Just pushed a new version of the iguana dPoW code that is a lot more efficient and should be fast enough to handle all 64 notary nodes. We are almost to 20 testnet notary nodes, but we still need to get them all participating in the group dPoW tx.
I also simplified the notaryinit process a bit and pruned out some unreliable test nodes.
Assuming we get a steady stream of notarized data, my next tasks are the half dozen remaining dPoW items:
a) updating latest notarized height based on the notarized tx data
b) prevent overwriting blocks below notarized height
c) detection of special transactions to update list of current notaries
d) award 5% APR for utxo older than a week when they are spent
e) round robin mining difficulty
between supporting the notary operators and GUI devs, not sure I will be able to get all 5 done before ICO start, but once the notary nodes are generating the group tx, the biggest amount of new code is complete.
I want to minimize the changes needed inside the komodo to minimize the chance of bugs. By splitting the changes in separate modules, we can start testing and qualifying it independently. The process of generating notarized data is quite independent of the main komodo code, especially as things need to work even if there are no notary nodes.
The trick will be to seamlessly start using the notarized data if and only if it is available.
Strictly speaking c), d) and e) are not needed for mainnet launch. However they would all require hardforks to update, so my preference is to get them done before mainnet genesis.