spawn is called while the SlintContext is still alive

Error message

spawn is called while the SlintContext is still alive

What it means

`xdg_desktop_settings::spawn()` starts the portal watcher that tracks desktop appearance settings (dark mode etc.). It upgrades the passed `SlintContextWeak`; if the context is already gone the function cannot proceed and panics, since spawn must be called while the context is alive.

Solutions

Keep the `SlintContext` (platform) alive until after backend startup completes; only drop it at process end
Ensure `slint::platform::set_platform()` is called before any teardown paths run
If this occurs during normal shutdown, defer dropping the context until after the event loop exits

Example fix

// before
{
    let ctx = slint::platform::Platform::context();
    drop(ctx); // context gone
    start_backend(); // spawn panics upgrading weak ctx
}
// after
{
    let ctx = slint::platform::Platform::context();
    start_backend(); // spawn succeeds
    // ctx dropped last
}

Defensive patterns

Strategy: type-guard

Validate before calling

// ensure the platform context is alive before backend startup
let ctx = slint::platform::Platform::context(); // strong reference held
backend_start(&ctx);

Try / catch

std::panic::catch_unwind(|| xdg_desktop_settings::spawn(&shared, &ctx_weak))
    .map_err(|_| anyhow!("SlintContext dropped before portal watcher spawn"))?;

Prevention

Hold the SlintContext until the process/event loop ends
Call set_platform before any shutdown paths can run
Defer dropping the context until after the event loop exits

When it happens

Trigger: Calling `spawn` (from backend startup) after the `SlintContext` was dropped or before it was fully created - i.e. backend shutdown/teardown racing with settings initialization.

Common situations: Dropping the Slint platform/context very early in embedded or custom-platform setups; shutting down the app during initialization; misordered platform creation in custom integrations.

Understand the failure class

Background: "This is a bug, please report it": internal invariant violations, unreachable panics, and SNH errors explained — this error's family across 47 libraries.

Related errors

AI-assisted analysis of slint-ui/slint@bb937076de (2026-09-16). Data as JSON: /api/errors/5b5b8e1153d8efd2. Report an issue: GitHub.

Appendix: source

Thrown at internal/backends/winit/xdg_desktop_settings.rs:279

            name.as_str(),
            "org.freedesktop.DBus.Error.ServiceUnknown"
                | "org.freedesktop.DBus.Error.NameHasNoOwner"
        ),
        zbus::Error::FDO(err) => matches!(
            **err,
            zbus::fdo::Error::ServiceUnknown(_) | zbus::fdo::Error::NameHasNoOwner(_)
        ),
        _ => false,
    }
}

/// Starts the portal watcher and the timeout that releases the first windows if
/// the portal is slow or missing. Returns the task so the backend can abort it.
pub(crate) fn spawn(
    shared_data: &Rc<SharedBackendData>,
    ctx: &SlintContextWeak,
) -> Option<i_slint_core::future::JoinHandle<()>> {
    let strong_ctx = ctx.upgrade().expect("spawn is called while the SlintContext is still alive");
    // Hold back the first windows until the query applies the real appearance.
    shared_data.desktop_settings.appearance_pending.set(true);

    let shared_weak = Rc::downgrade(shared_data);
    let ctx_weak = ctx.clone();
    strong_ctx
        .spawn_local(async move {
            if let Err(err) = watch(&shared_weak, ctx_weak).await {
                if !portal_unavailable(&err) {
                    i_slint_core::debug_log!("Error watching for xdg desktop settings: {err}");
                }
                // The portal is unavailable; create the waiting windows anyway.
                finish_appearance_query(&shared_weak);
            }
        })
        .ok()
}

View on GitHub (pinned to bb937076de)