Build a revolutionary chess engine in V (Vlang) — contributors wanted

[Chessonaut]

Howdy!

I’m starting an open collaboration to build a **next-gen chess engine written in V (Vlang)**. The goal isn’t “yet another clone,” but a clean, modern engine that’s fast, understandable, and fun to extend—while still being competitive.

If you like compilers, performance, search algorithms, or just want to ship something real in a newer language, this is for you.

### Project vision (what we’re building)
A chess engine with:
- **Strong baseline play** (solid search + evaluation, good move ordering, correct endgame handling)
- **Modern architecture** (clear modules, testable components, reproducible builds)
- **Speed** (bitboards, tight loops, profiling-driven optimization)
- **Experimentation-friendly design** (easy to plug in new eval terms, search ideas, or NNUE later)
- **Great tooling** (bench suite, perft correctness gates, UCI protocol support)

### Why V (Vlang)?
V is a statically typed, compiled language designed for **simplicity + speed**. It feels a bit like Go in readability, while offering C-like performance when written carefully.

What makes V appealing for a chess engine:
- **Fast compilation** and a simple build story (great for iteration)
- **Clean syntax** (easy for new contributors to read quickly)
- **C interop** when you need it (SIMD, tablebases, etc.)
- **Single-binary output** (nice for distributing engine builds)
- Built-in formatter and testing tools

If you’re new to V, you can still contribute—there are tasks ranging from documentation and test harnesses to core search code.

V language links:
- V homepage: https://vlang.io/
- V GitHub org: https://github.com/vlang/v
- V docs: https://github.com/vlang/v/tree/master/doc

### Who I’m looking for
All skill levels welcome. Specifically:

**Engine / algorithm folks**
- alpha-beta / PVS, aspiration windows, LMR/LMP
- move ordering, killer/history heuristics
- transposition tables, repetition detection, time management
- evaluation design (material, PST, mobility, king safety, pawn structure)
- endgames: perft correctness, basic tablebase probing integration later

**Performance / systems folks**
- bitboards, magic bitboards, attack generation
- profiling, cache behavior, memory layout
- concurrency for search (lazy SMP style)

**Tooling / UX**
- UCI protocol implementation
- test harness (perft suite, STS/EPD, self-play scripts)
- CI workflows, reproducible benchmarks

**Docs / community**
- contributor guides, “good first issue” curation, examples

### What’s already available (repos)
If you want to coordinate on GitHub, here are starting points:

- **Main repo (current work / coordination):** https://github.com/hedgeg0d/vchess
- **V language (compiler + stdlib):** https://github.com/vlang/v

If you already have (or want) a dedicated engine repo name (e.g., `vengine`, `vchess-engine`, etc.), say so and we can structure it cleanly from day one.

### Initial roadmap (first milestones)
1. **Core correctness**
- Legal move generator
- Perft (with standard test positions)
- FEN/PGN basics (at least FEN to start)
2. **UCI engine skeleton**
- `uci`, `isready`, `position`, `go`, `stop`
3. **Search v1**
- iterative deepening alpha-beta with TT
- quiescence search
- basic move ordering
4. **Eval v1**
- material + PST + a few simple terms
5. **Benchmarks + CI**
- perft gate in CI
- basic speed benchmark

### How to join
Reply with:
- your background (engine / perf / V / tooling / docs),
- what you want to work on first,
- and your GitHub username.

If you want a starter task, say “give me a good first issue” and I’ll propose something small but meaningful (perft harness, Zobrist hashing, TT layout, UCI parsing, etc.).

### Contact / coordination
We can coordinate via:
- GitHub Issues + PRs (preferred)
- A lightweight chat (Discord/Matrix) if people want real-time sync

If you’re interested, jump in—let’s build something ambitious in V and make it a project people can learn from *and* play against.

[Ciekce]

an entire wall of AI slop

[towforce]

an entire wall of AI slop

What makes you confident that Chessonaut's post was written by a chatbot?

[chrisw]

an entire wall of AI slop

What makes you confident that Chessonaut's post was written by a chatbot?

Same detection reasoning that most of your posts are written by chatbot.

[towforce]

an entire wall of AI slop

What makes you confident that Chessonaut's post was written by a chatbot?

Same detection reasoning that most of your posts are written by chatbot.

Ahhh... the note at the top (link)!

[tapio]

A guy with no sufficient skills, trying to appear bigger ("we") asks for people with skills. In a field where there's no money to be made.

We laugh.

[syzygy]

an entire wall of AI slop

What makes you confident that Chessonaut's post was written by a chatbot?

Do you really not see the dead giveaway?

[towforce]

an entire wall of AI slop

What makes you confident that Chessonaut's post was written by a chatbot?

Do you really not see the dead giveaway?

The two online tools I've tried both say it's highly likely that the text is AI generated. I'm not well-practised in the art of differentiating AI text from human writing (but if I knew the author's own writing style then I would be able to tell in a heartbeat): I'm guessing you're talking about the heavy use of "markdown" markup?

Btw - I've looked at the GitHub repo, and some of the files on there are surprisingly old for a "new" project.

[syzygy]

I'm guessing you're talking about the heavy use of "markdown" markup?

Yes, nobody does that in a forum post. In particular if it's the wrong type of markup.

Then there's also the general style with em dashes and way too many details that don't really make sense but contain words picked from the correct jargon. I mean, really, is this novice programmer looking for people wanting to focus on memory layout?

And then this sentence:

If you already have (or want) a dedicated engine repo name (e.g., `vengine`, `vchess-engine`, etc.), say so and we can structure it cleanly from day one.

Huh?? This makes negative sense to put in.

Or this:

If you want a starter task, say “give me a good first issue” and I’ll propose something small but meaningful (perft harness, Zobrist hashing, TT layout, UCI parsing, etc.).

The LLM is prompting Chessonaut there.

[towforce]

Once pointed out, it's all obvious!