Masaki Moriguchi (a.k.a. Michel Morin)

I also encountered this when running a benchmark in [joaquintides/fxa_unordered](/joaquintides/fxa_unordered). Turning off the special handling of AArch64 improves benchmark results (on Apple M1 Pro) from ``` [uint64.cpp] absl::node_hash_map: 5208 ms,...

Agreed. For "dirty" soupy triangles and point clouds, continuous signing might be a viable option. Adding a new signed distance type `SIGNED_DISTANCE_TYPE_FAST_WINDING_NUMBER_DIRTY` looks easy.

The STI doc ([mirror site](https://www.boost.org/sgi/stl/advance.html)) does not use interval notation there and provides a concise description (which is not divided into positive and negative cases): > Preconditions > - Every...

And `transform_view` has the same problem in general...?

[Ruby's way](http://ruby-doc.org/core/Array.html) might be interesting. It has - `flatten`: flattens (all levels of) nested ranges. - `flatten(level)`: flattens nested ranges (up to the specified level). - `join(delimiter)`: similar to range-v3's...

> I have never needed to flatten more than one level, and certainly not an arbitrary number of levels. My gut says that we don't really need that. Agreed. Flatten-one-level...

> but with a runtime level how would you specify the associated types of the result range? Ah, good point! Its `value_type` gets inpractical. Thanks.

> A model of `input_iterator` may fail to meet the *Cpp17InputIterator* requirements by being move-only, or by not being *Cpp17EqualityComparable*. That's right, ouch... > We should use conccepts to decide...

> A model of `input_iterator` may fail to meet the *Cpp17InputIterator* requirements by being move-only, or by not being *Cpp17EqualityComparable*. So, this means that the specification of `iterator_category` for many...

This is now [LWG 3301](https://cplusplus.github.io/LWG/issue3301).