Issue with RobinBC

[gentine]

Hi Chris, Differential Equations is a neat package. Thanks for providing this to the community.
I have been having some issue with the RobinBC and General boundary condition. I have u(0)=0 and u'(L)=0 in my problem. RobinBC works well on the rhs but bot on the left hand side. Also when using the General BC I get another behavior even with the same BCs.
Any idea what the problem could be?
My BC is
const bc1 = RobinBC((1., 0., 0.), (0., 1., 0.), dx, 2)
or
const bc1bis = GeneralBC([0.0, 1.0, 0.0], [0.0, 0.0, 1.0], dx, 2)

And the equation I am solving for is:
step!(u,p,t) = Ksat/νe*((cos(α)*Δ+sin(α)∂x)bc1bisu)
tmax = 365.086400.0
prob = ODEProblem(step!, u0, (0.0, tmax))
alg = KenCarp4()
sol = solve(prob, alg)

with initial conditions:
t0 = 0.0
#u_analytic(x, t) = sin(2πx) * exp(-t*(2π)^2)
hmax0 = 100.0
u_analytic(x, t) = hmax0sin(2πx/(4L)) * exp(-t(2*π)^2)
u0 = u_analytic.(knots, t0);

and parameters
CFL = 0.5 # =u.dt/dxCourant-Friedrichs-Lewy CFL stability
α = π/10
Ksat = 2.0e-5; # saturated hydraulic conductivity
νe = 0.4 #porosity
u = Ksat/νe
dt = CFL*dx/u # 0.03 #3600.0

Thanks a lot

[ChrisRackauckas]

Hey,
Sorry it took awhile to take a look at this. It seems that knots is not defined so I am unable to run this. Could you provide a full MWE? I assume the issue is that the solution is defined only on the interior of the domain so the issue is in the definition of knots, but since it's not defined here I cannot be certain.