Re: [boost] High number of integer_sequence/index_sequence implementations

Subject: Re: [boost] High number of integer_sequence/index_sequence implementations
From: Peter Dimov (lists_at_[hidden])
Date: 2019-01-14 20:41:36

Stephan T. Lavavej wrote:
> [Nikita Kniazev]
> > Bunch of Boost libraries have their own integer_sequence/index_sequence 
> > implementation.
> > Wouldn’t it be better if there was 
> > boost::integer_sequence/boost::index_sequence?
>
> And how many of them are powered by __make_integer_seq (when available) 
> for throughput?

I haven't bothered because for N=8192 the difference is between 1.0s and 
0.7s and doing anything nontrivial with 8192 is probably going to fail 
anyway, but I might do so if only to be able to answer in the affirmative.