For what's worth, I think those two should be merged into one entity, not carried forward in parallel. I don't know why the g_uncompress/mkulzma has been added in the first place if it provides a superset of g_uzip / mkuzip's functionality and should have been just the new code / CLI option for the latter. I don't know if @ray is available for the comment and svn logs provide no clue, but I see no reason for such code bloat.

I just did lot of work on improving both g_uzip / mkuzip and I did not even realize we have its almost complete twins in the tree. :(