Optimistic Updates

Optimistic updates enable highly responsive and fast interfaces by avoiding network wait times. An update is optimistic by assuming the network is successful. In the case of any errors, Rest Hooks will then roll back any changes in a way that deals with all possible race conditions.

Partial updates

One common use case is for quick toggles. Here we demonstrate a publish button for an article. Note that we need to include the primary key (id in this case) in the response body to ensure the normalized cache gets updated correctly.

ArticleResource.ts

import { MutateEndpoint, SchemaDetail, AbstractInstanceType } from 'rest-hooks';
import { Resource } from '@rest-hooks/rest';

export default class ArticleResource extends Resource {
  readonly id: string | undefined = undefined;
  readonly title: string = '';
  readonly content: string = '';
  readonly published: boolean = false;

  pk() {
    return this.id;
  }

  static partialUpdate<T extends typeof Resource>(
    this: T,
  ): MutateEndpoint<
    (p: Readonly<object>, b: Partial<AbstractInstanceType<T>>) => Promise<any>,
    SchemaDetail<Readonly<AbstractInstanceType<T>>>
  > {
    return super.partialUpdate().extend({
      getOptimisticResponse: (snap, params, body) => ({
        // we absolutely need the primary key here,
        // but won't be sent in a partial update
        id: params.id,
        ...body,
      }),
    });
  }
}

PublishButton.tsx

import { useController } from 'rest-hooks';
import ArticleResource from 'ArticleResource';

export default function PublishButton({ id }: { id: string }) {
  const { fetch } = useController();

  return (
    <button
      onClick={() =>
        fetch(ArticleResource.partialUpdate(), { id }, { published: true })
      }
    >
      Publish
    </button>
  );
}

Optimistic create with instant updates

Optimistic updates can also be combined with immediate updates, enabling updates to other endpoints instantly. This is most commonly seen when creating new items while viewing a list of them.

Here we demonstrate what could be used in a list of articles with a modal to create a new article. On submission of the form it would instantly add to the list of articles the newly created article - without waiting on a network response.

ArticleResource.ts

import { MutateEndpoint, AbstractInstanceType } from 'rest-hooks';
import { SchemaDetail, Resource } from '@rest-hooks/rest';

export default class ArticleResource extends Resource {
  readonly id: string | undefined = undefined;
  readonly title: string = '';
  readonly content: string = '';
  readonly published: boolean = false;

  pk() {
    return this.id;
  }

  static create<T extends typeof Resource>(
    this: T,
  ): MutateEndpoint<
    (p: Readonly<object>, b: Partial<AbstractInstanceType<T>>) => Promise<any>,
    SchemaDetail<Readonly<AbstractInstanceType<T>>>
  > {
    const list = this.list();
    return super.create().extend({
      getOptimisticResponse: (snap, params, body) => body,
      update: (newResourcePk: string) => ({
        [list.key({})]: (resourcePks: string[] = []) => [
          ...resourcePks,
          newResourcePk,
        ],
      }),
    });
  }
}

CreateArticle.tsx

Since the actual id of the article is created on the server, we will need to fill in a temporary fake id here, so the primary key can be generated. This is needed to properly normalize the article to be looked up in the cache.

Once the network responds, it will have a different id, which will replace the existing data. This is often seamless, but care should be taken if the fake id is used in any renders - like to issue subsequent requests. We recommend disabling edit type features that rely on the primary key until the network fetch completes.

import { useController } from 'rest-hooks';
import uuid from 'uuid/v4';
import ArticleResource from 'ArticleResource';

export default function CreateArticle() {
  const { fetch } = useController();
  const submitHandler = useCallback(
    data =>
      // note the fake id we create.
      fetch(ArticleResource.create(), { id: uuid(), ...data }),
    [create],
  );

  return <Form onSubmit={submitHandler}>{/* rest of form */}</Form>;
}

Optimistic Deletes

Since deletes automatically update the cache correctly upon fetch success, making your delete endpoint do this optimistically is as easy as adding the getOptimisticResponse function to your options.

We return an empty string because that's the response we expect from the server. Although by default, the server response is ignored.

import { Resource, SimpleResource } from '@rest-hooks/rest';
import { MutateEndpoint } from 'rest-hooks';

export default class ArticleResource extends Resource {
  readonly id: string | undefined = undefined;
  readonly title: string = '';
  readonly content: string = '';
  readonly published: boolean = false;

  pk() {
    return this.id;
  }

  static delete<T extends typeof Resource>(
    this: T,
  ): MutateEndpoint<(p: Readonly<object>) => Promise<any>, schemas.Delete<T>> {
    return super.delete().extend({
      getOptimisticResponse: (snap, params, body) => params,
    });
  }
}

Optimistic Transforms

Sometimes user actions should result in data transformations that are dependent on the previous state of data. The simplest examples of this are toggling a boolean, or incrementing a counter; but the same principal applies to more complicated transforms. To make it more obvious we're using a simple counter here.

class CountEntity extends Entity {
  readonly count = 0;

  pk() {
    return `SINGLETON`;
  }
}
const getCount = new Endpoint(
  () => fetch('/api/count').then(res => res.json()),
  {
    name: 'get',
    schema: CountEntity,
  },
);
const increment = new Endpoint(
  async () => {
    const body = JSON.stringify({ updatedAt: Date.now() });
    return await (
      await fetch('/api/count/increment', {
        method: 'post',
        body,
        headers: {
          'Content-Type': 'application/json',
        },
      })
    ).json();
  },
  {
    name: 'increment',
    schema: CountEntity,
    sideEffect: true,
    getOptimisticResponse(snap) {
      const { data } = snap.getResponse(getCount);
      if (!data) throw new AbortOptimistic();
      return {
        count: data.count + 1,
      };
    },
  },
);

function CounterPage() {
  const { fetch } = useController();
  const { count } = useSuspense(getCount);
  const [clickHandler, loading, error] = useLoading(() => fetch(increment));
  return (
    <div>
      <p>
        Click the button multiple times quickly to trigger the race condition
      </p>
      <div>
        {count} <button onClick={clickHandler}>+</button>
        {loading ? ' ...loading' : ''}
      </div>
    </div>
  );
}
render(<CounterPage />);

Live Preview

Loading...

Store▶

Try removing getOptimisticResponse from the increment Endpoint. Even without optimistic updates, this race condition can be a real problem. While it is less likely with fast endpoints; slower or less reliable internet connections means a slow response time no matter how fast the server is.

The problem is that the responses come back in a different order than they are computed. If we can determine the correct 'total order', we would be able to solve this problem.

Without optimistic updates, this can be achieved simply by having the server return a timestamp of when it was last updated. The client can then choose to ignore responses that are out of date by their time of resolution.

Tracking order with updatedAt

To handle potential out of order resolutions, we can track the last update time in updatedAt. Overriding our useIncoming, we can check which data is newer, and disregard old data that resolves out of order.

We use snap.fetchedAt in our getOptimisticResponse. This respresents the moment the fetch is triggered, which is when the optimistic update first applies.

class CountEntity extends Entity {
  readonly count = 0;
  readonly updatedAt = 0;

  pk() {
    return `SINGLETON`;
  }

  static useIncoming(existingMeta, incomingMeta, existing, incoming) {
    return existing.updatedAt <= incoming.updatedAt;
  }
}
const getCount = new Endpoint(
  () => fetch('/api/count').then(res => res.json()),
  {
    name: 'get',
    schema: CountEntity,
  },
);
const increment = new Endpoint(
  async () => {
    const body = JSON.stringify({ updatedAt: Date.now() });
    return await (
      await fetch('/api/count/increment', {
        method: 'post',
        body,
        headers: {
          'Content-Type': 'application/json',
        },
      })
    ).json();
  },
  {
    name: 'increment',
    schema: CountEntity,
    sideEffect: true,
    getOptimisticResponse(snap) {
      const { data } = snap.getResponse(getCount);
      // server already has this optimistic computation then do nothing
      if (!data) throw new AbortOptimistic();
      return {
        count: data.count + 1,
        updatedAt: snap.fetchedAt,
      };
    },
  },
);

function CounterPage() {
  const { fetch } = useController();
  const { count } = useSuspense(getCount);
  const [n, setN] = React.useState(count);
  const [clickHandler, loading, error] = useLoading(() => {
    setN(n => n + 1);
    return fetch(increment);
  });
  return (
    <div>
      <p>
        Click the button multiple times quickly to trigger the potential race
        condition. This time our vector clock protects us.
      </p>
      <div>
        Network: {count} Should be: {n}
        <br />
        <button onClick={clickHandler}>+</button>
        {loading ? ' ...loading' : ''}
      </div>
    </div>
  );
}
render(<CounterPage />);

Live Preview

Loading...

Store▶