Re: [boost] [review] The review of Boost.DoubleEnded starts today: September 21 - September 30

Subject: Re: [boost] [review] The review of Boost.DoubleEnded starts today: September 21 - September 30
From: Hans Dembinski (hans.dembinski_at_[hidden])
Date: 2017-09-26 16:09:50

> On 26. Sep 2017, at 03:39, Zach Laine via Boost <boost_at_[hidden]> wrote:
> 
> On Mon, Sep 25, 2017 at 12:17 PM, Thorsten Ottosen via Boost <
> boost_at_[hidden] <mailto:boost_at_[hidden]>> wrote:
> 
>> For devector:
>> -------------
>> 
>> Is it it not obvious that an O(1) push_front performs better than an O(n)
>> vec.insert( vec.begin(), x ) ?
>> 
> 
> I find the use of the word "obvious" when applied to optimization to be a
> little suspect.

I wanted to say the same. I got my fair share of surprises when I benchmarked code that I expected to run faster than some other code. I can recommend Google Benchmark for micro-benchmarks, it is quite neat.

https://github.com/google/benchmark

Best regards,
Hans