I posted this as a separate submission but Nico posted the followup post [1] mentioned at the end of the opening paragraph: > In particular, the existing borro

I posted this as a separate submission but Nico posted the followup post [1] mentioned at the end of the opening paragraph: > In particular, the existing borrow checker rules, which were aimed at preventing dangling pointers, turn out to be extremely well-suited to this task. I find this very interesting and very heartening as well, and I think it points to a kind of deeper analogy between memory errors in sequential programs and data races in parallel programs. I will elaborate on this theory in a later post. [1] http://smallcultfollowing.com/babysteps/blog/2013/06/11/on-the-connection-between-memory-management-and-data-race-freedom/ http://smallcultfollowing.com/babysteps/blog/2013/06/11/on-t...