On 25.08.2019 16:39, Joerg Sonnenberger wrote: > On Sun, Aug 18, 2019 at 08:58:15PM -0000, Valery Ushakov wrote: >> Joerg Sonnenberger <joerg%bec.de@localhost> wrote: >> >>> On Sun, Aug 18, 2019 at 05:05:33PM +0200, Kamil Rytarowski wrote: >>>> Ping? Can we switch from intptr_t to void*? >>> >>> Can we just go back to the original state? >> >> Anything wrong with the templated solution for c++11 I posted a few >> days ago, keeping the original state for older c++ versions (since >> nullptr that, afaiu, started all this is only in 11). > > I'm kind of waiting for the next obscure fallout. The original version > has easy to understand handling. It works without cast for one common > use case (passing indices) and needs a reinterpret_cast in the other > (passing pointers). Now we are trying to do both and play whack-a-mole. > There is no technical reason really for prefering void * and in fact a > couple of them for prefer a plain integer. Lack of potential trap > representations for example. I maintain that going back to the original > state is actually the better situation. > > Joerg > That patch had no handling of reinterpret_cast<>. I want to see the original state of void* so all casts will be unneded. Since B lang times we used pointers and integers as the same data type. They were so indistinguishable that C lang invented unsigned to make any difference.
Attachment:
signature.asc
Description: OpenPGP digital signature