John M. Kuhn

What version of Julia and Formatting are you using? With Julia 1.0.0 and Formatting 0.3.4 I get: ``` julia> printfmt("{1:.4e}", 40000.0) 4.0000e+04 ```

What do you get for the following? ``` julia> printfmt("{1:.4e}", 39999.9) 4.0000e+04 julia> printfmt("{1:.4e}", 40000.1) 4.0000e+04 ``` I don't have access to a Windows 7 machine but it seems odd...

Your 2 examples expose 2 different issues. The first is Julia issue [#37171](https://github.com/JuliaLang/julia/issues/37171). ``` julia> round(9.99999999e-308, sigdigits = 4) 9.99999999e-308 ``` when the desired result is `1.0e-307`. The second example...

Fixed in Julia Master (1.8) https://github.com/JuliaLang/julia/pull/42033

I would use [Calendar Versioning](https://calver.org/) using the date of the commit as the version number (2022.03.25 or in the case of MPIR 2020.12.25. In your Release Notes you should indicate...

Some discussions about standardizing this in C: https://www.austingroupbugs.net/view.php?id=1166 https://www.austingroupbugs.net/view.php?id=258

The results vary run to run. Most likely a buffer overflow in the C. I'll look into it. ``` julia> Dec128("1" * "0"^106, RoundUp) 1.000000000000000000000000000000001e106 julia> Dec128("1" * "0"^105, RoundUp)...

DecFP is a wrapper for the [Intel Decimal Floating-Point Math Library](https://software.intel.com/en-us/articles/intel-decimal-floating-point-math-library). It only supports the IA-32, Intel 64 and IA-64 architectures. It doesn't support ARM. There has been some interest...

It compiles, but I think there are run time errors. I will look at this in the next few days.

I'm out of town at the moment but later this week I can test 32-bit ARM locally.