On 28/03/2021 22.32, Corentin Jabot via SG16 wrote:
>
>
> On Sun, Mar 28, 2021 at 10:20 PM Alisdair Meredith <alisdairm_at_[hidden] <mailto:alisdairm_at_[hidden]>> wrote:
>
> I’m still not sure I see it. The UB in [lex.phases]p2 is entirely when reverting changes
> in raw string literals, and my understanding of this paper is the only raw string reversion
> that remains is line splicing, and the presence of the newline character means that this
> reversion could never produce a UCN - hence, that particular UB is no longer possible
> as its circumstances can no longer occur.
>
>
> \ub \
> cd
>
> Can still appear in a source file.
> We could/should strike this sentence, but it's still a behavior change
> One that affects MSVC. https://godbolt.org/z/xK8j3dM1e <https://godbolt.org/z/xK8j3dM1e>
>
> I would be happy if Jens did that in his paper but I think doing it separately makes sense

With this paper, I'm trying to minimize behavior changes in general,
and in particular I'd like to leave new-line handling alone.

Jens