We use redux-optimist only for the editorial workflow which is inconsistent with the rest of the codebase, and makes the code more complicated.
Also, that package is no longer maintained.

We should remove it as unless we can think of a major benefit it produces.

I've read many similar closed issues, but haven't found a good documented solution on what can I do as a consumer of ky package if I need to support e.g iOS 10.0 browser (that breaks on async/await support currently).

I agree that it's not related to ky itself, but can be a nice help for the consumers of the package.

In our particular can we don't use babel (we use typescript), so I'm

Hi, im trying to use Hashids with Typescript and Node.

import Hashids from 'hashids';

I'm not getting any Typescript errors in the editor, but Im getting this error when the code is compiled.

Using ts-node version 8.5.4, typescript version 3.7.5

`Error: No valid exports main found for 'node_modules\hashids'
at resolveExportsTarget (internal/modules/cjs/loader.js:625:9)

The goal of this project is to make itself obsolete. We should try to get browsers to fix their own styles so we don't have to work around them indefinitely.

If you want to help out:

• Pick a style in modern-normalize.css
• Go to the relevant browser issue trackers ([Chrome](http://dev.chromium.org/for-test

My goal is to have the best documentation for each provided type.

If something is unclear in any of the types, please comment or open a PR to improve it.

Some things you could contribute:

• Fix typos.
• Improved description of the type.
• More examples.
• Provide more real-world use-cases. This helps the user understand where a type might come in handy.
• Add links to relevant discu

For popular things that makes sense. Suggestions welcome, but I'm not going to add more than maybe 30. So make it worth it.

Don't do a PR. I'll add them all at once when we've decided on what to include.

While the correlation id used to match network requests with responses is always supposed to be unique at any given time, some tests were misbehaving because that wasn't the case.

Since this is one of those things that should never happen, we could add an invariant that throws a non-retriable error if this happens anyway, since something has gone terribly wrong and we won't get valid results