Willow Ahrens

Valentin, unfortunately I think that code is lost to time. However, it sounds like the code y'all have is already more sophisticated than what I was doing.

Here is an example to reproduce the problem. File `spmv2.cpp`: ``` #include "taco/tensor.h" #include #include #include #include #include template double benchmark(double time_max, int trial_max, Setup setup, Test test){ auto time_total...

Sure thing! I think this would be great to include with the next version, and I wanted to file the PR before I forgot the details. I chose `splitargdef` because...

I will clarify that there is no current work-around for literal `nothing` input to `splitarg`, and no workaround to differentiate an `Any` typeassert from no typeassert based on the output...

Congratulations! You have discovered one of the more complicated issues with multiple dispatch: method ambiguity. Here's what's happening: ForwardDiff has defined `+(x::Real, y::DualNumber).` and ArbFloats has defined`+(x::ArbFloat, y::Real).` Since both...

Currently, Memoize uses only one cache per function. It stores this cache in a global variable in the module that expands the macro. If the global variable is already defined,...

@rikhuijzer It's true that if you define the NonexistentCacheType memo function first, you get the error. However, if the function already has a cache (if you use the LRU cache...

For correctness, I don't think we need to empty the cache if we simply formalize the expectation that the cache passed to memoize only contains correct values (this would be...

Reflecting a bit, I realize now that my previous comment was written under the assumption that only the memoized function gets redefined, and that it doesn't reference any changing global...

> Come to think of it, is there an argument against emptying the (now unreachable) Dict on method redefinition? It makes the code a bit uglier, but that's it, no?...