24 Oct
2011
24 Oct
'11
8:33 p.m.
On Mon, Oct 24, 2011 at 7:20 PM, Peter Dimov <pdimov@pdimov.com> wrote:
Hi Olaf,
Yes, I can review and apply patches. Regarding #4637, the patch makes the following assumptions: (1) that an array of size 0 should create a NULL pointer, which is not what new T[0] does, and (2) that the array should be value-initialized, which may not be what the majority of the users want.
Valid points. Why didn't you raise them in the tracker?
These design decisions should be discussed here among the scoped_array users and settling them is a prerequisite for this patch to go in.
Incidentally, you'll need to make these same choices for make_shared_array as well.
I know. I'll start a thread. Olaf