Stefan Slapeta <stefan_nospam_@slapeta.com> writes:
I don't know if that's the same as in boost.python but the serialization library now shows up white cells, too: http://tinyurl.com/3pvzv
I don't see any problem there now. The failure has been fixed - the test_level_not_serializable is passing in sslapeta runs. We don't keep history - so there is no way for me to go back and see what happened.
It's remarkable that all these are results of compile or compile-fail tests and the tests would succeed if they are reported correctly. I attach an excerpt of my bjam.log: please have a look at it, it seems to me that everything is correct except the interpretation of the results.
Everything looks OK there. For sure it is not the same problem as Boost.Python has. There are more missing test results in other libs, most of I've looked at are just new tests not yet picked up by some regression runs - the artifact of merging of regression runs made on different CVS snapshot. I cannot explain the missing results in Martin's graph library intel-7.1-linux, intel-8.0-linux toolsets. If they don't disappeat tomorrow, I will probably ask Martin to send us the whole bjam.log (or I will modify the regression run scripts to include it in uploaded archive). -- Misha Bergal MetaCommunications Engineering