You might possibly be able to measure a small performance cost to  
using inject:into: because it passes down to a virtually identical  
loop (ie uses a temp, a do: loop etc) and costs you one extra message  
send or so.  If someone chose to write the compiler optimiser code to  
inline inject:into: then it would not cost even that.

Given that modern machines are typically able to send 3-10 million  
messages a second, it probably isn't worth anyone's effort.

tim
--
tim Rowledge; tim at rowledge.org; http://www.rowledge.org/tim
Oxymorons: Airline Food