Relax signature of ==(::MPolyIdeal, ::MPolyIdeal)

[joschmitt]

Small change, but requires a long explanation:
The current signature of == for ideals in multivariate polynomial rings is

==(::MPolyIdeal{T}, ::MPolyIdeal{T}) where T

so this function is not called if one tries to compare ideals in polynomial rings of different type (T), instead the generic == is called:

julia> R, (x, y) = QQ["x", "y"];

julia> S, (u, v) = grade(R, [ 1, 1 ]);

julia> ideal(R, [x]) == ideal(S, [u])
false

julia> @which ideal(R, [x]) == ideal(S, [u])
==(x, y)
     @ Base Base.jl:127

The answer false is of course not wrong, but in my opinion this should error because the comparison doesn't make sense; this is what I added in this PR:

julia> ideal(R, [x]) == ideal(S, [u])
ERROR: Base rings must coincide.
Stacktrace:
 [1] error(s::String)
   @ Base ./error.jl:35
 [2] check_base_rings(I::MPolyIdeal{QQMPolyRingElem}, J::MPolyIdeal{MPolyDecRingElem{QQFieldElem, QQMPolyRingElem}})
   @ Oscar ~/.julia/dev/Oscar/src/Rings/mpoly-ideals.jl:75
 [3] ==(I::MPolyIdeal{QQMPolyRingElem}, J::MPolyIdeal{MPolyDecRingElem{QQFieldElem, QQMPolyRingElem}})
   @ Oscar ~/.julia/dev/Oscar/src/Rings/mpoly-ideals.jl:993
 [4] top-level scope
   @ REPL[6]:1

julia> @which ideal(R, [x]) == ideal(S, [u])
==(I::MPolyIdeal, J::MPolyIdeal)
     @ Oscar ~/.julia/dev/Oscar/src/Rings/mpoly-ideals.jl:992

[joschmitt]

I also added a check_base_rings to is_subset(::MPolyIdeal, ::MPolyIdeal); this function would work for rings which "look the same" but aren't (which is quite convenient but also leads to more problems than it solves in my opinion).

[fingolfin]

Thank you, good idea!