On Tue, Nov 11, 2014 at 10:03 PM, Belcourt, Kenneth <kbelco@sandia.gov> wrote:

On Nov 11, 2014, at 7:46 PM, Rene Rivera <grafikrobot@gmail.com> wrote:

> Yes, it does look similar to output I saw. Didn't realize it would show up for others. But I don't know where it's coming from.
>
> On Tue, Nov 11, 2014 at 8:27 PM, Belcourt, Kenneth <kbelco@sandia.gov> wrote:
> Hi,
>
> Here’s a snippet of output from the nightly testing process.  I looks like someone’s dumping pointer addresses, or something, and an awful lot of them at that.  Rene, could this be related to the new testing capabilities you’re standing up?  Any ideas?

Looks like it’s coming from this commit Beman made a hour or so ago:

c41521ceadb

He seems to have added an ostream overload that is getting picked up by the testing system.

        include/boost/endian/detail/cover_operators.hpp

This one around line 100 in particular looks suspect:

      // Stream inserter
      template <class charT, class traits>
      friend std::basic_ostream<charT, traits>&
        operator<<(std::basic_ostream<charT, traits>& os, const T& x)
      {
        return os << +x;
      }

Could this be it?

I haven't been able to reproduce the problem for either Boost.Endian or Boost.Timer running on Linux with gcc 4.8.2.

What platform and compiler are you seeing the problem with?

--Beman