Question: Newton Iteration Not Converging

I'm trying to plot the solution of the following differential equation:

theta * y''(t) - y'(t) = 0  with boundary conditions  y(-1)=alpha  and  y(1)=beta. Theta is supposed to be very small, say 0.01, and for starters I put  alpha=1  and  beta=-1. But Maple does not compute. I use the following source code:

> singularperturbation:={0.01*diff(y(t),[t$2]) - y(t)*diff(y(t),t) = 0, y(-1)=1,y(1)=-1};

> differentialsolve:=(singularperturbation,y(t),type=numeric,method=bvp[midrich],abserr=1e-2);

>plots[odeplot](differentialsolve,[t,y(t)],-1..1,color=green,style=line,symbol=circle,symbolsize=10);

My difficulty is that once I press "enter" after writing the second line of the code,I get the message:

"Error,(in dsolve/numeric/bvp) newton iteration has not sufficiently converged after max iteration count." 

It suggested to me that instead of writing abserr=1e-2 at the end of the second line, I should write abserr=1e-1. But still, the error message this time reads:

"Error,(in dsolve/numeric/bvp) Newton iteration is not converging."

Can someone please tell me how I can overcome this difficulty? If the second line of the code keeps giving me error messages,obviously I cannot plot the solution. And what's the deal with this "abserr"?  Why do I need to include absolute error at all?  Thanks kindly.

AntipodeanMan