Moved TypeConst function implementation out of header file. Hopefully should fix GCC visibility issue.
Fixed GetIDString virtual function missmatch that was causing the name not to show up when the function was called from a NiObjectRef.
Added Python-specific operator functions to Ref to prevent the wrapper from crashing on Null Refs.

Replaced _WRITE, _READ, _STRING, AND _FIXLINKS defines with the corresponding internal function call.
Replaced STANDARD_INTERNAL_FUNCTIONS with the internal function signatures.
Made my Doxygen config file use relative paths and added it to SVN so everyone can generate the same documentation for themselves.
Removed includes WZ added to ComplexShape.  These were not missing, they are up to the user to include if they wish.

Fixed the array bug by reversing the template arguments and re-doing the declarations in the Python script.

Added support for sub version 4 headers, object storage, and link method.  Some objects like NiNodes already read correctly.

Split some things that were previously part of NIF_IO.h into their own files to make SWIG wrapping work better.

Tried to implement MergeNifTrees for KF files with a NiControllerSequence root.  Untested.

Made ShortString an intrinsic type to fix problems with writing them.
Removed extra byte in HeaderString NifStream function.

Moved version constants to a separate file to clear up problems with no longer including NIF_IO.h in niflib.h.

Commented out functions written by Amorilia that I need to figure out how to fix.

Major update to python generation scripts and others.  Currently just namepace and mutable size members with appropriate generated code.

Some other DAoC related fixes to XML.

Implemented Color3 object.
Adjusted project settings to make the Maya importer link faster.

Enabled precompiled headers in MS Visual C++ project.  Also tried some tweaked optimization settings to make it compile faster.

Header and footer code now in the structs. Also small update and fix to NIF_IO HeaderString read/write.

Made niflib more verbose: prints blocks as it reads the NIF file (helps identifying buggy blocks); fixed output of bytes.

Updated license text in remaining files.  Short text in all files except niflib.h, and year updated to 2006.

Moved things around and updated xml_extract.h so that everything compiles again.  Other than NIF_Blocks.h/cpp which will be removed in the future, of course.

Implemented some of the object hierarchy the new way in nif_objects.h and started updating niflib.cpp.

Removed interfaces and moved documentation to real classes.  NIF_Blocks.cpp and niflib.cpp no longer compile.

Begun work on the new type of Attributes which store pointers to their data rather than the data itself.  Moved some code to new files to fix problems with cross-includes.
Need to decide what to do about TexSource, a struct in the external interface that should have a link in it.

Key streaming in NIF_IO.
Added xml_extract.cpp to SConstruct.
!! The extracted files are based on a modified XML, with CrossRef and Link implemented as uint, and char implemented as byte.

TODO: implement CrossRef and Link in niflib.h?

Did some research in the C++ FAQ and found out how to have vectors of objects that have no default constructor:  http://www.parashift.com/c++-faq-lite/ctors.html#faq-10.5  Therefore, LinkGroup and CrossRefGroup were removed and the NifStream functions and << operators were shifted from blk_ref and IBlock to Link and CrossRef directly.  The Python will now just have to worry about inserting the owner into the constructor of all links.

Created NifStream and << operators for blk_ref and IBlock *.  Special code will need to be generated by the Python to handle reading of links.

Implemented HeaderString class.
Removed more legacy code from NIF_IO files.