Change tokenId encoding to use small-endian bits

[ilikesymmetry]

Before:

toTokenId("base") -> 44498839750228103637605521814503088468024911857695849731344316461305001148416
toTokenId("basecamp") -> 44498839760694099857589941137367510834861780992994609044330424891558904987648

After:

toTokenId("base") -> 62617365
toTokenId("basecamp") -> 6261736563616D70

This improves devx by:

1. making the strings more easily human readable
2. making the strings more easily human differentiable (notice the clarity in difference between the two codes)
3. reduce clutter on NFT rendering sites that will likely use tokenIds