I agree. I find this snippet interestingly incorrect. > with very few regressions, up until now. the idea that this is a regression is silly. you can't have a

I agree. I find this snippet interestingly incorrect. > with very few regressions, up until now. the idea that this is a regression is silly. you can't have a regression unless old code is slower as a result. which is clearly not the case. its just a less than ideal outcome for generics. which will likely get resolved.