Mina Protocol Community Spotlight Townhall Events, 30 Jun 2021

Previous Meeting

⏯

youtube image

►

From YouTube: Engineering & Network Updates | Mina Protocol Spotlight Townhall

Description

Learn about Mina's engineering and network updates from ecosystem partners at O(1) Labs and ChainSafe.

🔦 Featuring 🔦
Brandon Kase - First Engineer at O(1) Labs
David Ansermino - VPE at ChainSafe

➡️ Chainsafe: https://chainsafe.io/
➡️ O(1) Labs: https://www.o1labs.org/

What is Mina?

Mina is the world’s lightest blockchain, powered by participants. Rather than apply brute computing force, Mina uses advanced cryptography and recursive zk-SNARKs to design an entire blockchain that is and always will be about 22kb, the size of a couple of tweets, ushering in a new era of blockchain accessibility. With its unique privacy features and ability to connect to any website, Mina is building a private gateway between the real world and crypto—and the secure, democratic future we all deserve. Mina is stewarded by the Mina Foundation, a non-profit organization.

Follow Us!

🐤 Twitter: https://bit.ly/MinaProtocolTwitter
💬 Telegram: https://bit.ly/MinaProtocolTelegram
📢 Discord: https://bit.ly/MinaDiscord
📧 Monthly Newsletter: https://bit.ly/MinaNewsletter

#Blockchain #ZeroKnowledge #Rust #Wasm #Crypto #Decentralization #Scalability #CryptoCommunities #MinaProtocol #EngineeringUpdates #NetworkUpdates

Mina Protocol – the world's lightest blockchain

A

Hey everyone: um my name is david ansemino and I am vp of engineering at chainsafe systems. uh We are a 60-person blockchain r d shop founded in toronto, canada.

A

Today, I'm here to introduce the rust implementation of the mena protocol that we've recently begun working on. First off. Let's take a look at some of the goals for this project.

A

Our primary goal is, of course, to build an alternative implementation of the mena protocol by increasing the diversity of nodes on the network, we aim to increase the resiliency of the network. Our second goal is to target additional architectures, specifically arm and wasm. This will enable us to run mina nodes in embedded and in browser environments.

A

Our third goal is to motivate the development of a complete specification. While this will most likely reflect the reference implementation, it should make it easier for additional implementations to be built through this process. It's very likely that areas of improvement will be identified either within the reference implementation or the protocol itself.

A

Now, let's answer some of the questions that you might have why rust rust is the most loved language in 2020, according to stack overflow's developer survey, it's well known for its excellent performance, its memory, safety and concurrency. It has a strong community driven open development process which is very active.

A

It has some of the best language, support and tooling for wasm, and it's already being used by a multitude of protocols for their primary or alternative implementations.

A

Why wasm well wasm is formerly described as a binary instruction format for a stack based virtual machine. The format is very compact, making it very portable.

A

The primary use case for wasm is to provide a safe execution environment within web browsers, which has comparable performance to running native software. This allows more sophisticated applications to be executed within a browser. However, since its inception, a number of other use cases have been identified outside of browsers and web applications.

A

Why arm well arm is a processor architecture that dates back about 30 plus years compared to the more common x86 processors used in most computers arm processors, have a lower cost, consume less power and produce less heat. They are very commonly found in iot devices as well as other socs.

A

Many of you have probably heard of the apple m1, that's an arm processor and a great example of the strengths of this architecture. So why chainsafe? Well, our company mission is to realize the potential of web3 technology through sustainable, open source development.

A

We see huge potential with mina in driving the blockchain space forward and we are very proud to be supporting the development of it. We have great experience building protocol implementations in addition to mina we're currently working on implementations for east 2, filecoin and polkadot as an open source first organization, we have a strong commitment to supporting community driven development.

A

Now, let's look at our current roadmap, which is very incremental after building the serialization primitives, we'll move into implementing the block selection algorithm, then we'll be implementing block verification and, after that, we'll be adding support for the ledger state and some additional state objects.

A

Finally, we'll be building out lib p2p support in the browser so that the node can receive blocks and submit transactions.

A

At the end of these initial milestones, we'll have a non-consensus node which can run in browser, although it won't yet participate in consensus, it can be used in browser-based applications to interact with the mina network. For many protocols this would be considered a light client, but thanks to the design of the media protocol, this will actually have the same security model as a full node.

A

We hope to continue this implementation and eventually complete a full consensus participating node. This project is still very young, but let's take a look at some of our progress so far for that first serialization milestone. We have built surte bin pro, which can be found in this github repo.

A

It's a rust implementation of the bin pro protocol using the serde framework. It's still experimental, but has almost complete functionality.

A

Next up we're going to be implementing the block selection, algorithm you'll be able to follow, along with our progress in our github repo chain slave slash, mina rs. Meanwhile, we'll also be working on a better name as well as some branding.

A

That's all I have for you today. Please feel free to reach out to me on telegram or discord. If you have any questions or feedback or interested in participating, we also have a channel in the meena discord called chain safe, rust client, where you can reach our team. If you love meena and you love rust, we would love to hear from you thanks. So much for watching.

B

Awesome uh thanks thanks david uh right, so I'm I'm brandon, I'm the first engineer at old one labs and I'm here to talk about all the engineering work. That's been going on on our team since, since mainnet this can be broken up really into two categories. I think one just general changes to the mina demon itself and the other is the work on snaps.

B

So we've done so much that I cannot possibly talk through everything, but I made this nice slide to sort of help show what's going on. Basically, there's there's a release, that's coming very soon and there's there's a current alpha, build where you can play with it on a devnet or test net, and I'm just going to walk through some of the some of the new things.

B

So, although on the left, you see the graphql and cli features, one one thing that we've done is: we've added a new vrf command that allows block producers at the beginning of an epoch, to figure out exactly which slots that they are supposed to win, that the vrf, the verifiable random function will tell their know that they've won a block and not only can block producers use this information themselves to have a better understanding of when they're about to create blocks, but they can even produce a file that can prove to others of when they're supposed to make blocks.

B

And- and perhaps you know, the community can build a tool that helps rank block producers and helps rank staking pools using this information. Another thing is we added a command to hash transactions? That was just missing, so it's there there's a new flag to specify a minimum block reward amount when you're creating a block, and this is useful to protect against the case where your node has just restarted. You haven't received enough smart work and someone tries to slip in a smartwork. That's super expensive, then.

B

uh Finally, on the graphql and cli side, we've added an api to detect stage, ledger proofs being emitted from blocks, and this is like the complicated concept, but but the general idea of it is this: is when the the transactions finish being snarked by snark workers and are actually added to the to the snark part of the blockchain and- and it's important to know this, because this is precisely when the information is checkpointed for the purposes of staking and delegating.

B

So so. With this api, you can figure out exactly when the last time this happened before an epoch, transition and better understand. What's going to happen in an upcoming fbi on the stability and bug fixes side, one thing that we've done is uh optimize the way that the menodaemon verifies snark work. So this happens in a separate process from the main menu daemon and basically the optimizations make it so that the mina demon can spend more time and energy and compute power doing all the other things it needs to do.

B

Rather than communicate with the start verification process, we've added more verbose, logging of long async cycles. This will help uh open labs and the community better understand where these bottlenecks, which trigger these long async cycles, are in the code base. So we can prioritize and fix them. We've also added more aggressive snark work, rebroadcasting logic to nodes.

B

So if your node is creating snark work and you want to sell it on the snorket place and you want block producers to see it, it's really important that you get your message out there and now it's more aggressively sent. Another thing is we're chunking transaction rebroadcasting now, so what this means is, if you have sent a ton of transactions from your node in a really short period of time, if you're, not careful, you can overload other nodes on the network and sort of get them to stop listening to you.

B

It won't actually break them, but it'll they'll just ignore your messages. So now uh the your node will be smart about the rate at which it sends those transactions out uh and it won't overwhelm uh the other nodes and everyone will listen to you uh and and finally, we've we've fixed an out of memory error that is happening from time to time. Based on an issue with urinal the journalist proof library, it's very complicated but point is on the alpha. We haven't seen this bug, which is really exciting and the last category here.

B

One thing I want to highlight is there have been 57 pull requests that have landed, that improved monitoring and alerting of the mena network. What this will do is it will make uh one labs, at least from the perspective of what our nodes see in the network better understand. What's going on and be alerted in the case of anomalies, so that you know we can take action urgently if necessary.

B

We've we've added a feature to the rosetta, which is uh the open protocol that allows information to be extracted from from different blockchains, so mina implements rosetta and now it supports querying for both liquid and locked balance, lookups on chain and then finally, we have now support for creating alpha builds, uh one of which is linked here and has already been released and, and the alpha builds are intended to be used on devnets and test nets when we feel that the alpha is stable enough, we'll promote it to beta beta is intended to use on mainnet, but you know not have all the nodes upgrade all at once and once we think that is stable, then we'll do an official release and uh just before we move on here, I want to give a shout out to andrew at the oberlin labs team, for making incredibly amazing release notes.

B

If you haven't seen them you should this slide would not have been possible without those release notes existing. So that's the meena demon. If we, if we go over to the snaps, this is just a glimpse into the stuff. That's already been completed on snaps, so one thing is: uh we have zero knowledge proving that works in the browser fully, so you can prove snarks in the browser all right. That's that's one one piece.

B

The other thing is not only can you prove them, but you can actually implement snark circuits using typescript or javascript in the browser and improve them, and even that's working there's, there's an rfc out for the new transaction model. That's needed to support snapseed protocol, there's just skip over a few of these things, there's a better hash function and then, finally, there's there's a ton of progress on https snaps. um This is this: is this concept where our websites that have public data behind an https certificate?

B

We can kind of get that data on chain on mina without the website really even having to do anything to integrate. So it's super super fascinating.

B

Looking forwards, uh the the 2021 priorities, the same same thing that we talked about at illuminate, we've already made a bunch of progress, but ob1 labs wants to enable the developer community to build and contribute to the ecosystem. In doing that, you know we're helping to support the rest implementation. That david talked about.

B

Chainsafe is working on we're working on snaps and various tooling associated with snaps and with mena, and then we're putting a lot of effort into improving performance uh and and fixing issues with the with the demon in order to improve performance on the network and to reduce hardware requirements for for running nodes. Yeah and there's gonna be a blog post, describing our priorities and the 2021 and 2022 roadmap coming really really.

B