Out of academic curiosity, does anyone know if scalar construction is likely to be more or less expensive than
scalar modification? As a matter of fact, it would be really
interesting to see performance analyses of many common
perl operations.. Maybe they're already out there.. Any
pointers, anyone?