Migrating from log to rlg

The log crate is the facade. rlg can either replace it (direct rlg API) or install as its backend (drop-in). Choose by whether you want the fluent API or minimal diff.

Option A: install rlg as the log facade backend

Zero call-site changes.

#![allow(unused)]
fn main() {
use log::info;

// Initialize once at startup:
rlg::init().unwrap();

// Every existing log::* call routes through rlg's engine now.
info!("user_id={user_id} authenticated");
}

You get structured storage, redaction, OTLP export — but records still look like the message-formatted strings your log:: calls produced. Attributes are not extracted.

Option B: rewrite to the rlg fluent API

Diff at the call site but you get first-class structured attributes.

#![allow(unused)]
fn main() {
// before (log)
info!("user_id={user_id} authenticated");

// after (rlg)
rlg::log::Log::info("authenticated")
    .with("user_id", user_id)
    .fire();
}

Level mapping

Setup

#![allow(unused)]
fn main() {
// before
env_logger::init();

// after
let _guard = rlg::init().unwrap();
}

Filter via RUST_LOG continues to work — rlg parses the same env var syntax.

Related

• from-tracing.md — larger diff, richer target.
• from-slog.md.