When running ctest on my current updates for v. 2.1.1 I encountered a large number of vcs test failures due to very small differences in the Mean value displayed in the vcs plot header: e.g. 73.171410 vs. 73.171354. I don't know what would be causing these differences, but in any case I think it would be scientifically more appropriate to only display significant digits in the plot hearers: e.g. display 73.17 instead. Displaying 8 significant digits could be taken to imply much higher accuracy in the value then is actually the case.
When running ctest on my current updates for v. 2.1.1 I encountered a large number of vcs test failures due to very small differences in the Mean value displayed in the vcs plot header: e.g. 73.171410 vs. 73.171354. I don't know what would be causing these differences, but in any case I think it would be scientifically more appropriate to only display significant digits in the plot hearers: e.g. display 73.17 instead. Displaying 8 significant digits could be taken to imply much higher accuracy in the value then is actually the case.