Robert,
First of all, thanks for the fixes in itensor. You did my work while I was
traveling. Wish all my projects were like that :-)
> we might consider creating another summation function
> which treats the index as a different symbol (by introducing
> a gensym or something like that).
I think that is generally a good idea. As our little problem demonstrated,
when you write a function intended for general use, the user may indeed
supply to that function arguments that contain a summation index internal to
that function, leading to potential conflict.
Viktor