The development team behind Core Solidity has released a comprehensive status update detailing the ongoing evolution of the new frontend for solc. Reflecting the state of the project as of early October 2026, the report emphasizes that Core Solidity is still strictly a prototype and is not yet intended for mainnet deployment. The update addresses common misconceptions, highlights significant syntax adjustments, outlines current capabilities alongside missing features, and shares valuable insights gathered from months of qualitative interviews with protocol developers, security auditors, and language researchers.
A central clarification emphasized in the update is that Core Solidity does not replace the entire compiler. Instead, it serves as a new frontend for solc, introducing a modern type system equipped with algebraic data types, pattern matching, generics, traits, and compile-time evaluation. By shifting much of the compiler’s built-in logic into a standard library written directly in the language itself, Core Solidity modernizes the developer experience while preserving a familiar foundation. Crucially, the backend remains unchanged: Core Solidity compiles down to Yul, handing off the generated code to the optimizer and code generator that solc already utilizes for Classic Solidity.
Over the summer, the syntax underwent a major overhaul to align much more closely with Classic Solidity. Consequently, code samples published in earlier design documents no longer compile directly, though the core programming concepts they describe remain entirely valid. A small contract example, such as an escrow managed by lifecycle states stored in an enum variable, demonstrates how types now follow names in a postfix fashion (phase: Phase, amount: uint256). Additionally, except for a minimal set of built-in types, symbols are not in scope unless explicitly imported, and enum variants are now capable of carrying data, fulfilling the definition of algebraic data types.
The prototype compiler, hosted within the solcore repository, currently supports a variety of advanced features and has successfully been tested on ports of existing smart contracts, including the Ethereum deposit contract, WETH9, and a minimal ERC20 token. These ported contracts are compiled through solc and executed on the Ethereum Virtual Machine as part of continuous integration testing, though events are temporarily omitted due to current prototype limitations. Furthermore, developers looking to experiment with the language can utilize the web-based playground, which features runnable examples such as multi-vault compositions and feature-module-based tokens. Parallel to these software developments, the team is actively formalizing the semantics of Core Solidity using the Lean 4 proof assistant within a dedicated repository.
Despite robust progress, several production-grade features remain absent from the main branch of the solcore repository. While pull requests are continuously merged to bridge these gaps, current limitations include the absence of custom errors, events, full value transfer mechanics, and complete access control implementations in default examples. Gas costs and code size metrics have not yet been a primary focus for optimization, meaning comparisons with Classic Solidity have yet to be conducted.
A deliberate structural departure from Classic Solidity is the complete absence of inheritance. Core Solidity features no virtual keywords, abstract contracts, or traditional inheritance hierarchies. Instead, contract extension and modularity are achieved exclusively through composition via traits and modules. Security researchers and auditors interviewed during the project’s qualitative research phase expressed general approval of this shift, noting that removing inheritance significantly simplifies the process of reviewing contract attack surfaces and identifying exposed public functions.
Between March and October 2026, the development team conducted 18 qualitative interviews with industry stakeholders, including protocol engineers, security auditors, educators, tool maintainers, and language researchers. Participants responded enthusiastically to algebraic data types and pattern matching, particularly highlighting the compiler’s safety feature of rejecting unhandled match cases that lack a default arm. Discussions regarding module composition revealed general comfort with the removal of inheritance, though certain teams requested more extensive real-world examples demonstrating how reusable modules can effectively manage state without interacting directly with raw storage slots.
Two significant open questions highlighted in the report concern ABI compatibility and the ultimate scope of the standard library. While Core Solidity is designed to maintain backward compatibility with Classic Solidity by supporting external calls to legacy contracts, bidirectional compatibility remains a complex challenge. Specifically, algebraic data types within external interfaces require custom encodings not currently defined by the standard ABI specification, preventing legacy contracts from seamlessly calling every Core Solidity interface without intermediary work. Regarding the standard library, community opinions remain divided between maintaining a minimal core versus offering a more comprehensive set of built-in utilities, with developers emphasizing that bugs in foundational libraries impact the entire ecosystem.
Looking ahead, the team is actively scoping an initial installable release of the prototype to facilitate early community feedback. Documentation updates remain a high priority for upcoming development cycles to support developers transitioning beyond the online playground. Engagement with the global Ethereum community will continue through upcoming appearances, including a presentation at the DeFi Security Summit and a dedicated workshop at Devcon in Mumbai this November. Developers and researchers interested in contributing or providing feedback can access the source code repositories, join open weekly team calls, or reach out directly to the core development team.
