Ok, fine, I'll do the actual curve fitting instead of just estimating.

Eyeballing it, were saying 1% in 2013, 2% in 2021, 3% in 2023?

Gives us a fit of..

0.873 * exp(0.118 * x)

So..

Correct the equation and solve for x

x_target = np.log(200 / a) / b

Calculate the actual year

year_target = 2013 + x_target

print(year_target)

In ~2058 everyone will be using two linux desktops at once.