TimeSeriesProperty should use boost::posix_time::ptime

[Peter Peterson]

Since many of the logs for SNS will be generated at 60Hz, it is necessary to migrate the time axis for TimeSeriesProperty to use boost::posix::ptime.

[Janik Zikovsky]

(In [6009]) Refs #1384: TimeSeriesProperty now uses boost::posix_time::ptime instead of time_t. Hey, nanosecond resolution! Seemingly all the code had to be fixed to get it to work.

[Janik Zikovsky]

(In [6012]) Refs #1384: Fixed one test and disabled the MuonNexus2 one, and a line in LoadRSaveN...test. Can't load ANY muon nexus files on my system, so its hard to test. No time to track the other tonight.

[Russell Taylor]

(In [6015]) Fix tests. One apparent failure was because of a print statement to cout, the other was failing on cleaning up a temporary file so we'll do without that for the time being. Re #1384.

[Janik Zikovsky]

(In [6022]) Refs #1575 and #1384: converted AlgorithmHistory to use boost::posix_time . This probably still breaks the mantidPlot build; I'm still working it.

[Janik Zikovsky]

(In [6023]) Refs #1384: Hopefully fix MantidPlot build and windows DateAndTime.cpp build.

[Janik Zikovsky]

(In [6028]) Refs #1384: Removed direct references to boost::posix_time in MantidPlot; should fix windows MantidPlot build error.

[Janik Zikovsky]

(In [6030]) Refs #1384: More dateAndTime conversion and tests

[Janik Zikovsky]

(In [6034]) Refs #1384: Fixed the sample log window so that it can plot the logs correctly; however this required removing the fractional seconds from the column with the date, since QTIplot does not handle these.

[Janik Zikovsky]

(In [6036]) Refs #1384: Fid to qtiplot.pro to link to boost libraries (under windows)

[Janik Zikovsky]

(In [6037]) Fixes #1384: Sample logs now uses seconds since start instead of absolute date. Fixed segfault for failed filtered log plot.

[Stuart Campbell]

This ticket has been transferred to github issue ​2231