Philippe Canal

In this case, I think it is okay to always request a superset: ``` LD_PRELOAD=$PWD/build/lib/libaccel_final.so cmake --build build -j 8 ```

@lmoneta When the function was changed to return the 'correct' value in case the search for value is smaller than anything in the collection, `stressTMVA` fails (see https://github.com/root-project/root/pull/17179#issuecomment-2517437510). Can you...

> Returning the interator to end(), also for the case the search value is on the lower side, might be confusing. However this is the C++ way to return an...

I have not confirmed nor denied. I'll attempt reproducing it today.

I get this: ``` Celeritas version 0.6.0-rc.2.40+wip.df0b0c3bd [==========] Running 2 tests from 1 test suite. [----------] Global test environment set-up. [----------] 2 tests from StatusCheckerTest [ RUN ] StatusCheckerTest.host status:...

So several error messages, one 'assertion failed' ... and a `0` exit code.

No, it the same result 100% of the time. It is just a hiding in plain sight bug? There are 2 potential separate issues: 1. Why is the tests issuing...

> [ ... the purpose of the test is to check the error ... ] Silly me, I missed that part :(. So indeed the issue is actually that the...

As another general reminder, one challenge in the migration from the `Makefile` to a `CMakeLists.txt` is to ensure that all the implicitly generated `Makefile` based test (i.e. those without an...

As seen/hinted here: https://app.codecov.io/gh/celeritas-project/celeritas/pull/2043?dropdown=coverage&src=pr&el=continue&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=celeritas-project It seems that the issue might be triggered by `Real3{0, 0, 0}` as the lines is counted correct for `Real3{0.1, 0.1, 0.1}`