Today Chia COO and President Gene Hoffman along with technical staff from Chia Networks gave a detailed overview for how the upcoming pooling protocol works for the pool developers interested in starting pools.
The presentation was extremely interesting, with a number of guiding design principles that I agree with wholeheartedly. The primary ones being both farmer and pool security, not just from a key perspective but an integrity perspective as well (something that Hpool has struggled with since inception). Although, the allowances for pool switching and running your own full node will significantly improve decentralization and allow make pool lock-in significantly less problematic. And the item I hadn’t even though of, “configless blockchain backup”. This will mean that you don’t need to setup your pooling again when configuring new harvesters. That is a really nice feature, and I’m almost positive it will further strengthen the integrity of the pooling protocol making sure nothing is client side and manipulatable.
They also explained exactly why the portable plots will be different, as they will no longer use a pool public key but instead will use a smart coin (singleton) that contains the address of the pool, the user public key and some other metadata for the pool that you are currently working with. This coin would be “spent” and recreated every time a pool is joined or changed, which is why you will need at least 1 mojo in your wallet to join a pool. You can ensure that by visiting the official Chia faucet here. There were also some technical details about how the payout process works. It seems well thought out to ensure no double spend or taking pooled money. I am not a cryptographer but the theory seems sound to me.
They then proceeded on some technical details about how the pool requests are executed. If you need that information you should really go watch the presentation.
The GUI demo followed, and that was very interesting. Due to the ongoing legal difficulties I am facing with Chia Networks I will not be putting any of their material on this site until that is resolved, but suffice to say it looks fairly polished at this point. It will allow for easy creation or join of a pool and seems to be fairly self contained without the need for complicated command line code and complex worker setups and the risk of mining for nothing like other cryptocurrencies.
There was an explanation about how the payout structure will be done, which is complicated and will likely require a post of its own. In a nutshell, users will be awarded points based on the difficulty of the proofs they submit to the network, and that total number of “difficulty points” will be used to determine pool payouts. This seems designed to ensure an equitable payout, but more analysis will be required to see if it achieves that goal.
The chia plots create command will also be seeing an update to support the new functionality, allowing you to use the singleton. We will detail that functionality when it arrives.
Afterwards we had a discussion with Bram and Gene and some question and answer period with a number of really good questions, about how the pools would discover their coins and how they plan on safeguarding against pools getting too big and misbehaving. I wont claim to be an expert here, but Bram’s answers seemed very well thought and the point about how having farmers sign the blocks and not the pools is extremely important and should very seriously reduce the likelihood of bad pool behaviour. Near the end we got the release schedule, which isn’t. It was a commitment to have things ready and a rollout schedule for when they are. And a commitment to not break any more commitments.
The session was very good and worth a watch if you are interested. I highly recommend anyone serious about Chia and pooled farming listen in, even if it was directed at pool developers.