Avoid unclear diameter terminology in AABB::from_center.

[adamreichold]

• I agree to follow the project's code of conduct.
• I added an entry to rstar/CHANGELOG.md if knowledge of this change could be valuable to users.

---

[adamreichold]

Trying to address #222 (comment)

[michaelkirk]

I was expecting "width". I guess it's a little funny because width == height.

To my (US_en) eye, "distance" is more or less synonymous with diameter.

In particular: the distance from the center to the edge is not the same as the distance from the center to the corner.

Is there another popular rectangle API that uses this "distance" term? Everything I find is either min/max points or origin,width,height.

I can't find another implementation of our funny width==height constructor.

[adamreichold]

My idea behind the choice of "distance" was that this is the smallest AABB which includes all points within the given distance of the center? (This should even be true no matter which metric on R^n is chosen, shouldn't it?)

However, I am personally not particularly invested in any terminology and will happily change this if you have a clear favourite? (Or anybody else for that matter.)

To my (US_en) eye, "distance" is more or less synonymous with diameter.

At least to me, "distance" would be synonymous to radius instead of diameter.

[adamreichold]

@urschrei @michaelkirk I think with this resolved, a new release would be a good thing(tm). Of course, we need to resolve the terminology discussion here in some way. I am not particularly invested in any specific wording and would happily apply changes if one of you has a strong opinion. Until then, I think "distance" as proposed here is the least bad option so far, especially considering

this is the smallest AABB which includes all points within the given distance of the center

So maybe I add this to the doc string and we run with this?