Frames White

Work around, is don't use `Complex` use `ComplexF64`, which will also make your code all round faster. Since `Complex` is an nonconcrete type. See https://docs.julialang.org/en/v1/manual/performance-tips/#man-performance-abstract-container If you do `ones(ComplexF64, 100,...

no this plan won't work right as as. You can't always mute in the references after the fact. Because a immutable type might have fields that are references. I think...

ok, so cycles are only possible with mutation. So the kinda extreme version of what we can do is 1. do cycle detection + reference counting 2. break the graph...

woah, very cool. Is this documented anywhere? I see the other end was added to the HolyLabRegistry in https://github.com/HolyLab/HolyLabRegistry/pull/161

I want to test exact equality, though, not approximate. And I want equality on tuples to work. (`isapprox` isn't defined for tuples) The hash complication is trivially resolved. I hope...

Not having this is causing issues for Diffractor I look into this further soon and see what (if any) invalidation issues there are.

I also have encountered this. I suspect it is relatively common when the tests are like lots of small unit tests at the start, and then some big slow integration...

Per https://github.com/JuliaCollections/DataStructures.jl/issues/479 we are not accepting new data structures til post-1.0.0 I guess a case could be made that this PR was open long before then and so should be...

why does this have some very aggressive compiler directives thoughout it? these are going to require careful hand checking. The compiler can normally work things out.

sorry, its just i am not sure if this belongs in this package or not. It's a very intricate function.