Don Syme
Don Syme
This is indeed fixed by #11681
I'm closing as a duplicate of this: https://github.com/dotnet/fsharp/issues/12038
@NinoFloris Do you know if this is fully covered by #13673?
@vzarytovskii @NinoFloris I'll need to review this very carefully. I'll be back at work properly next week and suggest we do it then
> if the compiler runs into a conversion that it cannot properly do it just spews box unbox.any due to the coerce. I assume this only happened because the type...
This is looking good - a couple of minor nits
> The downside for this implementation in particular is ... I think we won't do this as part of this PR, partly because emitting the warning (even optional) would take...
@NinoFloris pointed out privately that this does emit the FS3389 off-by-default warning in an additional case, seen here https://github.com/dotnet/fsharp/pull/13673/files#diff-0020979b647768e28019703c22e81771d97b9e9b26eb7c3cb518c692e6596beaR264 [2:25](https://fsharp.slack.com/archives/DJC7FM64X/p1661174715549369) This is ok and I consider this a bug fix...
@albert-du This is fantastic work We need some more testing * We have to test loops near Int32.MaxValue. These can be really difficult to get right since adding the step...
> @dsyme It doesn't seem too much related, but maybe you could also take a look at https://github.com/dotnet/fsharp/issues/7931 or https://github.com/dotnet/fsharp/issues/5795 while working on this PR, since they're also about issues...