Vinzent Steinberg

@mlindner That's approach (b) from the issue description. One possible issue might be that `libm` can yield different results than `libc`, and that you can already use it in your...

What are the differences?

I wonder whether it is intentional.

I don't think it is the latter, because the default (`FLT_EVAL_METHOD = 0`) is the behavior you would expect. Increasing `FLT_EVAL_METHOD` should result in more correctly rounded results. From Wikipedia:...

Indeed, supporting derive would be more user-friendly than the current macros and makes sense. Implementing `Merge` for containers is an interesting idea, I need to think a bit about the...

An advantage of moving `lazy_static` to `std` (or rather `core`) would be that `no_std` could be supported on stable, because the required unstable features can be used for the implementation.

bc.. I don't know whether it's useful. It's nice to have. (And when you have the choose between returning +0 or -0, it's trivial to use it.) p. Original comment:...

bc.. I really have to say that I'm very used to this ordering: cn_x^n + ... c1_x + c0 I'm even so used to it, that it would be more...

bc.. Sorry, I meant exactly this ordering. I did not name the coefficients correctly. p. Original comment: "http://code.google.com/p/mpmath/issues/detail?id=102#c3":http://code.google.com/p/mpmath/issues/detail?id=102#c3 p. Original author: "https://code.google.com/u/[email protected]/":https://code.google.com/u/[email protected]/

bc.. Why? For sparse polynomials? p. Original comment: "http://code.google.com/p/mpmath/issues/detail?id=102#c5":http://code.google.com/p/mpmath/issues/detail?id=102#c5 p. Original author: "https://code.google.com/u/[email protected]/":https://code.google.com/u/[email protected]/