Herb Sutter

Editors call: We agree, it should be a string literal. We'll fix the Guidelines to show that.

Editors call: Our current policy is that GSL's `gsl::span` follows the `std::span` interface exactly plus adds the ability to guarantee bounds checking. We are trying to avoid adding more convenience...

Thanks Anna, some more context Anna provided separately: MSVC understands [[gsl::suppress(something.x)]] Clang understands [[gsl::suppress(“something.x”)]] GCC doesn't understand them on a line inside a function

Note that this is specified to be quoteless in [In.force](http://isocpp.github.io/CppCoreGuidelines/CppCoreGuidelines#inforce-enforcement)

Editors call: Thanks! Sorry for the lag -- this has been addressed with #2178.

Yes, yes, and possibly yes, respectively. This question happened to come up in a recent conversation among the editors, and I heard sympathy for the position that chaining might be...

Resolved with the merged change.

Yes, I agree this is a bug, lazy initialization shouldn't be allowed in a loop (loops can be entered zero times).

> it would be nice to only error if the variable is actually used after the branch. It would be possible to make this work, but wouldn't that be equivalent...

OK, fixed. The following errors are now diagnosed: ```cpp error1: () = { i: int; while true { i = 42; // ERROR: can't initialize i in a loop }...