On Thursday, 24 January 2013 at 22:28:04 UTC, Jonathan M Davis wrote: > However, I do think that would be able to mark variables such > that they're treated as rvalues such that swapping a > variable out with a property function dosen't break code agreed