Solve inconsistencies between "paused" mutations and "paused" queries

[crutchcorn]

While working on adding in a banner to display offline metadata to our users, I noticed that there was no easy way to fetch the paused mutations count, even while using the useMutationState hook.

This seems to be due to the fact that mutations do not have a paused status enum, despite having a isPaused state property.

This seems to not align well with how queries are handled. It might be nice to have a paused status enum on mutations as they are on queries.

References to "paused" mutations:

https://github.com/TanStack/query/blob/main/packages/query-core/src/mutation.ts#L35
https://github.com/TanStack/query/blob/main/packages/query-core/src/mutation.ts#L63-L65

Where I'd want to add support for paused status type:

• https://github.com/TanStack/query/blob/main/packages/query-core/src/mutationObserver.ts#L135
  • Add a isPaused derived state, like there is for useQuery
• https://github.com/TanStack/query/blob/main/packages/query-core/src/mutation.ts#L288
  • Add status: 'paused' to the ...state spread
• https://github.com/TanStack/query/blob/main/packages/query-core/src/types.ts#L686
  • Worth mentioning that QueryStatus is broken into FetchStatus as well. Maybe we want to do the same for mutations?

[TkDodo]

I noticed that there was no easy way to fetch the paused mutations count

Yeah, queries and mutations have diverged a bit. There is no fetchStatus because with mutations, we can and want to have an idle primary state.

What we could do is add an isPaused: boolean filter to MutationFilters. What you can do right now is to use:

predicate: mutation => mutation.state.isPaused

MutationFilters can be passed to e.g. useMutationState. We do this internally, too:

query/packages/query-core/src/mutationCache.ts

Line 166 in 5290aa1

const pausedMutations = this.#mutations.filter((x) => x.state.isPaused)

[crutchcorn]

For sure! I am now using isPaused in my codebase, so this isn't a blocker for today.

Yeah, queries and mutations have diverged a bit.

Is this something we may want to investigate for a potential Query v6 release? I'm happy to take on this work