Three autotests fails during validation on Windows 10. The errors appear to be exit codes that are out of the expected range when exceptions are caught.
ERROR [INIT] : c:\gridlabd\trunk\assert/autotest/test_complex_assert_imaginary_err.glm exception unexpected, code 127 (shell failed) in 6.9 seconds ERROR [INIT] : c:\gridlabd\trunk\assert/autotest/test_complex_assert_real_err.glm exception unexpected, code 127 (shell failed) in 5.5 seconds ERROR [INIT] : c:\gridlabd\trunk\assert/autotest/test_enum_assert_err.glm exception unexpected, code 127 (shell failed) in 6.4 seconds
Important note: the only difference between the Windows 7 and Windows 10 validation tests are that the Windows 10 test was performed immediately after the Windows 7 system was upgraded to Windows 10. Rebuilding gridlabd made no difference on the validation test on Windows 10.
The assert autotests all pass fine for me on a trunk build of GridLAB-D on a Windows 10 machine (native install, not a migration/upgrade). Encountering a couple other autotest issues, that I'll document here, just in case they are related. They may be more "user error" or an issue with my environment.
1 -- executing --validate in the base folder runs no autotests. Individual projects/modules must be changed to first, then validated.
2 -- Prior to updating this morning, core autotests all passed fine. After the update (coming from approximately r5416), gridlabd.h does not appear to be being found by the runtime compiler autotests
These problems do not block the knothole process itself so I suggest moving this to later milestone