`eps` not defined for `LogFloat64` or `ULogFloat64`

[stecrotti]

Any chance we can define eps(::LogFloat64) and eps(::ULogFloat64)? I don't have enough knowledge of the workings on eps to propose a sensible fix but i can try to follow suggestions.

Use case: I have a complex number whose real and imaginary parts are Logarithmics. When trying to take the absolute value i get an error because eps is not defined for ULogFloat64 and for some reason abs calls eps.

MWE

julia> using LogarithmicNumbers

julia> x = Logarithmic(1.0)
+exp(0.0)

julia> z = x + im*x
+exp(0.0) + +exp(0.0)*im

julia> eps(z)
ERROR: MethodError: no method matching eps(::Complex{LogFloat64})
The function `eps` exists, but no method is defined for this combination of argument types.

Closest candidates are:
  eps(::Type{Dates.DateTime})
   @ Dates C:\Users\sky_9\.julia\juliaup\julia-1.11.5+0.x64.w64.mingw32\share\julia\stdlib\v1.11\Dates\src\types.jl:452
  eps(::Type{Dates.Time})
   @ Dates C:\Users\sky_9\.julia\juliaup\julia-1.11.5+0.x64.w64.mingw32\share\julia\stdlib\v1.11\Dates\src\types.jl:454
  eps(::Type{BigFloat})
   @ Base mpfr.jl:1075
  ...

Stacktrace:
 [1] top-level scope
   @ REPL[13]:1

My use case

julia> using LogarithmicNumbers

julia> x = Logarithmic(1.0)
+exp(0.0)

julia> z = x + im*x
+exp(0.0) + +exp(0.0)*im

julia> abs(z)
ERROR: MethodError: no method matching eps(::Type{ULogFloat64})
The function `eps` exists, but no method is defined for this combination of argument types.

Closest candidates are:
  eps(::Type{Dates.DateTime})
   @ Dates C:\Users\sky_9\.julia\juliaup\julia-1.11.5+0.x64.w64.mingw32\share\julia\stdlib\v1.11\Dates\src\types.jl:452
  eps(::Type{Dates.Time})
   @ Dates C:\Users\sky_9\.julia\juliaup\julia-1.11.5+0.x64.w64.mingw32\share\julia\stdlib\v1.11\Dates\src\types.jl:454
  eps(::Type{BigFloat})
   @ Base mpfr.jl:1075
  ...

Stacktrace:
 [1] _hypot
   @ .\math.jl:729 [inlined]
 [2] hypot
   @ .\math.jl:706 [inlined]
 [3] abs(z::Complex{LogFloat64})
   @ Base .\complex.jl:286
 [4] top-level scope
   @ REPL[8]:1

EDIT: the same issue also comes up while trying to use logarithmic numbers in combination with finite differences.