Unhelpful return code for IntervalNonlinearProblem

[TomRottier]

Trying to solve the following IntervalNonlinearProblem:

f(t,p) = sin(t) - 10.0
prob = IntervalNonlinearProblem(f, [0.0, 2pi])
sol = solve(prob, ITP())

gives a FloatingPointLimit return code and the solution as 0.0. Which according to the docs (https://docs.sciml.ai/SciMLBase/stable/interfaces/Solutions/#retcodes) means a successful solve:

ReturnCode.FloatingPointLimit
The success state of the solver. If this return code is given, then the solving process was successful, and the closest floating point value to the solution was given.

Given this equation cannot be solved this return code doesn't seem appropiate. It also feels like the "closest floating point value to the solution" should be pi/2 rather than 0.0.

On Julia 1.10.4 and:

(jl_ajmNJc) pkg> st
Status `/tmp/jl_ajmNJc/Project.toml`
  [8913a72c] NonlinearSolve v3.13.1

[oscardssmith]

it should probably be an initialfailure since I think intervalproblem requires opposite signs on the boundary

[ChrisRackauckas]

Yes, it should give an initial failure with a warning about not being an enclosing interval.