Why we decided against GraphQL for local state management

Thinking about local state in the same way as external service calls just made a part of my brain say “Yeah, that’s the right abstraction”. I completely sympathise with the reasons it wasn’t worth the trouble from their perspective, though. Being able to say “Yes, this improves X dimension but we might have issues with other dimensions” shows a mature engineering org.

Tangentially related, they linked to RecoilJs which I wasn’t familiar with. The presentation on the RecoilJs site finally helped me say “Ok, I get the local state problem”. Worth a watch and reading the previous HN discussion on it [0-1]

[0] https://recoiljs.org/

[1] https://news.ycombinator.com/item?id=23181375

Re: Why we decided against GraphQL for local state management

#5

Post by jtsiskin » Fri, Aug 21, 2020, 4:32 PM UTC

egwynn wrote: post #3

“OkCupid”, not “OkayCupid"

Yes haha when I read the title I thought “how has this company not been sued by OkCupid yet??”

Re: Why we decided against GraphQL for local state management

#6

Post by gavinray » Fri, Aug 21, 2020, 4:52 PM UTC

As someone who works daily with GraphQL, it's neat to see other people talking about this.

The tools for client-side state management with GraphQL are really powerful, but require a lot more involvement than something like a standard state management store, IE an action that calls your GQL endpoint and stores the data in state.

  async function fetchUsersAction(state) {
    const users = await gql.getAllUsersQuery()
    state.users = users
    return state
  }

I generally set up a regular client-side GQL library (urql, Apollo) and use this kind of pattern for store state.

This approach has worked well for me in a number of fairly decent-sized apps, and it seems to be a pretty common one.

Re: Why we decided against GraphQL for local state management

#7

Post by Matthias247 » Fri, Aug 21, 2020, 5:04 PM UTC

I wish they would just revert their site to where it was 2 years ago instead of doing anything new.

The new version just looks like yet another Tinder clone, and doesn't resemble anymore what Okcupid was good at: Good profiles, which allowed to get an initial impression about people beyond "I like their looks".

Re: Why we decided against GraphQL for local state management

#8

Post by carabiner » Fri, Aug 21, 2020, 5:14 PM UTC

Matthias247 wrote: post #7

I wish they would just revert their site to where it was 2 years ago instead of doing anything new. The new version just looks like yet another Tinder clone, and doesn't resemble anymore what Okcupid was good at: Good profiles, which allowed to get an initial impression about people beyond "I like their looks".

I wish it was more like 6 years ago, or whenever the Match buyout was. It's become hot people clickbait like all the swipe apps now.

The entire internet is being optimized to either outrage you or sexually arouse you as quickly as possible.

Re: Why we decided against GraphQL for local state management

#9

Post by syncbehind » Fri, Aug 21, 2020, 5:17 PM UTC

carabiner wrote: post #8

Matthias247 wrote: post #7

I wish they would just revert their site to where it was 2 years ago instead of doing anything new. The new version just looks like yet another Tinder clone, and doesn't resemble anymore what Okcupid was good at: Good profiles, which allowed to get an initial impression about people beyond "I like their looks".

I wish it was more like 6 years ago, or whenever the Match buyout was. It's become hot people clickbait like all the swipe apps now. The entire internet is being optimized to either outrage you or sexually arouse you as quickly as possible.

> The entire internet is being optimized to either outrage you or sexually arouse you as quickly as possible.

I absolutely hate how accurate this is. It feels more and more that the optimization and efficiencies are making the internet less useful but more addictive.

Re: Why we decided against GraphQL for local state management

#10

Post by droopyEyelids » Fri, Aug 21, 2020, 5:18 PM UTC

carabiner wrote: post #8

Matthias247 wrote: post #7

I wish they would just revert their site to where it was 2 years ago instead of doing anything new. The new version just looks like yet another Tinder clone, and doesn't resemble anymore what Okcupid was good at: Good profiles, which allowed to get an initial impression about people beyond "I like their looks".

I wish it was more like 6 years ago, or whenever the Match buyout was. It's become hot people clickbait like all the swipe apps now. The entire internet is being optimized to either outrage you or sexually arouse you as quickly as possible.

Or both, as in the case of Netflix's new semi-child abuse documentary.