Jared Parsons

> Just so I understand, this means there will be two distinct diagnostic codes, one that's an error and one that's a warning, and the compiler will choose which to...

@miyu No. This scenario is effectively breaking the `readonly`-ness of your local. At the moment LDM decided against taking the change to this level. It wasn't a hard no but...

> My guess is that it was viewed as simpler/more useful if the local RS rs = default; had the same escape scope as RS rs;. I don't believe there...

> i get the difference in scope, the ternary being one scope entirely while the if statement has separate scopes per branch, it just seems like something that should 'just'...

> Heh I noticed you had a similar "why doesn't this work" moment in the email ;) Yeah it is subtle but it does follow from how all the temp...

@RikkiGibson, @dotnet/roslyn-compiler PTAL

Have you all given any thought to how you're going to fix this problem? Been thinking about this for a few days. Most of the generators that are out there...

> Note: i think we would also need to do this in the reverse direction. Namely, instead of having generators opt-out of skeleton generation, have generators opt into them. Agree....

That example is very cachable as it comes purely from additional files. I would expect that contributes really nothing for skeleton assemblies. Yes first time we need to run it...

> And a question unrelated to the exact implementation, would params ReadOnlySpan be also permitted? Yes that is one of the types we would expand to allow with `params`. >...