Fix problem with switching between file loaders that have properties with same name but different default. Tidied up the logic in the same place. Removed an if clause that could never succeed. Re #2476.

Fixes #2486. Should stop the Load Dialog taking so long to appear if a previously loaded file was on a network drive.

Fixes #2839. Make the error message from the Load algorithm much clearer if it can't find an appropriate loader.

Added a setFacility()
refs #2760

Fix issues with Load Dialog and file filters. They are now defined by the Facilities extensions + a couple of frequently used others. Fixes #2602

Refs #2572: Added an initDocs() for all algorithms that sets the summary string to display in the dialog, as well as the wiki summary for possible later auto-wiki-fication of documentation.

Refs #2572: Ran into issue with un-escaped characters that my compiler did not catch. Commented out all optional messages until I can fix it.

Fix for load algorithm that would redeclare properties which already existed and reset their values if they had been set. Added a test case for this in the unit test. Fixes #2458

Correction to the Load algorithm so that it will pick up any output workspace on a sub algorithm, not just ones whose name starts with OutputWorkspace. Re #2169

Add more flexible simple API command for Load. Re #2330. Clean up for memory leak in LoadSNSNexus. Re #2394

Modified LoadEventPreNeXus so that it is compatible with the Load algorithm, Refs #2168. Also made a couple of adjustments to the layout of the MWRunFiles widget to make it better at resizing with the dialogs it is attached to. Re #2329

Implemented a custom dialog for the Load algorithm. It now dynamically generates its properties from those that the concrete loaer has, Refs #2329. Also fixed a problem with loading grouped nexus files through the Load algorithm, Re #2362

This should be everything I need for the Framework to make using the Load algorithm much simpler (including from C++). Re #2329

Proper fix for the static const declaration debacle in IDataFileChecker. The class was hard marked DLLExport rather than switching between the two depending on the IN_MANTID_API flag. This is the case in quite a lot if our code where DLLExport is used instead of the correct EXPORT_OPT_IN_{PACKAGE} (which I think need a more concise name for). Re #2329

Some minor changes to neaten up the IDataFileChecker children and to move the static contants that were not attached to anything in particular. Re #2329

Update to the Load algorithm to allow running solely to discover the concrete loader type. Also updated workspace property to respect the optional flag in other cirumstances. Re #2329

Load() now has an extra optional property, the entry number to load in a multi-period file. Currently only ISIS Nexus files support this re #2209

Rename Framework code directory to just that. Also delete Main and VS & Eclipse project file. Re #1881.

re #1538 - Refactored generic load algorithm.Implemented quick file check and file check in all load algorithms.

re#1289 - group name is removed from workspace group vector in the Workspacegroup class and changed algorithms which uses group vectors and iterates over it.
After installing visual studio 2010, unit testing is not working for all categories of algorithms on my machine, some of the unit tests may break after this check-in.

Bumped FileProperty up to API so that it can take advantage of the new FileFinder. Code has just been moved here, not reimplemented yet. Re #1398