Import Oxide engine (Stages 0–10) under MIT license

Full project snapshot migrated to new Gitea remote without history:
engine, editor, physics, script, examples, tests, docs, and assets.
Relicensed from GPLv3 to MIT and updated repo URLs.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

engine/benches/app.rs

@@ -0,0 +1,31 @@
+//! Benchmark for the Stage 5 schedule/module overhead.
+//!
+//! Stage 5 criterion: module/system scheduling overhead must be negligible
+//! compared to the Stage-4 hardcoded loop. There is no per-frame work here — the
+//! benchmark measures the *frame overhead itself*: advancing timing, walking the
+//! phase lists, and the fixed-timestep accumulator, with a realistic handful of
+//! empty systems registered. Check that `app_empty_update` is in the low
+//! nanoseconds (i.e. lost in the noise next to any real system's work).
+
+use criterion::{black_box, criterion_group, criterion_main, Criterion};
+use oxide_engine::app::{App, Schedule};
+
+fn empty_update(c: &mut Criterion) {
+    let mut app = App::new();
+    // A few no-op systems spread across phases, as a trivial game might have.
+    for _ in 0..4 {
+        app.add_system(Schedule::Update, |_| {});
+    }
+    app.add_system(Schedule::FixedUpdate, |_| {});
+    app.add_system(Schedule::Render, |_| {});
+
+    c.bench_function("app_empty_update", |bencher| {
+        bencher.iter(|| {
+            app.update(black_box(1.0 / 60.0));
+            black_box(app.time.frame)
+        });
+    });
+}
+
+criterion_group!(benches, empty_update);
+criterion_main!(benches);