Add a note about word size to the MersenneTwisterEngine changelog entry

[WebDrake]

This was an oversight when finalizing #5011.

[dnadlinger]

Is there an implementation in the C++ standard? Otherwise, the wording here is slightly confusing.

[WebDrake]

Ah, good point. Obviously I compared against an implementation of the C++11 standard. However the standard itself does mandate the expected behaviour (which is really only re-stating the expected behaviour of the algorithm itself). See p. 899 (numbers; p. 913 in your PDF browser) of http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2012/n3337.pdf.

[WebDrake]

I'll rephrase to an implementation of the C++11 standard.

[WebDrake]

Amused to see that code coverage metrics can change because of a tweak to a changelog entry ... :-P

[WebDrake]

Updated to improve the wording:

diff --git a/changelog/std-random-MersenneTwisterEngine.dd b/changelog/std-random-MersenneTwisterEngine.dd
index 7d97481..0fa3d23 100644
--- a/changelog/std-random-MersenneTwisterEngine.dd
+++ b/changelog/std-random-MersenneTwisterEngine.dd
@@ -5,7 +5,7 @@ tempering parameter `d` and the initialization multiplier `f`).
 Handling of the word size `w` has been fixed so that the generator will
 now properly handle cases where it is less than the number of bits in
 the chosen `UIntType`.  This has been validated against the behaviour
-of the C++11 standard implementation.
+of a widely-used implementation of the C++11 standard.
 
 For anyone using the standard template instantiation `Mt19937` this will
 have no noticeable effect.  However, these will be breaking changes for

[WebDrake]

... and now it's somehow gained code coverage, despite not touching any actual code :-P

There must be a rounding error out there somewhere ...

@JackStouffer assuming you're happy with the new wording, dlang-bot seemed to dislike your tag ;-)

[wilzbach]

dlang-bot seemed to dislike your tag ;-)

Well, that's a feature that auto-merge gets disabled once the approved diff changes as otherwise I would be a potential hole to slip in unreviewed code ;-)
@JackStouffer simply re-add it then

[WebDrake]

Well, that's a feature

Ah, good point :-) Thanks for the clarification!

[wilzbach]

... and now it's somehow gained code coverage, despite not touching any actual code :-P
There must be a rounding error out there somewhere ...

The problem is that CircleCi merges the PR with upstream/master before it runs the coverage tests. Thus a change in master in the meanwhile (aka from the ref your PR is based on) will yield a different project result. I contacted the guys from CircleCi to accommodate this behavior in their pipeline.