jaredwray / flat-cache

A stupidly simple key/value storage using files to persist the data
MIT License
165 stars 30 forks source link

FIX: replace circular-json with flatted #23

Closed Teamop closed 5 years ago

Teamop commented 5 years ago

Changes

ai commented 5 years ago

If everyone approved it, can we merge it and release as a patch?

royriojas commented 5 years ago

@ai, I believe this will have to be released as minor at least, since we're changing dependencies. Otherwise we will have to let the eslint team know about this been released as a patch, since this package is a dependency in file-entry-cache which in time is a dependency in eslint, just to avoid causing issues there.

platinumazure commented 5 years ago

Hi @royriojas, I'm from the ESLint team.

As long as the functionality of your package is unchanged, you can release as patch or minor if you wish. It would just help us to also have a new release of file-entry-cache with the flat-cache dependency set to use the updated version, so we can solve the deprecation warning on the ESLint side. But definitely no rush on this at all. Thanks!

royriojas commented 5 years ago

published as flat-cache@1.3.1

Thank you guys!

lenovouser commented 5 years ago

@royriojas if I understand you correctly this fix will be released with the next major version?

platinumazure commented 5 years ago

I think we need to figure out why the cache structure changed with the upgrade and see if there's a way to avoid that, so that we could publish this as a minor release. I don't think there was an intention to change the cache structure, but it just happened due to the dependency exposing a slightly different API or something.

I'm happy to try to look into that when I have time.

None of this is to speak for @royriojas, who may have a different plan entirely.

royriojas commented 5 years ago

@lenovouser yes, it should be on a new major, also we should upgrade the tests in eslint so they don't read the cache directly, because the format is an implementation detail, so it should not be something our tests rely on, I believe.

@platinumazure I'm not having enough spare time these days. But I'm sure the issue is that we should not read the cache directly from eslint tests but to find another way to test.