Open jlaurens opened 1 year ago
According to the Qt docs, "When comparing floating-point types (float, double, and qfloat16), qFuzzyCompare() is used for finite values." So there is no strict equality for floating point numbers anyway.
From a binary point of view, float (i.e. IEEE 754 binary32) has (slightly more than) 7 decimal digits of accuracy, so 112,28 is indeed significantly different from 112,39 from a numeric point of view (difference in the 5th digit). It would be helpful to know which particular test in test_UI
actually fails, as I couldn't find either of the mentioned numbers as hard-coded values.
Bug description:
On macOS Ventura 13.3.1(a) this test sometimes fails. The first idea is to replace QCOMPARE with a less strict equality test.
I will post a pull request soon.