- Simplicity of design - there should be one obvious way to do it.
- The code is the spec.
- Readability is paramount - beautiful is better than fast.
- Minimal code - keep total footprint small.
- Minimal dependencies - all dependencies must get audited, and become part of the repo.
- Modular dependencies - wherever reasonable, make components modular.
- Finished - software projects that don't become finished are projects that are forever vulnerable. One of the primary goals of the Gno language and related works is to become finished within a reasonable timeframe.
- Maintainable, debuggable, and future-proof codebase.
- The interpreter serves as a spec for how the AST is meant to be interpreted.
- The (virtual) machine is designed to interpret an immutable AST which matches the language.
- The interpreter is meant to become independent of the host language, Go.
- After the Gno interpreter can interpret itself, we will implement bytecode compilation.
- Each node can run on a commodity machine. Corollarily, for scaling we focus on sharding & forms of IBC.
- Correct, debuggable software is more important than extreme performance.
- Multicore concurrency makes Tendermint within the range of theoretical performance.
- Go is chosen for faster development of modular components, not for maximum speed.
- Real bottleneck is in the application layer, not in supporting large validator sets.
- Focus on feature completeness, debuggability, and maintainability over extreme performance.
- No envs.
- No short flags.
- No /bin/ calls.
- No process forks.
- Struct-based command options.
- Single base token.
- Deflationary is sufficient when tx fees are imminent.
- Int64 is sufficiently large to handle human numbers; for everything else, use denom conversions.
- Truth is revealed in the open light.
- Everybody has the potential to see the light.
- Those who see the light are moved to expand it.