Drew Powers

Love all these thoughts. I don’t disagree with any of it! > the openapi-fetch benchmarks currently mostly benchmark msw. This could somewhat explain the high fluctuations. That is a _very_...

Just removed MSW from the benchmarks and that indeed seemed to contribute heavily toward the fluctuations. I’m sure there will still be a little flaking, as is normal, but it...

This is tricky because there’s no way to store that information in a way that’s currently-accessible by the static types. You’ve stated a clear problem that all users of this...

Ah @ghchaosfae I did see that error as well! But I didn’t know if I was moving “forward” or “backward” so-to-speak. Wanted to find the most minimal issue, and work...

@ghchaosfae great find! Yes that’s a good workaround, but would love to keep using `projectService` if possible, because like you said that’s what ESLint recommends (and seems to be the...

As you discerned, this error does appear when there’s an actual error. It’s tough to control the output for what TypeScript shows beyond throwing an error. There are no known...

I think this is a breaking change, but one most seem to be in favor of. Love the direction of this so far! Please let us know if you need...

@gzm0 I completely agree—`parseAs` is not an ideal solution; it’s just a bandaid to not introduce performance issues. Many people find it hard to use, and want it to be...

> Have we considered simply returning the fetch Response object with a more narrowed type directly? That was one of the earliest decisions I made to just save hundreds of...

@Ginxo your issue is likely valid! But we haven’t had time to review it yet. We review things when we are able to. And your issue does not have a...