Due to the current status of EOS, the EarnBet team (previously EosBet) has announced that they will leave the EOS mainnet if the situation is not resolved within 30 days.
The extreme cost of resources in the mainnet was caused by the EIDOS mining mechanism, which dramatically increased the total number of transactions executed on the mainnet in the last 4 weeks. At the same time, however, all these transactions limit all other dApps that are no longer able to maintain an adequate level of User Experience for their users.
EarnBet found itself in exactly this situation, once they had pushed hard on marketing and rebranding their platform they had been able to attract many more users, but given the high cost of resources at the moment they can only support 3-4 bettors online at a time.
"We're finally in a place where we can reap real returns from our marketing efforts and significantly increase our user base. But instead, our marketing programs must grind to a halt for fear of too many players."
Unfortunately in this situation they are not the only ones, there are other dApps that are thinking of leaving the mainnet, including EDNA and probably KARMA(?). Currently also some Reward Proxies are having trouble paying resources to distribute rewards to their voters, including Colin Talks Crypto's Reward Proxy.
According to the EarnBet team, Block Producers should block the transactions used to mine EIDOS tokens (considered "spam"), to favour dApps. For this reason they consider it a problem of Governance and not of algorithm/resource efficiency.
"Our goal has always been to be the first dApp to reach mainstream adoption. In order to do this, we need to be able to sustain over 10 bets per second. This is currently not possible on EOS."
However, we must point out that this situation is mainly caused by the high number of transactions that are executed, regardless of whether they are carried out to mine EIDOS or not. If we were to find ourselves in a similar situation with transactions resulting from a multitude of dApps (mass adoption?), we would probably end up in the same scenario, and in that case the Block producers could not decide to block the transactions of some dApps to favor others. So even blocking EIDOS transactions now would only be a short-term solution.
To date we know that Dan Larimer with Block.One are working on a solution to improve the CPU approach and bring improvements to REX. In the next days Dan should propose in a blog post a new approach to the CPU that should drastically reduce the cost of the CPU. Could this be the solution we are looking for?