On Sun, 30 Oct 2005 12:00:53 -0500, Douglas Gregor wrote

Boost regression test failures ------------------------------ Report time: 2005-10-30T14:55:09Z

This report lists all regression test failures on release platforms.

Detailed report:

http://engineering.meta-comm.com/boost-regression/CVS-RC_1_33_0/developer/is...

8 failures in 4 libraries: date_time (1) filesystem (2) range (3) serialization (2)

|date_time| testtz_database: intel-win32-8_1

...snip other failures...

|serialization| test_shared_ptr_132_text_warchive: intel-win32-8_1

Ok, since the date-time problem didn't self-resolve I guess we have to do something. I notice that one of the date-time and serialization tests have the same weird failure mode -- code 4 abort. This just started happening a couple days ago in the date-time test that had been working previously. Not sure when the serialization problem started. Anyway, I think we may need to go to intel on this one although I see John M. had been tracking this previously. In the meantime, I guess I'm reluctantly inclined to mark these as expected so 1.33.1 can be shipped. Ideas/suggestions anyone? Jeff ------ ...failure details.... Test output: sslapeta - serialization - test_shared_ptr_132_text_warchive / intel-win32-8_1 Report Time: Sun, 30 Oct 2005 07:03:22 +0000 Compiler output [2005-10-30 05:02:36 UTC]: "C:\Programme\Intel\CPP\Compiler80\Ia32\bin\icl" /Zm800 -nologo /EHsc -c -DBOOST_ARCHIVE_TEST=text_warchive.hpp -DBOOST_LIB_DIAGNOSTIC=1 /Z7 /Od /Ob0 /EHsc /GR /MDd /Op /DBOOST_ALL_NO_LIB=1 /wd4675 /Zc:forScope /Zc:wchar_t /Qwn5 /Qwd985 /Qwe556 -Qoption,c,--arg_dep_lookup /Qansi_alias /Qvc7.1 -D_NATIVE_WCHAR_T_DEFINED -I"C:\br\results\bin\boost\libs\serialization\test" -I"C:\Programme\Microsoft Visual Studio .NET 2003\VC7\include" -I"C:\br\boost" -Fo"C:\br\results\bin\boost\libs\serialization\test\test_shared_ptr_132_text_warchive.test\intel-win32-8_1\debug\threading-multi\test_shared_ptr_132.obj" -Tp"..\libs\serialization\test\test_shared_ptr_132.cpp" test_shared_ptr_132.cpp (0): internal error: backend signals compilation aborted for ..\libs\serialization\test\test_shared_ptr_132.cpp (code 4) Report Time: Sun, 30 Oct 2005 07:03:22 +0000 ---------------- Compiler output [2005-10-30 05:02:24 UTC]: "C:\Programme\Intel\CPP\Compiler80\Ia32\bin\icl" /Zm800 -nologo /EHsc -c -DBOOST_DATE_TIME_POSIX_TIME_STD_CONFIG -DBOOST_DATE_TIME_STATIC_LINK -DBOOST_ALL_NO_LIB /Z7 /Od /Ob0 /EHsc /GR /MDd /Op /DBOOST_ALL_NO_LIB=1 /wd4675 /Zc:forScope /Zc:wchar_t /Qwn5 /Qwd985 /Qwe556 -Qoption,c,--arg_dep_lookup /Qansi_alias /Qvc7.1 -D_NATIVE_WCHAR_T_DEFINED -I"C:\br\results\bin\boost\libs\date_time\test" -I"C:\Programme\Microsoft Visual Studio .NET 2003\VC7\include" -I"C:\br\boost" -Fo"C:\br\results\bin\boost\libs\date_time\test\testtz_database.test\intel-win32-8_1\debug\threading-multi\testtz_database.obj" -Tp"..\libs\date_time\test\local_time/testtz_database.cpp" testtz_database.cpp (0): internal error: backend signals compilation aborted for ..\libs\date_time\test\local_time/testtz_database.cpp (code 4)

|range| array: vc-6_5 vc-6_5-stlport vc-7_0

I believe this is fixed now: there was a broken compiler workaround that had been commented out on the release branch: looks like it was a mistaken commit. There is still some (new) code in boost/range/value_type.hpp that's commented out on the release branch but not on the development branch, I don't know which is correct, but it doesn't seem to be causing regressions as it is, so I suggest we leave it. Unless I'm mistaken that should fix the last failures: once the daylight-saving issues have resolved themselves. John.