On 30. Nov 2018, at 08:43, Olaf van der Spek
wrote: On Thu, Nov 29, 2018 at 5:58 PM Hans Dembinski via Boost
wrote: Dear Paul, dear Olaf,
On 29. Nov 2018, at 14:13, Paul A. Bristow via Boost
wrote: you need a new name for signed axis size?
int axis_size()
perhaps ;-)
both of you basically propose to have two similarly named "size" methods which return `unsigned` and `int` respectively. I really hope that you are kidding as indicated by the smileys in your messages.
Not really: http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2018/p1227r0.html
Ok, I see, thank you for the link. I am not sure if the proposal is the best solution for the show-case problem outlined in the proposal. My personal bias against having to similar "size" methods is high, since I feel that the high level interface now exposes a low-level problem, namely, how basic operators should work on signed and unsigned types. Best regards, Hans