write into memory.
I didn't know how the memory layout actually works, and how it expands dynamically based on where someone is trying to access it. Now I know why accessing memory which is "far away" is more expensive: you pay for each 32-byte word expansion that happens to get there.
MLOAD also gave me some trouble because I couldn't figure out how to read a single byte on the tail (with something like a PUSH1 31 hex and then MLOAD). Nonetheless, solving it has been rewarding as always, thanks to the tests provided by this challenge.