I tend to agree with you as well. I've also come to the conclusion that conceiving of software as some new 'kind' of non-linear writing is more productive than

I tend to agree with you as well. I've also come to the conclusion that conceiving of software as some new 'kind' of non-linear writing is more productive than understanding it through the traditional engineering discipline. But it is both and it is neither. Software is a new thing that we have yet to fully understand.