Welcome to the Elastos Blog
Read it here first
Elastos 4th Anniversary WeChat AMA | Shunan Yu, Head of the Gelaxy Team
On August 30, Shunan Yu, Head of the Gelaxy Team, was invited to share their work summary and strategy including improvements to the main chain’s consensus, the new DID sidechain, the oracle machine and the Lightning Network with the Chinese community.
The following is the translated transcript:
Hello, everyone. I am Shunan Yu, the Head of the Elastos Gelaxy team. Our team is mainly responsible for the development of the Elastos blockchain. Today, I would like to share with you the development progress of our team during this time.
In terms of the main chain, we introduced the Schnorr signature technology, which has been accepted by the BTC community. This technology can effectively improve the privacy of multi-signature transactions and compress the signature space on BTC. We also introduced the Schnorr signature to the ELA main chain, which not only realizes multi-signature based on the Schnorr technology, the same as BTC, but also compresses the space of cross-chain signatures and space for main chain and side chain DPOS consensus signature based on this technology. The first part will be launched sometime after September, and the cross-chain signature and DPOS consensus are in the process. Further discussions and scheme authentications are required.
At the same time, the main chain has the ability to degrade to POW in extreme circumstances and then recover to POW+ DPOS. There are a total of 36 nodes in the DPOS consensus, 2/3 of which can participate in the consensus and output blocks normally. If more than 2/3 of the nodes fail to participate in the consensus, it may lead to the inability to output blocks. To cope with this situation, we have prepared the above backup plan. The purpose of returning to the POW mode is to leave time for resume running of DPOS nodes. After resuming normal operation, these nodes can send the Ready request, then the DPOS consensus can be recovered.
In addition, we have achieved the random backup node function. Previously, the first 24 nodes participated in the consensus. To encourage backup nodes to participate in the consensus, the first 23 nodes are generated by voting, one node is generated randomly, which can also be involved in DPOS consensus and enjoy the same benefit as the other 23 nodes. This function has already been developed and will be launched after we perfect it, as soon as possible.
Improvement in ELA production reduction. We have reduced the production based on the CR proposal and started to make an additional issuance with 20 million as the constant base since June 22 of this year. By referring to the Bitcoin model, the ratio of the additional issuance is halved every four years.
Change of CRC consensus. Among the 36 super nodes, 12 nodes are reserved for the CR committee members, and the elected committee member will obtain a node automatically. It was maintained by our team before, but now it is open to be claimed. Each elected CR committee member claims one node for maintenance, which will in the future promote decentralization.
In terms of the proposal, we have made a new sidechain proposal. The use and operation of the sidechain relies on the new sidechain proposal, which includes the identification information of the sidechain and a sidechain for a unique identification to ensure that all nodes that operate the sidechain participate in the same chain. That is to say, a new sidechain has to be voted on by the entire community to be put into service.
In terms of cross-chain, the function of small-sum quick accounting based on the arbitrator’s signature is realized, and the accounting time can be controlled within 10 seconds. The confirmation of the current cross-chain requires a long time. This function only supports small-sum quick accounting, and the principle is based on the multi-signature scheme of the arbitrator so that the small amount can be quickly transferred to the sidechain and the main chain does not need to be an entry block. This can be controlled within 10 seconds, and vice versa. Furthermore, we have added the two-way automatic reversal transaction to avoid the transaction failure in the receiving end but the transmitting end has completed the transaction due to abnormalities.
In terms of DID side chain, we have already got a DID sidechain, and now we have developed a DID sidechain based on the Ethereum sidechain. The new sidechain shares the same DPOS consensus with Ethereum. Compared with the old ID chain, the new chain has better block output speed, and the time of block output is 5 seconds. It naturally supports intelligent contracts, and a browser has been developed for this chain. In addition, the relationship between these two chains is that the old chain is still operating, yet the ID data has been transplanted on the new chain, and the assets have to be transferred by the user. Next, the transfer function of this chain will be canceled.
For the oracle machine, the Chainlink oracle machine has been transplanted based on the DPOS mechanism on the ESC side chain, and some changes have been made in the credit mechanism of the oracle machine, which mainly changed the arbitrator “credible” to the DPOS arbitrator mode. Then the information of the DPOS arbitrator is transferred to the sidechain in a cross-chain manner, and the Chainlink oracle machine on the sidechain can use the information as the credible basis. The oracle machine is currently under verification.
In order to get through the ecology of Ethereum and ELA, we are attempting to combine the ELA sidechain with the Layer2 of Ethereum by referring to the idea of Layer2 of Ethereum, thereby achieving the flow of ELA from main chain to side chain and to Ethereum as well as the flow of ETH from Ethereum to ELA sidechain. In this model, the sidechain still takes ELA as the basic token, which means that ELA is used in gas fees. ETH exists in the form of a token contract, and there will be token contracts corresponding to ELA on Ethereum.
We plan to complete our goal step by step. The first step is to achieve a cross-chain flow between ELA and ETH. Here, we refer to the DPOS consensus mechanism of the ELA main chain and consider the arbitrator generated by the main chain as the cross-chain arbitrator between ELA sidechain and Ethereum, thus avoiding repeated selection of arbitrators on Ethereum. The next round of the arbitrator’s information will be sent to Ethereum through the last round of the arbitrator’s signature to ensure information reliability. The cross-chain transaction between Ethereum Layer1 and Layer2 will be guaranteed by these arbitrators’ signatures. This part is scheduled for completion.
The second step is to refer to the optimistic convolution or the zero-knowledge proof schemes. As our goal is not exactly the same as that of Layer2 of Ethereum, we have to take into account the cross-chain ability of ELA sidechain and ELA main chain, which means that two Layer1 share one Layer2, thus it will be more complicated. Therefore, we can only refer to some implementation principles to help achieve the cross-chain ability of complete contract calculation.
In the aspect of the Lightning Network, the Lightning Network on BTC has better development. The office-train transaction processing ability of BTC can be greatly improved, the transaction speed can be improved and the transaction costs can be saved through the Lightning Network. Our team is now working on the preliminary research in this regard. We hope to integrate with the ELA wallet and find a way to integrate with ELA. On the other hand, it is also our hope that ELA transaction processing ability can be improved.
Currently, we are focusing on the wallet-end integration as well as the access of the node-end to multiple accounts. We learned that the operation of the Lightning Network wallet requires a one-to-one corresponding node. A Lightning Network node can only have one account for channel construction, which is inconvenient for users. For this reason, a solution is to place the node at the mobile end, and the mobile end directly participates in the Lightning Network. This direct decentralized participation mode of the mobile end does not achieve an ideal effect compared with the application effect of our previous decentralized wallet, and it is difficult to improve its usability. Another solution is to provide multiple mobile ends and connect to the back end. This solution has two modes, i.e. escrow account and non-escrow account. Our priority is given to explore the non-escrow account.
That’s all for my presentation of our team’s recent work and direction. Thank you very much!