In that case project selection is important. Unpaid work towards a project who's main effect will be to reduce licensing costs for fortune 500 companies is very different to working on something to (say) improve educational outcomes for low income children.