Guy Bedford

I was under the impression previously that Webpack used the `"module"` condition to avoid the dual package hazard by explicitly matching that condition for both `require` and `import` since Webpack...

@evanw it would be very interesting if you could expand somewhere on the exact symbol naming technique you converged on here. I'm sure it will make sense looking at the...

> This seems possibly related to this line of code? > https://github.com/snowpackjs/rollup-plugin-polyfill-node/blob/main/src/index.ts#L8 If your build was exposing `import '\0node-resolve:empty.js'` I don't think the bug is sanitization, but that that is...

Ahh, I see that the `files` field is in fact loaded with SystemJS.import. This seems like a non-standard thing to do with Karma though as it is changing the meaning...

If you look at the RequireJS instructions (https://karma-runner.github.io/0.13/plus/requirejs.html) they advocate using the `included: false` option in files. I do think it would make sense to follow the precedent set there....

@rolaveric I've got something just about working at https://github.com/guybedford/karma-systemjs-experiment, exploring some ideas which would complement what is happening in this repo. I'm aiming to do a writeup on the principles...

Yes it's a different approach, but there could certainly be splicing done between the two. I guess it could be a major release if you wanted to do a replacement,...

I like the premise here. If `module.json` is to take over all the responsibilities of `package.json` then there is a risk of a lot of duplication though as interim packages...

Did you consider just using an environment variable for this like `NODE_PROJECT_ROOT` or something?

This feature comes down to specifying a single path only within which Node.js should execute code - I think rather than a file signifier, we should just have an out-of-band...