From Komodo Slack on march 24, 2018 one of the first high level view of the Smart Contracts system for KMD
…the “smartness” logic for each “contract” will be embedded into a dedicated blockchain, I guess a smartchain. So only people who have that chain will bother with tx for that chain
Each chain’s code will run native executable code, i.e. not even interpreted and not some abstracted script virtual machine, i.e. 100x to 1000x faster.
Oh, and we have a way to do things cross chain, so any chain can rely on any other chain’s results, i.e. the sum of all these smartchains can be combined into a large network. So instead of having a single giant network and having to figure out how to shard it and still have things work running very slow interpreted virtual machine, we go the other way: we make each contract as fast as possible, already isolated and then allow them to be combined as needed. If not needed to be combined, it won’t be, and this is on a tx by tx basis.
For those worried about usage, don’t! I am worried now more about what to do when KMD blocks are full… Even with all this isolation, everything always goes back to KMD and with just hundreds of smartchains that will generate thousands of dPoW notarizations per day and who knows how many cross chain contract tx per day… Value of a currency is proportional to its network effect (usage). Already with BarterDEX just in test phase our usage is above XMR and about the same as ZEC.
We will write a paper on all the 2.0 tech after its working, these things are impossible to know ahead of time, at least for me as blockchains are tricky things to get working. So all these whitepapers that claim they will make something really complex working have coders that are far better than me… or maybe the reality of what they end up with will be significantly different than what the whitepapers describe.