> On July 30, 2013, 4:59 p.m., Benjamin Hindman wrote: > > 3rdparty/libprocess/3rdparty/stout/include/stout/protobuf.hpp, line 165 > > <https://reviews.apache.org/r/12951/diff/1/?file=327960#file327960line165> > > > > Yes, my intuition tells me we'll want to omit them if not present.
The problem is that people will rely on the JSON format. For example, with resource monitoring information, users may rely on "cpus_user_time_secs" being present in the output. However, it is sometimes not available (e.g. on OS X with insufficient permissions) and will thus be not set in the protobuf message. This means they'll have to know which JSON fields are optional and handle them being omitted. So for now I'll leave as is. - Ben ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/12951/#review24247 ----------------------------------------------------------- On July 25, 2013, 10:05 p.m., Ben Mahler wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/12951/ > ----------------------------------------------------------- > > (Updated July 25, 2013, 10:05 p.m.) > > > Review request for mesos, Benjamin Hindman and Vinod Kone. > > > Repository: mesos-git > > > Description > ------- > > JSON::Number was output using a precision of 10 digits, whereas double > precision guarantees 15 digits of accuracy. > > Digits 16 and 17 may be bogus, so kept it at 15. See: > http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2006/n2005.pdf > > > Diffs > ----- > > 3rdparty/libprocess/3rdparty/stout/include/stout/json.hpp > e2ef0d3a50e698192ba182d53dbcc9186542df06 > 3rdparty/libprocess/3rdparty/stout/include/stout/protobuf.hpp > bd5bc7372fafb1aba757ac08acceeb9689e2fc09 > 3rdparty/libprocess/3rdparty/stout/tests/json_tests.cpp > c582d7084747f4f1804cc554cc544c8bc02602ac > > Diff: https://reviews.apache.org/r/12951/diff/ > > > Testing > ------- > > make check > > > Thanks, > > Ben Mahler > >
