Hello, @gorbak25 it’s more than a month since the last update. Can you share more about the progress of hyperchain?
We are looking forward to more details about hyperchains. White paper? Popular science articles? Video? Update FQA?
Is there a better place to get these updates? I know the developers probably aren’t hanging around on these forums.
@erik.chain maybe you can help with some update? Anything is better than silence from @gorbak25 since people usually tend to get nervous if there is no response for quite some time
Even line: “don’t bother me, there is no update now, I will keep you posted when I have something to say”, would be better than silence.
Hi @Amadeuss, as far as I know @gorbak25 was working on an update that should be released soon
And @hadees.chain here is the place to get the updates, though we will soon also have public update calls again to give you regular updates on the development as well
Hi @erik.chain , I knew that you would be the right person to give some quick feedback, I appreciate your swift response
Looking forward for future development
Update #17 [09 Feb - 20 April]
Time for a long, long, long overdue update
Behind the scenes a lot was going on, we wrote a lot of documentation, we prepared a lot of demos, we got the private HC testnet up and running, we helped with the IRIS HF and we have a clear roadmap for the near future. This update is deliberately not complete to not clog this thread - some new threads will appear in the forum in the next days.
HC Private testnet
The long awaited private HC testnet is here The MVP of the HC protocol is here and operates nominally - we’re testing everything using 3 BTC regtest nodes as the parent chain and 3 HC validator nodes as the Hyperchain. We decided to keep this private as long as we’re still fidling with the protocol and our docker scripts.
HC Roadmap
We prepared a clear way forward - we will post it in a separate thread in the forum
Team expansiton
We’re hiring and expanding our dev resources. Right now I’m happy to announce that @cytadela8 and @wj2016 joined our team
HC Protocol
With the Proof Of Concept(Private HC Testnet) ready we went to the drawing boards and ironed out a lot of details about the protocol:
- Compared to the protocol in the V1 Whitepaper it’s much more secure and cheaper than before - we found a nice difficulty formula which ensures that HC has a lot of nice security properties
- The HC protocol was hardened against “shadow forks”
- We evaluated the economic feasibility of the HC Protocol
- We found a way to make delegated staking safe and sound
- We designed a way to easily migrate AE mainnet to the HC protocol via an onchain protocol enforced vote
- etc…
HC Article
We prepared a detailed marketing friendly writeup of the HC protocol - it’s the preliminary step before preparing the V2 of the whitepaper. This was the main reason why this update took so long to publish - I’ve planned to publish everything at ONCE, or at least the Article and Roadmap alongside the update
IRIS HF
We helped with the Iris HF - We developed contract factories, cloning and bytecode introspection in the FATE VM and introduced syntax support for it in Sophia - Contract factories and bytecode introspection by radrow · Pull Request #305 · aeternity/aesophia · GitHub
Some performance tweaks in the node were made.
Erlscripten
Amazing progress here - right now one can just npm install erlps-aesophia
to get a client side sophia compiler in JS :3 Some small JS compatibility layer is still missing but this will be done after the IRIS HF. Here is an SDK with a client side sophia compiler: [DON'T MERGE] Client side Sophia compiler by gorbak25 · Pull Request #1117 · aeternity/aepp-sdk-js · GitHub
While working on erlps we also found some rare bugs in the sophia compiler.
We reworked the IR of the purescript compiler to improve the compiler performance, and we also started some discussions on their pattern matching mechanism: Proposal: Replace data structures with tagged arrays · Issue #4020 · purescript/purescript · GitHub
Stay tuned
thanks for the update and your amazing work! can’t wait to actually test hyperchains
I noticed that the bundle size currently is around ~7mb according to the analysis of @davidyuk. is it possible and planned to shrink that to a size that is suitable for common web applications?
really great to see progress on that side. this will enable us to implement a “feature complete” mobile client for state channels, right?
so no more PoW for AE?
In the long term probably yes.
- AEVM support can be removed from the bundle
- Tree shaking can be done
- We can remove a lot of unused stuff from aebytecode(aeb_generate_fa_ops)
Generally I think it will be hard to get the bundle size down to <1MB. ~2MB might be feasible. The way forward is to get the SDK working in one of 3 modes:
- Full dynamically downloaded compiler with everything
- By default a slim calldata encoder/decoder using erlps
- Legacy/Deprecated HTTP compiler
does hyperchain block is immutable?
Hyperchains blocks are always immutable. Keep in mind that like in any blockchain an immutable block might still be rolled back due to an 51% attack.
But still - that would be 51% attack on the parent chain, right? To the best of my understanding - HyperChains are a hybrid and thus immune to traditional long-range attacks that could happen on PoS networks…
HC are immune to long range attacks
More than that I’d love to see a CDN we can all trust. 7mb isn’t bad if only the first app has to download it locally. Maybe something tied into the chain?
In theory though you could use multiple chains and thats where I think it gets really interesting because the attack has to be on more than one chain.
so no more double spend attack at exchange?
When will hyperchain be released?
Have a look at the latest roadmap update here:
You can already try out an early version of the private testnet