7 Oct
2017
7 Oct
'17
9:29 p.m.
AMDG On 09/26/2017 02:03 PM, Joaquin M López Muñoz via Boost wrote:
Also, the reference statement that insertion is "linear in half the size of *this [...]" is currently not true (when the favored side is full insertion takes longer than that).
Technically it is true, since "linear in half the size" is equivalent to "linear in the size."
I'd rather drop this optimization and document that shifting always happen to the direction where the least operations are needed.
In Christ, Steven Watanabe