copyto!(Eye(4), I)

[cossio]

This seems like a weird error in this context. Since Eye(4) is already an I, this could just return without erroring?

julia> copyto!(Eye(4), I)
ERROR: ArgumentError: Cannot fill! with 0.0 an AbstractFill with value 1.0.
Stacktrace:
 [1] fill!
   @ ~/.julia/packages/FillArrays/xqile/src/FillArrays.jl:54 [inlined]
 [2] fillstored!
   @ ~/.julia/juliaup/julia-1.8.5+0.x64.linux.gnu/share/julia/stdlib/v1.8/LinearAlgebra/src/special.jl:356 [inlined]
 [3] fill!(A::Diagonal{Float64, Ones{Float64, 1, Tuple{Base.OneTo{Int64}}}}, x::Int64)
   @ LinearAlgebra ~/.julia/juliaup/julia-1.8.5+0.x64.linux.gnu/share/julia/stdlib/v1.8/LinearAlgebra/src/special.jl:367
 [4] copyto!(A::Diagonal{Float64, Ones{Float64, 1, Tuple{Base.OneTo{Int64}}}}, J::UniformScaling{Bool})
   @ LinearAlgebra ~/.julia/juliaup/julia-1.8.5+0.x64.linux.gnu/share/julia/stdlib/v1.8/LinearAlgebra/src/uniformscaling.jl:376
 [5] top-level scope
   @ REPL[42]:1

[dlfivefifty]

trying to write to immutable arrays isn't really supported anywhere:

julia> copyto!(1:5, 1:5)
ERROR: CanonicalIndexError: setindex! not defined for UnitRange{Int64}
Stacktrace:
 [1] error_if_canonical_setindex(#unused#::IndexLinear, A::UnitRange{Int64}, #unused#::Int64)
   @ Base ./abstractarray.jl:1352
 [2] setindex!
   @ ./abstractarray.jl:1343 [inlined]
 [3] copyto_unaliased!
   @ ./abstractarray.jl:1038 [inlined]
 [4] copyto!(dest::UnitRange{Int64}, src::UnitRange{Int64})
   @ Base ./abstractarray.jl:1018
 [5] top-level scope
   @ REPL[1]:1

[jishnub]

That's not really the case for AbstractFillArrays, though

julia> Fill(2, 2) .= 2
2-element Fill{Int64}, with entries equal to 2

Nonetheless, this should be fixed in LinearAlgebra rather than in this package

[dlfivefifty]

Ha interesting! I suspect I added that behaviour for Infinite arrays.....but can't remember the exact details.