Matt Fishman

Currently NDTensors is still more of an "internal" package for use within ITensors, though it is a lot more stable now than when I wrote that warning initially. It implements...

That is a neat idea. I guess if an Index doesn't have a `dim=*` tag it is treated as O(1)?

Ah I didn't read carefully enough, that makes sense to report the actual dimensions of indices that don't have a "dim=*" tag. Just a note that right now, tags do...

> @mtfishman do you happen to know a good way to set up a macro environment sort of thing for this? Something like: > > ``` > @showcost begin >...

Or maybe something more subtle, like: ```julia curshowcast = showcast() setshowcost!(true) D = A * B * C setshowcost!(curshowcast) ``` since you want to keep the current state is was...

Perhaps we can just start linking people to the documentation page anyway, and just have a short description in the Github README section (we probably don't need these long examples...

I think it is better to do this as part of the rewrite.

For dispatch it is helpful to have special ranges that start at one since that is used as axes by most arrays, like how Base has `UnitRange` and `Base.OneTo`. Handling...

To be more specific, if we write most code in terms of `AbstractGradedUnitRange` and `AbstractUnitRange`, that should allow us to have a lot of subtypes (such as `GradedOneTo`, `GradedUnitRage`, `GradedUnitRangeDual`,...

> I realize that with the current design, there are already `GradedUnitRangeDual{ I will remove `UnitRangeDual` and use self-dual axes as the default. Sounds good, that's what I was going...