Tanner Gooding
Tanner Gooding
This, notably, also represents another issue, in that sub response files must be relative to the working directory and not relative to themselves. This effectively breaks the ability to include...
This has some overlap with code contracts.
The behavior would likely need to be undefined in the scenario where a user compiles against an assembly that guarantees the throw but runs against an assembly where it does...
@HaloFour, I haven't finished going through the proposal yet, but I think `throw` and `never` have a small section where they do not overlap. It seems like `never` is used...
(also ignoring the fact that `d` isn't in scope for 2/3)
How is 2 definitely assigned? If the exception caught is the one thrown by the method, then `obj is Delegate d` was false and `d` is definitely not assigned. If...
I'm probably going to keep this open, since it appears the primary motivation between #538 seems to be to expand the set of things which can be treated as expressions,...
The 'issue' with that convention is that throwing the exception directly in the method prevents the JIT from inlining that method (I think CoreCLR may have relaxed this for some...
@CyrusNajmabadi, I think that this is something that cannot be fully addressed by the runtime. Not only is some analysis outside the scope of what the JIT can do (due...
> having a throw in a method does not explicitly prevent inlining Interesting, I've always seen, even in simple cases, that it prevents my method from being inlined (and thought...