Re: [ Hyperledger ] Proposal - BitXHub: inter-blockchain technology platform
Hi BitXHub Team,
Thank you for your responses. It’s always nice to see more people interested in blockchain integration.
At Fujitsu, we have also built a blockchain interoperability platform. It’s quite similar to yours, actually—much more so than the current, main BIF platform which was contributed by Accenture. We use a “middleman blockchain” which you refer to as a relay-chain in order to manage interchain transactions in a way that is probably very similar to what you are doing (although I haven’t dug deeply into your code, so I could be wrong). You can view it here:
As you can see, we have contributed this to the BIF lab and made the decision to work together with the Accenture folks (and any others that care to join) on blockchain interoperability. Despite some core architectural differences, we think there are a lot of core functionalities that will be common to both our (and your) “middleman blockchain” approach and Accenture’s “overlay” approach. Ideally, we could come up with a set of modular components that allow anyone to set up a blockchain integration platform that meets their specifications. Maybe this is too ambitious, but we think that it is worth trying.
I’d also like to respond to your comments on the BIF. I certainly don’t think it is necessarily deficient, but just that different design decisions have been made.
>>> Each blockchain must implement various verification logics of heterogeneous interchain transactions, which is too complex for a participant ledger to develop.
The alternative is what you are seemingly doing (please correct me if I’m wrong): build a language for contracts that runs on top of all blockchains. From my perspective, this will require a separate shim for every different blockchain, meaning that it will also be complicated to develop. Is this the correct assumption? I do agree with you that some sort of universal verification language is probably the best way to go, and I think this is where the BIF project is headed (although I can’t speak for everyone or for certain). You may want to ask some of the core BIF guys like Peter and Jonathan what they think about this. But the fact that everyone (and all code involved) in a transaction across multiple chains needs to know the semantics of transactions in all of the chains involved is something that is certainly clear.
>>> No middleman leads to a lack of track for the execution status of inter-Hi chain transactions, which makes it difficult to guarantee the transactional consistency of inter-chain transactions.
A middleman can often be viewed as a drawback. Indeed, I think Accenture’s overlay network requires less trust overall than ConnectionChain or your solution since you don’t have to trust an additional middleman network, which may be much smaller or less trustworthy than the networks it is connecting. The overlay network certainly has issues—it can’t work with cryptocurrencies or, more generally, proof of work based systems. But it has definite benefits for connecting large, permissioned networks.
>>> The receiver's verification of the legality of interchain transactions is too simple (just verify signatures).
It’s fine for an overlay network, where the overlay nodes will have to be familiar with transaction execution on both blockchains involved in a transaction. I guess your opinion here depends on whether you want the client or the protocol to be responsible for ensuring the transaction semantics are correct across blockchains.
>>> We propose an efficient and pluggable validation engine in relay-chain s, every app-chain only needs to write its own verification logic and send it to the relay-chain, thus the legality of interchain transactions can be verified uniformly and efficiently.
This is something I’d definitely be interested in seeing in more detail. I can’t view your whitepaper at work (a lot of .cn sites are blocked unnecessarily, including some universities and typically even the Asiacrypt submission website (!!) which causes me a lot of grief) so maybe you address that in there. If it is expressive enough to handle most transactions and easy to implement on a lot of different blockchains, it would be a very nice contribution.
>>> The relay-chain plays a role as a decentralized trusted middleman, and other blockchains interact through the relay-chain with interchain gateway. Moreover, relay-chains track the execution status of interchain transactions to ensure transactional consistency of interchain transactions, which greatly reduces costs of communication and trust.
Again, this is very similar to the ConnectionChain design. You may want to take a look.
Thanks again for your time, your interest in this topic, and for coming to Hyperledger with a proposal. Please let me know what you think about what I’ve said here—I’d love to hear your feedback.
Thanks a lot for your time, and have a great day.
From: tsc@... [mailto:tsc@...] On Behalf Of "???
Sent: Thursday, February 6, 2020 6:14 AM
To: Arnaud Le Hors <lehors@...>; tracy.a.kuhrt <tracy.a.kuhrt@...>
Cc: tsc <tsc@...>
Subject: Re: [Hyperledger TSC] [ Hyperledger ] Proposal - BitXHub: inter-blockchain technology platform
Thanks for reminding us of this matter, we're interested on the workflow and principle of BIF, and make a comparison between BIF and our BitXHub.
Actually we consider that BIF is still deficient in some ways:
On the contrary, BitXHub adopts the architecture (relay-chain + interchain gateway) which is quite different from BIT:
We're waiting for more feedbacks and suggestions after your reading of our proposal!
------------------ Original ------------------
Date: Thu, Feb 6, 2020 01:23 AM
Subject: RE: [Hyperledger TSC] [ Hyperledger ] Proposal - BitXHub: inter-blockchain technology platform
I think Tracy is touching on a very important point. While it's exciting to see so many people interested in tackling the difficult and important problem of integration
across blockchain platforms we ought to try and converge efforts where it makes sense. This means that we ought to understand how different proposals compare to one another and determine whether they warrant being pursued concurrently rather than through joining
Hi, Shutian Bao.
Thank you for the proposal. This proposal is similar to the work that is ongoing in the Blockchain Integration Framework (BIF) lab. I know that the BIF community is discussing different options for interoperability during their bi-weekly calls. Have you had an opportunity to sync up with that community to see if this work might fit with what they are doing?
<tsc@...> on behalf of
This message is from an EXTERNAL SENDER - be CAUTIOUS, particularly with links and attachments.
We'd like to propose a universal inter-blockchain technology platform 'BitXHub'. Our current main efforts are focused on improvement of flexibility, high availability and scalability. If we could get some feedback on our proposed design from members involved in Hyperledger TSC, it'll be quite useful for further improvement of BitXHub.
This enhancement for Hyperledger is aiming to:
- Provide interactions among ledger projects (e.g., Fabric, Sawtooth and Iroha) in Hyperledger community when applied in different scenarios.
- Support interchain transactions including asset exchange, information sharing and service complementation.
- Our proposal
- Our whitepaper