TABConf 6: A Retrospective

For Localhost Research, TABConf 6 will retain a special place in our history. In the BitDevs Socratic Village at 10:30AM on Friday October 25th, we made our first public appearance. Of our team, Justin and Murch were present, alongside two of our board members, Mark Casey and Matt Corallo.

Prior to our appearance, Adam Jonas of Chaincode Labs gave an excellent presentation on the main stage that covered many frequently asked questions for those considering a career in open source development (Bitcoin + FOSS = BOSS). It was the perfect preface for the public reveal of Localhost which was immediately followed by a review of the research report recently published by the 1A1z team. Part 1 of their research explored who funds Bitcoin Core contributors, how their funding is deployed, and what the global distribution of funding recipients is. This report is excellent and worth a read.

From here, the BitDevs Socratic Village was underway. Each event was unique and exciting. It was an honor to sponsor the BitDevs Socratic Village and to participate in various events across the conference. For a full list of events that the Localhost team and board members participated in, please see here.

Because the Socratic Village operates under Chatham House Rules, we can’t give direct comment attributions to specific individuals, however we can provide some summaries of the events our team was present for.

The first Socratic Session touched on some major project updates in Bitcoin Core. Led by Justin, Murch, and David Gumberg, the following topics were covered: Erlay, Multiprocess, Kernel, CVE-2024-35202 and mutation-core.

It was exciting to note that Erlay has been picked up by Sergi Delgado. We reviewed his feedback of the existing code, modifications that have been made to it, and some simulations he is running with the Hyperion project. There was also some light coverage of minisketch. Multiprocess was also of interest to the crowd. We summarized the work to-date, looked at the new Cap’n Proto dependency and discussed various binary and packaging options. There was significant interest in the Kernel project and mutation testing as well.

Next up was the Changing Channels Socratic Panel, which took a look at consensus-compatible L2s that used primitives distinct from those found in channel-based protocols. Justin moderated discussion between the crowd and the panelists (Robin Linus, Jonas Nick and Erik De Smidt). Each panelist gave a presentation on a subject of interest: BitVM2, Shielded CSV and Ark. We learned about how Shielded CSV might use BitVM2, what properties Ark and Shielded CSV have and some of the tradeoffs they make to produce those features. We covered the UX of these two projects, some of the mechanics of BitVM2 peg ins/outs, as well as how the various projects handle worst-case scenarios (reorgs, liveness failures, etc).

We also learned a bit about the Shielded CSV nullifier construction and how sign-to-contract and other primitives help minimize the on-chain footprint of this protocol. We spent some time reviewing the properties of out-of-round Ark payments and their statechain-like security model. One individual from the crowd also mentioned WabiSabi as an early design consideration for in-round Ark payments. While not currently being implemented, they present an option in the future to introduce strong privacy within an Ark. Lastly, there was a short treatment on what future script upgrades might improve these protocols.

The Modern Mempool Architecture Socratic Panel with Matthew Zipkin, Gloria Zhao, Murch, and Justin was great. P2A/Ephemeral Anchors, Package Relay+TRUC, and Cluster Mempool were each covered. After giving the audience an overview of each of these projects, we learned exactly why, for example, package relay is dependent on cluster mempool and how various roll-out scenarios might effect users attempting to broadcast transactions that leverage these new policies. There was some limited discussion on how new script mechanisms might remove the need for anchors in some contexts.

On the main stage, Matt Corallo debated audience members about MEV, MEVil and Mining Centralization. The discussion was respectful and many different perspectives were shared. There is a lot of research required on this subject and we hope the community was left with a desire to think more deeply about these issues.

On Day 4, Murch spoke on the Main Stage about the Cluster Mempool proposal. He described how blockbuilding and eviction work with the current mempool, highlighted issues with those and then walked the audience through the approach taken by the Cluster Mempool proposal: by tracking transactions in the context of clusters and determining the order in which transactions would get mined in each cluster, it would become possible to predict the feerate at which all transactions would be included into blocks, block building would be sped up, and eviction would first affect the transactions to be mined last.

We also hosted two more Socratic Panels on day 4: one on miner centralization and another on Discreet Log Contracts (DLCs) in production. Our mining centralization panel was moderated by Justin. Panelists included Matt Corallo, skot9000, average-gary, and Evan Price. While spirits on this panel were high, there remained a somber tone that many hard problems remain unsolved. There was a base assumption that until ASIC manufacturing is commoditized, we will have some unavoidable centralization. Ideally, however, end-users and third party companies can take ASICs directly off-the-shelf and produce their own rigs such that users have more choice and control over their hardware.

On the software side, protocols like Stratumv2 and Hashpools balance some of the centralization pressures we have seen in the network. We covered recent allegations of hashrate laundering and the deleterious effects of FPPS on the mining ecosystem writ large. There was some additional conversation around MEV and how the topic intersects with decentralized and client-side templating protocols. Finally, questions of how the community could incentivize miners to use non-FPPS pools and client-side templating protocols was discussed.

Our DLCs in Production Socratic Panel was also moderated by Justin and had three panelists: Matthew Black (Atomic Finance), Shehzan Maredia (Lava Loans) and Benny B (dlcdevkit). To start, each panelist shared some perspectives on DLCs and their individual projects. Discussion topics that were covered included some of the UX complexities of DLC implementations (e.g. signing and exchanging thousands of CET signatures, lack of graceful recovery if DLC setup phase fails, novation, and poor spreads because of limited competition amongst market makers) balanced against the positive qualities of DLCs (e.g. private betting when compared to global ETH contracts, lower attack surface, and limited oracle price manipulation risk). We also discussed how DLCs might be improved with new script upgrades (e.g. CTV would reduce the state management burden for backups as well as eliminate the large set of CETs that need to be signed).

Aside from the panels, talks and sessions in the Socratic Village on days 3 and 4, at the Bitcoin Core Builders Day Table on day 2, Murch collected feedback from users of Bitcoin Core, answered general questions about Bitcoin Core, and guided prospective Bitcoin Core contributors through a code review. Matt Corallo also gave a follow-up talk to his 2022 talk about all the things broken in lightning. This time with as more positive spin, with a focus on all the things that have improved across the ecosystem to make the lightning UX better and the network more robust.

We finished off the event with the BitDevs Socratic Village Hardcore Trivia, sponsored by Localhost. It was a close one, with ChillDK-OG coming in first with 17 points out of 28.5. Three teams tied for second place with 15 points each: Team Monad, Zero Knowledge and Kick Flip. However, team Zero Knowledge collected a bonus point for most creative team name and ended with 16 points.

There were a number of other events in the Socratic Village which did not include Localhost participants (full list). To all those who contributed content to the village, a special thank you! To Hannah Rosenberg, Evan Price, coinward and Leo, thank you for overseeing village operations and making sure everything ran smoothly.

And to the conference organizers: you are the life blood of this event and The Real Ones. Mike Tidwell, Brandon Iglesias, exfrog and Brianna, thank you for everything you do. We’ll see you next year.