NOTES.txt

I was pissed because simple solvers were difficult to use. So I starte hacking solver_low_degree.
  * [x]  So far, I just un-uglified stuff.  
  * [ ]  I need to flip Iter and Real and extract Real =  by default  
  * [ ]  The latest NRiC puts bounds on Newton and Halley to keep them from taking off.  
  * [ ]  Instead of returning std::array<Solution<>> maybe have an OutIter argument - more algoish?
This fails:
  template<typename Iter, typename Real = emsr::value_t<std::decay_t<decltype(std::declval<Iter>()[0])>>>
    quadratic...
Maybe go off and inch forwards on a toy.
  * [ ] The low-degree solvers are real coeffs only!!!! That's why you can't have nice thinga!

I'm also trying to get the Skowron-Gould solver up and running.  
  * [ ]  S-G uses little endian polynomials! I should be able to use tools like Polynomial!.