Renamed the instance field of the ArchiveFile into diskInstance\nAdded a comment for the lastUpdateTime field of ArchiveFile\nWord wrapped comments of the data structures

Added comments in the EOS interface doc.\nRenamed MountGroup into MountPolicy and removed MountPolicy (adapted SQL code).\nRemoved amount of data to warrant a mount from LogicalLibrary.

Added last update time in the ArchiveFile and split the updatefileinfo command in two: updatefileinfo (run by EOS during resynchronization process) and updatefilestorageclass (run by EOS on behalf of a user who wants to change the storage class of a file)

Changed the documentation fro the eos interface and the instance crosstalk, renamed User into Requester, refactored all the SQLITE code to have a GROUPNAME column in the requester, changed the tests

Re-enabled the helgrind tests that triggered it.

relies on global variables managing files which were deleted on child process exit.
Helgrind is currently disabled for the multiprocess unit test as it makes it fail.

Added silencing of valgrind information for child processes.
Added protocol buffers cleanup at the end of unit tests.

Added ProcessManager class which owns the main loop.
Fixed a bug when closing the wrong side of the socketpair in SocketPair.
Minor fixes to tape daemon code.

Changed the interface to EOS doc and changed the type of the first parameter of catalogue.getArchiveFiles from uint64_t to string

Added a VO explanation and requirements document, and removed the last bits of dr_instance from the help texts

Finished renaming eosid to diskid and move drInstance out of drData and renamed to "instance" simply

we have out the mount policy parameters in the scheduler constructor temporarily we will remove them once we know where to put them