slaven@rezic.de wrote: ># Failed test 'rounding' ># at t/008integerize.t line 208. It's failing due to the new correct output for sprintf("%.0f", -0.0). It is only the test that is erroneous. The processing in that file uses the module's integerize_coordinate_sets() sub to round a -0.4 value, for which it gets the entirely reasonable result of -0.0. It compares this against an expected result of 0 (implemented as an integer zero) by applying sprintf("%.0f", $_) to both and string-comparing the results. There are two possible fixes for the test. One is to change the value in its expected-result data structure to a negative zero. That would have the effect of testing that the sub under test produces the negative zero where expected. The other option is to alter the way values are canonicalised for comparison to drop the sign from zeroes. That would have the effect of ignoring the sign of zero. -zeframThread Previous | Thread Next