follow up PR #9274 "Mesh_3 - speedup protecting balls placement"#9366
follow up PR #9274 "Mesh_3 - speedup protecting balls placement"#9366
Conversation
off-topic, but an improvement nonetheless
|
Why does this PR modify |
Because I sneaked aafe3b5 ( |
Let me rephrase: Why are you doing this modification in this PR? A dedicated PR with some explanations would be better. Properties are not even used AFAIK in Mesh_3. |
`CGAL::sqrt` for native types (`float`, `double`, `int`...) does the same but through the abstraction of `Algebraic_foundations`.
... and make it explicit that `weight_modifier` is the square of another constant.
- fix headers, - comments, - factorize the display of a range of ids into a function
I do not understand why the code is trying to recompute that value equal to 1. Let's remove that complexity for now. I hope it was not a way to take floating point inexactness into account.
move that block of code to where it is actually used
- rename `prims` to `curve_primitives` - add a scope block, that is the "creation" of `curve_primitives` (should be an [Immediately invoked function expression] - rearrange block of comments, to distinguish the parts of the code that should be in different sub-functions [Immediately invoked function expression]: https://en.wikipedia.org/wiki/Immediately_invoked_function_expression
Co-authored-by: Sebastien Loriot <sloriot.ml@gmail.com>
|
Successfully tested in CGAL-6.2-Ic-141 |
lrineau
left a comment
There was a problem hiding this comment.
Here are the three points that currently block the integration of this draft pull-request. Those three points need to be discussed.
Once those three points are addressed, the pull-request could be merged. As least it fixes the issue #9369, and it will fix issue #9371 if we agree to document the is_parallel static data member of Sequential_tag and Parallel_tag.
|
This pull-request was previously marked with the label |
|
Successfully tested in CGAL-6.2-Ic-144 |
All fixed, right? |
Summary of Changes
a few leftovers after PR #9274
TODO
See issue #9367.
Release Management