| Known languages | GitHub Statistics | Languages used in public repositories |
|---|---|---|
Hello - I'm Richard. I'm a Year 12 A-Level student who builds backend systems, programming tools, and infrastructure projects in my spare time.
I work primarily with Lua/Luvit, Go, and C, and I enjoy designing things close to the protocol layer - HTTP internals, routing engines, storage formats, and distributed infrastructure (the last one is my favourite).
Despite still being in school, I have independently built several production systems used by thousands of users, along with language tooling and low-level runtime components.
Backend systems, high-performance infrastructure, datastores, and protocol tooling. Some language design too, if you would like to include the Nue Programming Language.
I enjoy:
- designing high-performance HTTP stacks,
- implementing RFC-compliant protocols like SMTP, TOTP, and OAuth2,
- building storage engines and KV databases,
- real-world systems optimuisation,
- Lua/Luvit internals, coroutines and event loops,
- and rewriting bottlenecks in C when necessary (although that might just be because I have outgrown Lua's purpose to the point where external intervention is required)
Of course, all of the above practically means nothing if I don't tell you about some real-world projects where I actually apply some theory.
Numelon Rubiลก - High-performance paste and content platform
- 2M+ stored documents
- 10K+ monthly users
- Low-overhead HTTP server through a heavy extension of
coro-http - Full telemetry with latency histograms, RPC counters, TTFB, and tag-based metrics
- MessagePack metadata pipeline with serialisation offloaded to C
- Burst-tested to ~12.5M requests without failure
- Used as the backend trust layer for sUNC's test result verification
Sklair - HTML compiler and build system
- HTML compilation with component expansion
- Head segmentation and deterministic ordering (resource hints, metadata, scripts, styles)
- Sandboxed pre- and post-build Lua hooks for data generation and build-time logic
- Strict filesystem sandboxing for hooks (project / cache / generated / built scopes)
- Live development server with filesystem watching and debounced rebuilds
- Builds that are only a few milliseconds
- Zero runtime dependencies in the browser
Currently being rewritten in Go using my in-progress framework, Rind
- Full OAuth2 authorisation code flow
- Sessions, grants, and application management
- Strict CSRF protection and protection against replay attacks
- Passwordless login architecture
- TOTP 2FA implemented from RFC 6238 / RFC 4226 using OpenSSL's HMAC
- Custom SMTP client, "Letterbox", with persistent connections, MIME encoding, and connection pooling
There is A LOT more in much more detail, however I would like to keep my GitHub README concise. For more information, please visit my website.
Website: richy.lol LinkedIn: richy-z Email: hello@richy.lol