Date: Fri, 11 Mar 2016 19:23:32 -0600
On Friday, 11 March 2016, Jens Maurer <Jens.Maurer_at_[hidden]> wrote:
>
>
> > There is also some sentiment for getting rid of the existing “capture
> this by value” semantics that we decided aren’t what people really want
> (i.e., at some point making capture this by value the same thing as the new
> capture of *this).
> The current "surprising" case is when you do [=] as that still effectively
catches non-static member variables by reference. Personally, I'd be
happy with future C++ failing to compile for member access (including
non-static member functions) w/o explicitly capturing this or *this, but I
have no proposal at this time.
>
>
> > There is also some sentiment for getting rid of the existing “capture
> this by value” semantics that we decided aren’t what people really want
> (i.e., at some point making capture this by value the same thing as the new
> capture of *this).
> The current "surprising" case is when you do [=] as that still effectively
catches non-static member variables by reference. Personally, I'd be
happy with future C++ failing to compile for member access (including
non-static member functions) w/o explicitly capturing this or *this, but I
have no proposal at this time.
-- Nevin ":-)" Liber <mailto:nevin_at_[hidden]> +1-847-691-1404
Received on 2016-03-12 02:23:35