Sam Verschueren

I guess the reason it throws is because a merge strategy is much more complex if it couldn't automatically resolve the conflicts. Basically every commit has to be replayed and...

I'll see what I can do!

I briefly tried looking into this because I could highly benefit from a shared cache over multiple processes (and thus reading/writing to disk). However, I don't think it's currently possible...

Just throwing in ideas here. But would it make sense if we could do `git.serializeCache(cache)` and `git.deserializeCache(cache)` 🤔 ? I'm really trying to figure out what the best approach is...

Yes I agree that it's currently somewhat tedious to pass in the same options all the time. So what you are proposing is to expose some kind of cache object...

I guess for now, it's up to the user to handle serialization/deserialization. File doesn't have to constantly be written to disk on every change. If the user is able to...

Not sure if this is what you mean but what I believe @sindresorhus means is that `.editorconfig` should always default to 2 spaces for `.yml` files and omit the settings...

Nice suggestion!

> but realized I dread getting bug reports without any stack and then having to constantly tell the user to run it with --verbose flag Realized this as well when...

Actually it's possible to assign `undefined` to `err.stack` ``` js function hello() { const err = new Error('Hello World'); err.stack = undefined; throw err; } hello(); //=> Error: hello world...