Serialization adjustments

[davxy]

• Fix serialization order in eq 283 (work report) to be coherent with eq 117:

  • 117: (s, x, c, a, o, r)
  • 283: WAS (a, c, o, x, s, r)

  The serialization order of items in the other types is consistent with the definitions of those types, so this may be a leftover

• Fix naming for work items entry in eq 284 ($x_i -> x_w$) to be coherent with eq 174

• Adjust naming in eq 277 of dispute verdicts to be coherent with naming used in eq 97/98 I.e. Dispute extrinsic is now $E_D ≡ (v,c,f)$ and $(r,a,j) ∈ v$

• Encode verdict age as $E_4$

---

• Not related to serialization: Add hashing algorithm argument to MMR accumulate call

[basedafdev]

As mentioned by @vekexasia earlier in [https://github.com/w3f/jamtestvectors/pull/12], for the Assurances Extrinsic, can we also add explicit $E_2$ for the validator index?

[davxy]

If this statement (GP 0.3.6 - just below eq. 272) is and remains true:

Note that at present this is utilized only in encoding the length prefix of variable-length sequences.

perhaps adopting a "catch all" rule might be easier.

Something like: $x \in \mathbb{N}_{2^k}$ encoded as $E_{k/8}$

[gavofyork]

That doesn't work in the axiomatic set theory in which GP is written since $\mathbb{N}{2^8} \subset \mathbb{N}{2^{16}}$, thereby making encoded integers variable length without introducing a discriminator.

To make it work we'd need to introduce a form of typing, which I'd prefer not to do as it would severely complicate the the base language of the GP.