Ed Lui

This is a blocker for us. We're rolling our own post-deploy data migration scripts. Having something similar to Rails's `ActiveRecord` and `data-migrate` would be fantastic. In the mean time, we're...

Hey @ijjk and @shuding , as of this canary release and more formally in `13.0.2`, our Serverless functions now exceed the deployable threshold of 50MB in our private repo. Any...

@balazsorban44 Thanks for the reply! Indeed, the problem is with API routes, as @JoeKarow also notes. Rolling back to 13.0.1 solved it for us as well. Notice that @JoeKarow is...

Hey @balazsorban44 @JoeKarow I've tried to update a barebones repro with some of the signals between my private repo where this is happening and @JoeKarow 's public repo. I can't...

> Hey! Can you share a list of asset/dependency sizes when deploying 13.0.1? Then we can tell which part actually became larger in the newer version. Here you go, @shuding...

@JoeKarow no, I've removed it. Though maybe the results were cached between builds. Let me re-run without a deploy cache.

@JoeKarow 🤦 Yes, it was nuked thanks to the build cache. An uncached build exceed the threshold: 13.0.3 - uncached, exceeding threshold ``` Large Dependencies Uncompressed size Compressed size --...

Here's the output from `npm explain @swc/core`. Doesn't seem to be anything that should be pulling it in that's not `dev` `npm explain @swc/core` ``` @swc/[email protected] dev node_modules/@swc/core dev @swc/core@"1.3.14"...

> Hi, any chance someone who has a repo that consistently fails with this above `v13.0.1` could try the patch from #43069 via `yarn add next@https://files-99l7vkuaq-ijjk-testing.vercel.app` and confirming whether this...

> we have landed a new `experimental.outputFileTracingIgnores` config that allows adding custom globs to ignore from the traces which can allow making the function size more controlled when certain dependencies...