01
Clearly typed.
Checked early.
Explicit types, immutable values and visible optional data help catch mistakes during checks.
name: String?
match name {
Zelyra 0.3.0
·
in active development
From idea
to your own application.
With Zelyra you define data models, business rules and web pages in a clear, statically typed language. Built in Rust.
customer.zyl
✓ checked in 4ms
01table customers {
02id: Id primary auto
03name: String(100) required
04email: Email?
05}
schema
→
types
→
forms
→
API
◇
A shared model
Goal: connect schema, types, validation & SQL
✓
Checks during development
Detected issues in supported code, before deployment
01 / the idea
Describe once.
Reuse everywhere.
Zelyra connects data models, types, validation, SQL and interfaces. From a clear description, proven building blocks for your application emerge step by step.
Understand the principle →
01
Foundation
Data Model
The central schema defining tables, constraints and types.
table customers { ... }
02
Logic
Types & Rules
Nominal types and contracts help describe and check valid states in supported cases.
type UserId = Id
03
Web
Web & Forms
The experimental runtime can generate routes and forms with schema-aware validation for supported declarations.
page "/customers" { ... }
04
SQL
Checked SQL
Explore the database tooling available in the current experimental release.
query { select id, name }
02 / why zelyra
Clearly typed.
02
The model takes
center stage.
One schema can become the foundation for domain types, SQL, forms, APIs and CRUD.
table→form→API
03
Start small.
Grow safely.
Begin with a small program and expand it step by step into a full web application.
Open learning path →03 / THE MANIFESTO & MEANING · DIGITAL SOVEREIGNTY
What Zelyra means:
A dedication to digital sovereignty & architectural craftsmanship.
Zelyra is an experimental language project for database-backed web applications. Its direction is to make data models, database work, and web features easier to describe and inspect. The current 0.3.0 release is an early prototype: its checks cover only implemented cases and do not prove an application correct or production-ready.
Z
Zero-Tolerance for Flaws
The compiler checks supported language constructs and reports detected issues. These checks are limited in scope and are not a formal proof that a program is correct.
E
Excellence & Sovereignty
The compiler is open source and projects can be developed locally. The current runtime and database support are still evolving; review the release notes for the exact scope.
L
Logical Transparency
Code that explicitly states what it expects and delivers. Explicit types, readable schemas, and verifiable contracts instead of hidden magic.
Y
Yield & Productivity
The project explores a more connected workflow for data models and web applications. Which parts are available depends on the feature and template used.
R
Resilience (Forged in Rust)
The compiler is written in Rust. That implementation choice is not a benchmark, a production-readiness certification, or a guarantee about application behavior.
A
Autonomy (Data Sovereignty)
Local development is possible, and MariaDB is the primary database target in the current project. Check the documentation for supported workflows and their limitations.
⚡ Vision & Roadmap
Development note: The principles and architecture outlined here describe the definitive goal and vision of Zelyra, not yet its full current completion. The language and ecosystem are actively being developed milestone by milestone towards this standard.
04 / getting started
The best way to learn
is by doing.
Four accessible steps guide you from your first program to a database-backed web application.
your first lesson
Say what you mean.
Zelyra code describes clearly what values a function expects and what result it delivers.
View example →
fibonacci.zyl
fn fibonacci(n: Int) -> Int {
if n <= 1 {
return n
}
return fibonacci(n - 1) +
fibonacci(n - 2)
}05 / try it out
Read. Modify.
Understand.
Here is a small Zelyra page. Change the code in the editor or the parameter in the route bar and watch how it turns into a safe HTML response in real time.
Quick test:
✦
Good to know & interactive
This browser demo illustrates selected examples; it is not the Zelyra compiler or a full application runtime.
07 / installation & quickstart
Ready in minutes.
Run locally or with Docker.
The CLI can run as a standalone binary. Database-backed applications still need configured services such as MariaDB.
Option 01 · Standalone
Native CLI Installation
Clones the repository and builds the CLI from source. Build time and application performance depend on your system and workload.
$ git clone https://github.com/sf1976/zelyra.git
$ cd zelyra && ./install.sh
Next command:
zelyra check fibonacci.zyl
Option 02 · Full Stack
Docker Compose & MariaDB
Builds and starts the generated web application and MariaDB services using the project’s Compose file.
$ zelyra new my-app --mariadb
$ cd my-app && docker compose --env-file .env -f docker-compose.mariadb.yml up -d --build
Then initialize the schema:
docker compose --env-file .env -f docker-compose.mariadb.yml exec web zelyra db setup main.zyl
open source & community
Build your next project
with true independence.
Explore the handbook, inspect the open-source compiler on GitHub or start with our 5-minute interactive tutorial.
Z