UPD 06/06/2019

Open a late-night livestream from Colorado to walk through the freshly released Cardano roadmap.

Details the month and a half of internal discussion that shaped the roadmap's Byron, Shelley, Goguen, Basho, and Voltaire sections.

Recaps Byron's purpose of building a community and learning exchange listings and APIs, judging it an overall success despite being slower than hoped.

Update on Shelley: the self-node testnet is being prepped and already independently vetted, with about 150 stake pool operator sign-ups and a task force channel past 1,500 members.

Announces upcoming hackathons in Israel, Argentina, and hopefully Wyoming to expose Plutus and Marlowe to developers and gather feedback.

Goguen's progress shifting from an R&D cycle into an actual developer-facing product.

The roadmap's design as a simple interface that can keep linking out to more detail, videos, and blog posts over time.

Previews an upcoming Cardano Effect episode with David and commits to a monthly or six-week AMA cadence with advance notice, plus product managers doing their own AMAs.

Goguen content on the roadmap will be filled in over the next four to six weeks before moving on to Basho and Voltaire.

Contrasts the roadmap's new realistic design with the original attempt to force research into traditional Gantt-chart project management.

Explains why a waterfall approach failed for research-driven crypto development, pushing the formal methods and research teams toward agility.

Revisits the 2017 vision of small, competing, focused agile teams whose work emergently builds up the whole ecosystem.

Points to the Haskell wallet backend team's late-2018 redesign as the model, now shipping a release nearly every week.

Adopting Bech32 for Shelley addresses and finding both bugs and spec design defects through property-based testing, plus the extreme-programming approach that lets it plug into both Rust and Haskell clients.

Other teams converging on the same agile process, producing more demos and, overall, a roadmap of frequent iterative updates instead of big-bang releases.

Turns to community questions about prioritization and Cardano's architectural direction, starting with the KEVM and K framework.

Recalls that in 2015 proof of stake's viability was a genuinely open question, one proof-of-work advocates still dispute today.

Calls Ouroboros research an overwhelming success, putting Cardano ahead on proof-of-stake design and now in the fine-tuning phase.

Introduces the second high-risk pillar, sidechains and interoperability via structures like NIPoPoWs, crediting Aggelos Kiayias's team for its progress.

The sidechain goal: verify an asset hasn't been double spent using only a small amount of data, not the whole foreign chain, via light clients.

Introduces the third pillar: a programming model for smart contracts that also governs how they interact with the world outside the blockchain.

Points to the DAO hack and the Parity hack as cautionary tales of getting on-chain and off-chain code reconciliation wrong.

Frames the scalability challenge as delivering good performance for everyone from a thousand to a million users, the way Netflix, Google, Amazon, and Facebook do.

The pragmatic decision to fund research on both the functional and imperative sides of programming language design, crediting Phil Wadler and later Manuel Chakravarty for leading the functional track.

Marlowe as a Turing-incomplete DSL for financial contracts and Plutus as a full Turing-complete on-chain language derived from Haskell.

Lays out the two-layer model, a DSL running on chain paired with a general-purpose off-chain language, enabling formal specification and property-based testing.

IOHK's contributions to compiling Haskell to WebAssembly and to the GHCJS project so off-chain code runs in browsers and nodes.

The Plutus and Marlowe model is being validated through hackathon developer feedback.

Explains why the imperative-side research avoided simply forking an existing VM like the JVM or WebAssembly, betting instead on Grigore Rosu's K framework and runtime verification research.

The vision of semantics-based compilation, where writing a language's K semantics once lets it run automatically on the KEVM.

Admits the K framework research wasn't mature enough to be viable for consumers within the 2020 timeframe, prompting a strategy change.

Points to other groups using K, including Quantstamp formally specifying ERC-20 and ERC-721, and Elrond writing a Go K backend for KEVM.

Clarifies KEVM, Yul, and Solidity interoperability was deprioritized rather than abandoned, noting Cardano's own past Ethereum VM builds like Mantis.

Downscoping RINA research into a practical mini-protocol network stack, built by the network team and published in the Ouroboros-network GitHub repo.

Lists other active research threads: incentives, stake pool rational-actor behavior, the CIP process, voting, and the Treasury system, with Ben Chang leading Treasury and voting product management.

Compares protocol development to phone releases like the Samsung S10, always followed by a next version, and states the belief that Shelley, staking, and Goguen will all ship in 2019 barring disaster.

Argues decentralization is an aspiration rather than an achievable target, so Shelley and Goguen's goals will keep evolving past 2020.

Previews new features unlocked by the rewrite: contingent settlement, DID-based identity, and Ouroboros services like timestamping and a cryptographic randomness beacon.

Clarifies the difference between Cardano SL and Cardano CL as a multi-layer split between settlement and more ephemeral data and computation.

Connects the Goguen era to multi-party computation, first raised in the 2017 Why Cardano paper, using a poker game as the canonical example of locking funds, playing privately off-chain, then settling back on chain.

Names Royale and Kaleidoscope as Cardano protocols for this pattern, alongside MIT's Enigma project pursuing similar off-chain computation.

Breaks down Cardano CL's model types: MPC, zero-knowledge verification, and permissioned ledgers like a Fabric or Atala instance.

Frames Cardano SL's philosophical purpose as a financial operating system needing scripting, identity, metadata, and scale to billions of users.

Recalls the open question from the Why Cardano paper of how expressive the base-layer scripting language could be without risking security, noting it ended up more expressive than the 2017 plan but still not a world computer.

Observes that the industry as a whole is now validating Cardano's early bet on a multi-layer architecture.

Reflects on the project maturing from aspirational to practical, accepting that focus disappoints some people while satisfying others.

Expresses pride in IOHK's roughly 200-person team and in keeping principles intact despite 2017 to 2018 hype pressure.

The coming need for a CIP process modeled on Bitcoin's BIPs and Ethereum's EIPs, since Shelley removes IOHK's ability to unilaterally push hard forks.

Points to Tezos's on-chain governance experiment as a chance to learn from its successes and failures, cautioning against introducing too much democracy too early.

Targets implementing the Treasury system around Q1 2020 to give the community an alternative source of development funding.

Being inundated with interest from Fortune 1000 companies, governments, and municipalities wanting to deploy on Cardano.

Lays out the last major trip of the year, Japan, Georgia, and Israel meeting heads of state, then Uganda's royal family, Korea, and Argentina.

Reflects on the crypto industry's harsh side while affirming belief that most people in it are well-intentioned.

Asks the community to stress-test the self-node testnet across many configurations and file GitHub issues before the beta testnet.

Envisions Shelley's launch handing the network to a community a hundred times more decentralized than Bitcoin, followed by Goguen and Treasury work.

Closes by thanking the community for roadmap feedback and encouraging continued engagement on Reddit and Telegram.