silverwind

Running both on the same port is a huge hack, and I think rightly so. A tcp port should only serve one protocol, it's how the internet was build. There's...

On Windows, I've found `graceful-fs` to be pretty much essential to avoid all kinds of strange errors that won't happen on other platforms.

https://github.com/tachyons-css/tachyons-sass/pull/63

Maybe we can actually remove altering `white-space` altogether.

As for syntax, I suggest using import attributes, see https://github.com/tc39/proposal-defer-import-eval/issues/5#issuecomment-2038863806.

To reduce code churn, maybe it should have an option to limit it to attribute selectors as indicated by `[` and `]` in the surrounding strings. Patterns like below are...

Here's a workaround that shuts up typescript. It's not much of use because it does not validate `rules`. In fact I'm not sure how `rules` validation could even be made...

Just had this false-positive with `this` in a class show up after switching eslint's parser from `espree` to `@typescript-eslint/parser`. So the bug looks to be parser-related.

> And it's generally an anti-pattern to use packages named the same a built-in packages as it makes it unclear to people reading the code that it's not actually a...

I would ask the authors of `assert` to publish to a unambiguous package name and then use that. `punycode` is also available as unambigous [`punycode.js`](https://www.npmjs.com/package/punycode.js) module.