Skip to content

fix: validate that the version.name config option is deterministic #14166

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 7 commits into from

Conversation

eltigerchino
Copy link
Member

@eltigerchino eltigerchino commented Aug 11, 2025

Related to gageracer/sudokusu#6 (comment)

The issue of a non-deterministic version name became more noticeable since we started using the app hash in the client script in addition to the server and the svelte config may get loaded twice.

This PR adds a validation that it's deterministic by loading the Svelte config twice and checking if the value is the same.


Please don't delete this checklist! 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
  • This message body should clearly illustrate what problems it solves.
  • Ideally, include a test that fails without this PR but passes with it.

Tests

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

Changesets

  • If your PR makes a change that should be noted in one or more packages' changelogs, generate a changeset by running pnpm changeset and following the prompts. Changesets that add features should be minor and those that fix bugs should be patch. Please prefix changeset messages with feat:, fix:, or chore:.

Edits

  • Please ensure that 'Allow edits from maintainers' is checked. PRs without this option may be closed.

Copy link

changeset-bot bot commented Aug 11, 2025

🦋 Changeset detected

Latest commit: 4e607b4

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

This PR includes changesets to release 1 package
Name Type
@sveltejs/kit 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

@Rich-Harris
Copy link
Member

I'm not too sure about this change. It's already a bit weird that we say 'the version should be deterministic' while defaulting to something that isn't deterministic; to double down on that with a test feels very much like 'do as I say, not as I do'.

Moreover, I'm not sure it's reliable. If I use Date.now() on my MacBook Pro from 2021 and load the config file twice, the timestamps are usually 2ms apart, sometimes 1ms. I bet if I was on a newer machine they would often be close enough for a false negative. Plus it's a shame to colour the function like that.

I think it might be worth us investing some time to think about what it would mean to have a deterministic version automatically. For example: I haven't thought this all the way through but could we generate the version dynamically by hashing the inputs to Vite? Maybe we need to account for externalised server-side dependencies, in which case perhaps we need to account for a lockfile/package.json in addition. Or maybe even if the code hasn't changed, we need to account for any data that changed in prerendered content, in which case perhaps we use a placeholder value at first, compute the version after prerendering, and then replace the placeholder in any outputs in which it appeared?

@eltigerchino
Copy link
Member Author

I'm not too sure about this change. It's already a bit weird that we say 'the version should be deterministic' while defaulting to something that isn't deterministic; to double down on that with a test feels very much like 'do as I say, not as I do'.

Moreover, I'm not sure it's reliable. If I use Date.now() on my MacBook Pro from 2021 and load the config file twice, the timestamps are usually 2ms apart, sometimes 1ms. I bet if I was on a newer machine they would often be close enough for a false negative. Plus it's a shame to colour the function like that.

I think it might be worth us investing some time to think about what it would mean to have a deterministic version automatically. For example: I haven't thought this all the way through but could we generate the version dynamically by hashing the inputs to Vite? Maybe we need to account for externalised server-side dependencies, in which case perhaps we need to account for a lockfile/package.json in addition. Or maybe even if the code hasn't changed, we need to account for any data that changed in prerendered content, in which case perhaps we use a placeholder value at first, compute the version after prerendering, and then replace the placeholder in any outputs in which it appeared?

Yeah, that's probably a better idea. I'll close this and create an issue for it so we can discuss more there.

@eltigerchino
Copy link
Member Author

issue here #14176

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants