Patrick Stevens

> I know where you're coming from with this one. But wouldn't you typically catch this as an error almost immediately afterwards when you try to bind the result to...

@cartermp This is why I propose making the warning be off by default: so that users must opt into the breaking change.

I looked at this, by the way, and got something "working" (caught a couple of bugs in our existing codebase!) However, there turns out to be some information dropped by...

The PR is sort of ready, except that it has a few quirks. * `null` is often inferred as `obj` and there's not really much we can do about it...

I spoke too soon: it works only by coincidence. In fact it's broken in the presence of SRTPs (e.g. `let add x y = x + y`, where the type...

I couldn't immediately find the release process for this repository - when will this make it into a release? `tokenizers` (for versions earlier than 0.13.0) had no wheel available for...

@nojaf Fixing this issue would fix the bug we found in Fantomas. It's ultimately thanks to the `TrimStart` on https://github.com/markashleybell/MAB.DotIgnore/blob/master/MAB.DotIgnore/IgnoreRule.cs#L171 .

@dsyme Do you have any opinions on the treatment of `Object.ReferenceEquals(x, null)`, where `null` is inferred as `obj`? I don't see how to suppress the "obj inferred" warning for an...

Would you prefer me to go through the F# compiler and get it free of this warning? There will be some noise with e.g. `typedefof` and `null`.

@dsyme Would you mind please taking a look at the diff of 97ffb686e4c0e445185a1b9703c88e2cd5ff23ca and letting me know whether this is a path I should pursue? It does work as a...