Florian Angeletti
Florian Angeletti
I tend to think that both versions are subtly wrong since doing comparison with floating numbers requires to take in account the current approximation error. It is true that the...
To resume an offline discussion, backporting the header fixes to 5.0 is fine with me. However, I think it is better to keep the tests and the cpp file support...
The issue still exists, and is exotic enough to warrant being recorded.
The bug still exists and is still a witness of an exotic type system corner case.
What are the motivations behind these changes? At first glance, I find the existing code significantly more readable.
I tend to agree with closing this issue: readability is a more important concern than horizontal space optimization.
Alerts are now part of the documentation generated by both ocamldoc and odoc. Concerning the name, I hesitated with having a more generic `experimental` or `unstable` alert. However, if the...
I have the impression that the summary of the discussion is that we should go go for a generic `unstable` alert as first step, and then spend some time later...
After more discussions, there was a consensus among the maintainers that a disabled by a default `unstable` alert was acceptable whereas an active alert would be counterproductive: domains and effects...
I meant `unstable` not `unsafe`, sorry. This is strictly documenting the more unstable than usual nature of the `Effect` and `Domain` modules.