Very big BigFloat behaves differently depending on OS

[Kolaru]

On Windows:

julia> big"-5.87565e+1388255822130839282"
-Inf

julia> versioninfo()
Julia Version 1.8.5
Commit 17cfb8e65e (2023-01-08 06:45 UTC)
Platform Info:
  OS: Windows (x86_64-w64-mingw32)
  CPU: 4 × Intel(R) Core(TM) i5-6600K CPU @ 3.50GHz
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-13.0.1 (ORCJIT, skylake)
  Threads: 1 on 4 virtual cores

On Ubuntu (WSL):

julia> big"-5.87565e+1388255822130839282"
-5.875649999999999999999999999999999999999999999999999999999999999999999999999983e+1388255822130839282

julia> versioninfo()
Julia Version 1.8.5
Commit 17cfb8e65ea (2023-01-08 06:45 UTC)
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 4 × Intel(R) Core(TM) i5-6600K CPU @ 3.50GHz
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-13.0.1 (ORCJIT, skylake)
  Threads: 1 on 4 virtual cores

It was first noticed on CI on JuliaIntervals/IntervalArithmetic.jl#555.

[vtjnash]

In C on the Win64 platform, the long datatype on Windows cannot represent as large of a value as it can on LP64 systems

[giordano]

I can't find other tickets about this, but I was pretty sure this was reported by other users before because it's easy to trip on this. But as Jameson said, this is a limitation of MPFR on Windows, there's little we can do about it.

[giordano]

Oh, here it is: #40661. Closing as duplicate