...
[01:49:02 CDT(-0500)] <Bosmon> And, despite a JSfiddle to the contrary, having a try/catch block in this very innermost part of the framework would be completely unacceptable performance-wise
[01:49:21 CDT(-0500)] <Bosmon> Although I guess we could try getting the JSfiddle working properly, I didn't spend too much time with it
[01:50:06 CDT(-0500)] <colinclark> remind me what the nature of the performance impact with try/catch blocks is?
[01:50:17 CDT(-0500)] <Bosmon> It causes the VM to "deoptimise" the function containing them