If I calculate the differential length $ds$(see picture) as $rd\varphi$ I get a different value than if I take the projection of $r$ and the projected angle $d\theta$ to get $Rd\theta$ (where $R$ is the projection of $r$).
Does $d\varphi$ not approach $d\theta$ in the limit or does $ds$ not approach $ds'$? What is wrong with my approach? Please help( I'm sorry for my notations).