This has nothing at all to do with the "diffusion equation" If the temperature at one side (take it to be x= 0) is T1 and the temperature at the other side (take it to be x= d, the thickness of the glass) is T2 then the stationary temperature field inside the window is linear: T(x)= T1+ x(T2 T1)/d.
When x=0, that is T(0)= T1+ 0(T2 T1)d= T1. When x= d that is T(d)= T1+ d(T2 T1)/d= T1+ T2 T1= T2.
(Well, it does have a little to do with the "diffusion equation". The diffusion equation is $\displaystyle \frac{\partial^2 T}{\partial x^2}= \kappa\frac{\partial t}$. If the temperature field is "stationary" then its derivative with respect to t is 0 so the second derivative with respect to x must be 0 from which we conclude that the temperature is a linear function of x.)
