Subject: Re: [boost] [Hana] Announcing Hana's formal review next week (June 10th)
From: Louis Dionne (ldionne.2_at_[hidden])
Date: 2015-06-12 10:23:50
Joel de Guzman <djowel <at> gmail.com> writes:
>
> On 6/12/15 3:39 AM, Louis Dionne wrote:
> > Instead, in a heterogeneous setting, you probably want it to return false_.
> > With Fusion/MPL, this is not a problem because they use `std::is_same` to
> > perform comparisons. However, `std::is_same` performs a very shallow
> > comparison and only allows comparing types. Instead, Hana uses a more
> > general system which also allows comparing values, but we have to return
> > false_ for unrelated objects, a bit like `std::is_same` returns
> > `std::false_type` on unrelated types.
>
> I might not be understanding this correctly, but Fusion == does not use
> std::is_same. It does value comparison much like tuples.
Joel,
Indeed, my answer was not very clear. What I meant is that Fusion uses
std::is_same for the fusion::{find, filter, remove, erase_key, ...}
algorithms. In addition, Fusion defines the usual comparison operators
for sequences (==, !=, ...), like you correctly pointed out.
Instead, Hana uses the same comparison function (`equal`) for all these
operations. However, for some algorithms, the comparison operator must
return an IntegralConstant.
Regards,
Louis