First batch of algorithms with Workspace2D workspace properties changed to MatrixWorkspace. Re #1368.

Added new class MaskedProperty to mask some of the properties of ICat algorithms
Added ICat download directory to Build/Tests/Mantid.Properties file

Added YUnitLabel to Nexus Processed file. re #1095. Changed labels of curves plotted by the fit browser. re #1347

re #1367

re #1367

All Load*Nexus algorithms and LoadRaw now read the run number and store it in the sample logs of all the matrix workspaces they create. re #1264 fixes #1354

Added eventworkspaces to AddSampleLogTest and ChainedOperatorTest

Reimplemented name & alias functions as static so that an object doesn't need to be of that type as this caused problems when the libraries were loaded outside of MantidPlot. Re #1362

Added eventworkspaces to binaryOperation test

Disable MantidQt alias mechansim temporarily as it causes problems with Mantid when the libraries are in the same directory as the plugins but are run outside of MantidPlot. Re #1362

Also, added additional ParameterMap method which can get
string-value by searching a component by name.

(swapped a few int by unsigned int in Interpolation)

Add line to workspace factory create-from-parent method so that an EventWorkspace parent spawns a Workspace2D child. Re #1267.

Didn't need the extra variable, it should have used the already implemented isInitialized() function. Re #1362

Fix MantidPlot build. Windows seem to let the accidental circular dependency through. In order to be able to delete the temporary interface that has been created an initialized flag has been added to the SANSRunWindow so that its destructor does nothing if it hasn't been setup. Re #1362

re #1331 - to avoid a bug in some versions of python that was causing builds to fail, switch to using urllib rather than urllib2.

SaveCanSAS1D now stores the run number. Currently it is stored via a Python, storage via load algorithms will follow re #1264

Implemented an alias mechanism for the custom interfaces so that upon renaming them a link to the old name can still be kept. Re #1362

When CalculateTransmission is called from a SANS reduction the same monitor can't be specified as both the incident and transmitted re #1364

The changes to cow_ptr solved one issue, but not the one at hand which is that you can still wind up reading from a vector that's been/being written to by another thread. Re #1353.

Added absolute normalisation to Direct instrument convert to energy scripts and updated the GUI to use them.
Homer has been removed as the top-level interface and the ConvertToEnergy interface now sets up the GUI for Direct mode.
Also fixed a couple of warnings in GUI classes. Re #1045 #1362