fix: store forked derived values

[paoloricciuti]

Closes #17206

For some reason, despite a situation like this being the input for the second commit in #17163 we still ended up ignoring it.

This fixes it, however it introduces a new Map which I'm not too happy with...we could maybe do something with the batch itself?

Before submitting the PR, please make sure you do the following

• It's really useful if your PR references an issue where it is discussed ahead of time. In many cases, features are absent for a reason. For large changes, please create an RFC: https://github.com/sveltejs/rfcs
• Prefix your PR title with feat:, fix:, chore:, or docs:.
• This message body should clearly illustrate what problems it solves.
• Ideally, include a test that fails without this PR but passes with it.
• If this PR changes code within packages/svelte/src, add a changeset (npx changeset).

Tests and linting

• Run the tests with pnpm test and lint the project with pnpm lint

[changeset-bot]

🦋 Changeset detected

Latest commit: b769107

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 1 package

Name	Type
svelte	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[github-actions]

Playground

pnpm add https://pkg.pr.new/svelte@17212

[sillvva]

This fixes #17206

Before: (hovering link causes dev console error)
https://stackblitz.com/edit/sveltejs-kit-template-default-j4cmhsez?file=package.json,src%2Froutes%2F%2Bpage.svelte,src%2Froutes%2Ftest%2F%2Bpage.svelte

After: (no error, link works as expected)
https://stackblitz.com/edit/sveltejs-kit-template-default-ypaqodke?file=package.json,src%2Froutes%2F%2Bpage.svelte,src%2Froutes%2Ftest%2F%2Bpage.svelte

[dummdidumm]

This fails if you update and then read because there's no chance for batch_values to clear itself, example. Still, I'm somewhat inklined to merge because this is still better than what we have right now

[paoloricciuti]

This fails if you update and then read because there's no chance for batch_values to clear itself, example. Still, I'm somewhat inklined to merge because this is still better than what we have right now

I feel like we should always update inside forks...we could technically do it, but would that be wasteful and possibly performance problematic?

[paoloricciuti]

@dummdidumm last commit fixes your playground while keeping CI green (maybe we should also update the test to include the second write)...but I'm really not convinced it's the right solution...it feels wasteful (but more correct so maybe is fine?)

[dummdidumm]

ah yeah that works - made it less expensive and updated the test. Thank you!