Topics Livepeer is Exploring at Devcon4
A number of members of the Livepeer project will be at Devcon 4 in Prague from October 30th-November 2nd. Here are some of the topics in the Ethereum ecosystem that we’re highly interested in, along with events or breakouts that we’ll be participating in around those topics.
There is a lot of attention in the Ethereum ecosystem on payment channels, generalized state channels, and plasma chains for scalable transfer of value. While these solutions provide sound economic security for guaranteed payments, and exit-ability on faults, for some use cases the overhead doesn’t justify the security. In particular, with regards to payment channels, if you need to pay for the overhead of opening up and closing channels with many providers (or trust centralized liquidity providers), and you need to rely on both parties being always online to watch for faults, these mechanisms can be infeasible or economically impractical.
For use cases such as video encoding, where you may be working with many providers concurrently whom you need to exchange value with, probabilistic micropayments (where the payer issues something akin to lottery tickets with an expected value to a service provider) may make more sense — especially if you’re exchanging tit-for-tat payments for work performed, and can accept the trust characteristics that a provider failing to do a small bit of work means they still get paid whilst losing the opportunity to perform future work.
In Livepeer’s protocol, this is certainly the case, and the exploration of moving to probabilistic micropayments to dramatically reduce the cost of using the Ethereum network to settle payments is a top priority in our research, design, and implementation. We’re looking to talk with people interested in the details of probabilistic micropayments at Devcon, and we’ll be hosting a breakout discussion on the topic.
In particular, we’ll be digging into the topics of ticket expiration, confirmation signatures, required deposit amounts to avoid accidental double-spends, and fire+forget schemes vs online schemes.
Event: Tuesday, October 30th at 3pm — 1 hour breakout discussion. Email us if you’d like to participate.
The data availability problem is often discussed in the Casper design dialogue, in verification protocols, and in incentivized file storage networks. We’d like to abstract this problem to the form of:
- A payer would like to issue a bunch of work to a set of service providers, along with guaranteed ability to settle payment upon work successfully performed.
- Though even if the service provider can prove that they performed the work correctly, how does a protocol ensure that the payer actually receives the output of the performed task?
In Livepeer this manifests itself in the form of true, fire+forget video-on-demand transcoding. These sorts of schemes work in an online context such as live streaming where payments are tit-for-tat, but fall apart due to the data availability problem in offline contexts.
There is an ETHResearch breakout on this topic that we’ll be participating in, however please get in touch if you’d like to discuss in the context of file storage or streaming content.
How can protocols reward participants for contributing value to the network in fair and verifiable ways beyond just proof of work? In Livepeer users can participate and be rewarded for:
- Transcoding video
- Performing QA on the set of nodes whom jobs are routed to through delegation
- Facilitating the wide and fair distribution of initial stake
- Campaigning based upon software development contributions or community contributions to the ecosystem through delegation
The active ways in which participants can accrue value across networks beyond just mining and validating are being collectively referred to as Mining 2.0 or Generalized Mining. We’ll be participating in workshops and discussions on this topic.
Event: Coinfund Generalized Mining Meetup, Tuesday, October 30th at 6pm.
There’s a lot of promise behind the moon math of zkSNARKs and STARKs, but Devcon will be a great opportunity to dig into the details and understand how these soon-to-be-practical technologies can help reduce costs and increase trust in decentralized protocols in a near term practical sense. In particular, Livepeer is interested in learning about the use of STARKs in two areas:
1. Transcoding verification
2. Aggregation of probabilistic micropayment tickets to be cashed on-chain with a single transaction
Event: Wednesday, October 31st 1:00pm, coding workshop on STARKs presented by Starkware team.
And of course…
Decentralized video infrastructure. While the above topics are all Ethereum specific, our team’s experience and expertise is in video, streaming, and infrastructure. We welcome any discussions and meetings on building video applications using Livepeer, and the future of decentralized compute, content delivery, and web 3 infrastructure in general. We’ll be presenting at the below event alongside Swarm, Fluence, and NuCypher.
Event: Web3 Stack, what’s next for developers? Tuesday, October 30th 6:00pm.
See you at Devcon4, and please get in touch if you’d like to catch up on any of the above topics.