optout

> For reference https://maflcko.github.io/b-c-cov/test_bitcoin.coverage/src/chain.cpp.gcov.html is pretty well covered otherwise, but these tests can help with the specifics. Thanks for the link! Indirectly it is tested extensively, as the skip mechanism...

Review comments addressed, mostly minor: - Reduce the number of iterations to 10000 (from 30000), for faster test execution. - Add comments on choice of test inputs. - Use BOOST_REQUIRE_EQUAL...

@sipa, as this is triggered by #33515 of yours, I would kindly ask for a review.

Changes applied: - Fix lint issue - Added check on the median as well - some minor formatting

CI failed, limits of the 3rd for cycle were off; fixed ```diff int count_diff_smaller_16 = 0; int max_diff = 0; int total_diff = 0; - for(int i{1}; i < n;...

Pushed minor fixes after comments: - simplify test assertions - simplify computing of total/avg and max (no for loop) - formatting

Minor nits applied git diff ea998e18f6d64f485d4e2d85a1028474fb35120a e2e25f4947e65869b014483be70b764a62bd0df7 ```diff @@ -114,7 +114,7 @@ BOOST_AUTO_TEST_CASE(skip_height_properties_test) for(int i{1}; i < n; ++i) { auto new_index = std::make_unique(); new_index->pprev = block_index.back().get(); - BOOST_CHECK(new_index->pprev); +...

Very minor touch: moving of a curly nit-fix, rebase (noop); mostly to get CI to run again (had some apparently temporary issue). (e2e25f4947e65869b014483be70b764a62bd0df7 -> 1a0423cf25bb60f98a29e06c39b7db2998860cd9)

> I don't think it tests anything useful: > > * If it wanted to test the exact behavior of the skip-height pointer building, it could test that directly. No...

I redid the test, following @sipa's remark. - The test on skip height properties -- median, avg -- is dropped - A new test is added, that tests the skip...