This one was a little bit different from the other All about Pools articles, since this is not about a set of farmers who decided to start a pool but about an established mining pool getting excited about Chia and deciding to become farmers. Flexpool.io is a Canadian Ethereum mining pool with just shy of 15 000 miners. When their miners and mods started to talk about Chia nonstop they decided it was a coin with a good potential and were going to branch out into pooling for Chia. These folk are on another level in a few ways, as you will see.
Right out of the gate in talking to them, they are very sure about both the longevity of Chia and the need to get good, reliable pools in place without massive leakage from farmers. Flexpool estimates (and I don’t disagree) that the current off-chain pools are losing as much as 20% of their block rewards to dishonest members. Not only do they expect to completely sidestep these issues with the on-chain pooling protocol, they also expect the reward differentials to draw in the members from the existing pools. In effect, they do not think Hpool and Core-Pool will be able to compete.
Because they have some experience here, I spoke to them at some length about the Chia Network involvement, and they think its been great. Maybe even too great. They are able to ask a question on Sunday and get an answer in a few hours, which may not be sustainable forever. From their perspective Chia Network is developing a reference spec that the pools are then responsible for building out their own software to, but some pools will just be implementing that reference with a few minor changes. Flexpool will not just be deploying the reference software and calling it a day, having developed their own toolset in golang, and they think that the pools that do just try and use the reference will have a number of issues.
Expanding on this, they do think the biggest issues are that the entire stack is written in python. This has implications as it relates to scaling out to the size they are talking and are considering a complete rewrite of the full_node in golang to help it scale up to their needs. I have my own thoughts on the architecture of the Chia distributed network, and will be sharing those soon.
Flexpool, having some experience running major pools, are a lot less confident that by simply using the on-chain pooling protocol that this will be perfectly smooth, even if the issues differ from those facing Hpool and Core-Pool. There are always issues, they say, and even the biggest pools in other cryptocurrencies have downtime and problems crop up. This is an interesting, and frankly more honest, take than I am hearing from the other pools.
When it comes to the legal and financial questions, Flexpool has a year of operation as an Ethereum pool with 15000 miners to fall back on. They have experience handling millions of dollars and they have already done their due diligence in setting up their corporation properly with the correct controls to operate legitimately. They have a dire warning for the pool operators that ignore this step and they worry a lot of innocent Chia farmers might get caught up in that kind of nonsense. I share these views, and this reinforces my prediction that the Chia pools will live or die based on their ability to navigate their own particular legal quagmires. They want people to be careful and not jump into “too good to be true” situations that end up costing.
My conversations with Flexpool suggest a company very much ready for Chia pools. They don’t want the dev team rushing anything because they know from experience that there be problems either way, so minimizing them up front is the best strategy. I didn’t see any red flags here, and I think between their established pool presence and their involvement with the Chia community on Keybase and Reddit that they are in a good position to be the pool to beat.
