@ -135,6 +135,54 @@ If a `<svelte:boundary>` with a `pending` snippet is encountered during SSR, tha
> [!NOTE] In the future, we plan to add a streaming implementation that renders the content in the background.
> [!NOTE] In the future, we plan to add a streaming implementation that renders the content in the background.
## Forking
The [`fork(...)`](svelte#fork) API, added in 5.42, makes it possible to run `await` expressions that you _expect_ to happen in the near future. This is mainly intended for frameworks like SvelteKit to implement preloading when (for example) users signal an intent to navigate.
```svelte
<script>
import { fork } from 'svelte';
import Menu from './Menu.svelte';
let open = $state(false);
/** @type {import('svelte').Fork | null} */
let pending = null;
function preload() {
pending ??= fork(() => {
open = true;
});
}
function discard() {
pending?.discard();
pending = null;
}
</script>
<button
onfocusin={preload}
onfocusout={discard}
onpointerenter={preload}
onpointerleave={discard}
onclick={() => {
pending?.commit();
pending = null;
// in case `pending` didn't exist
// (if it did, this is a no-op)
open = true;
}}
>open menu</button>
{#if open}
<!-- any async work inside this component will start
as soon as the fork is created -->
<Menuonclose={()=> open = false} />
{/if}
```
## Caveats
## Caveats
As an experimental feature, the details of how `await` is handled (and related APIs like `$effect.pending()`) are subject to breaking changes outside of a semver major release, though we intend to keep such changes to a bare minimum.
As an experimental feature, the details of how `await` is handled (and related APIs like `$effect.pending()`) are subject to breaking changes outside of a semver major release, though we intend to keep such changes to a bare minimum.
Often when encountering this issue, the value in question shouldn't be state (for example, if you are pushing to a `logs` array in an effect, make `logs` a normal array rather than `$state([])`). In the rare cases where you really _do_ need to write to state in an effect — [which you should avoid]($effect#When-not-to-use-$effect) — you can read the state with [untrack](svelte#untrack) to avoid adding it as a dependency.
Often when encountering this issue, the value in question shouldn't be state (for example, if you are pushing to a `logs` array in an effect, make `logs` a normal array rather than `$state([])`). In the rare cases where you really _do_ need to write to state in an effect — [which you should avoid]($effect#When-not-to-use-$effect) — you can read the state with [untrack](svelte#untrack) to avoid adding it as a dependency.
### experimental_async_fork
```
Cannot use `fork(...)` unless the `experimental.async` compiler option is `true`
```
### flush_sync_in_effect
### flush_sync_in_effect
```
```
@ -140,6 +146,18 @@ The `flushSync()` function can be used to flush any pending effects synchronousl
This restriction only applies when using the `experimental.async` option, which will be active by default in Svelte 6.
This restriction only applies when using the `experimental.async` option, which will be active by default in Svelte 6.
### fork_discarded
```
Cannot commit a fork that was already committed or discarded
```
### fork_timing
```
Cannot create a fork inside an effect or when state changes are pending
Often when encountering this issue, the value in question shouldn't be state (for example, if you are pushing to a `logs` array in an effect, make `logs` a normal array rather than `$state([])`). In the rare cases where you really _do_ need to write to state in an effect — [which you should avoid]($effect#When-not-to-use-$effect) — you can read the state with [untrack](svelte#untrack) to avoid adding it as a dependency.
Often when encountering this issue, the value in question shouldn't be state (for example, if you are pushing to a `logs` array in an effect, make `logs` a normal array rather than `$state([])`). In the rare cases where you really _do_ need to write to state in an effect — [which you should avoid]($effect#When-not-to-use-$effect) — you can read the state with [untrack](svelte#untrack) to avoid adding it as a dependency.
## experimental_async_fork
> Cannot use `fork(...)` unless the `experimental.async` compiler option is `true`
## flush_sync_in_effect
## flush_sync_in_effect
> Cannot use `flushSync` inside an effect
> Cannot use `flushSync` inside an effect
@ -108,6 +112,14 @@ The `flushSync()` function can be used to flush any pending effects synchronousl
This restriction only applies when using the `experimental.async` option, which will be active by default in Svelte 6.
This restriction only applies when using the `experimental.async` option, which will be active by default in Svelte 6.
## fork_discarded
> Cannot commit a fork that was already committed or discarded
## fork_timing
> Cannot create a fork inside an effect or when state changes are pending
## get_abort_signal_outside_reaction
## get_abort_signal_outside_reaction
> `getAbortSignal()` can only be called inside an effect or derived
> `getAbortSignal()` can only be called inside an effect or derived
consterror=newError(`experimental_async_fork\nCannot use \`fork(...)\` unless the \`experimental.async\` compiler option is \`true\`\nhttps://svelte.dev/e/experimental_async_fork`);
*Returnsan[`AbortSignal`](https://developer.mozilla.org/en-US/docs/Web/API/AbortSignal) that aborts when the current [derived](https://svelte.dev/docs/svelte/$derived) or [effect](https://svelte.dev/docs/svelte/$effect) re-runs or is destroyed.
*Returnsan[`AbortSignal`](https://developer.mozilla.org/en-US/docs/Web/API/AbortSignal) that aborts when the current [derived](https://svelte.dev/docs/svelte/$derived) or [effect](https://svelte.dev/docs/svelte/$effect) re-runs or is destroyed.
Rich Harris
4 weeks ago