craig at netjam.org
Thu Nov 15 08:29:13 UTC 2007
> Oh, interesting. Did you ever run the macro benchmarks to see how much
> of a difference that makes? I've always wondered about the speed
> impact in a realistic kind of setting.
Ah, in normal use there's just an additional check to see if the
class of the receiver is the proxy class (which is in the special
objects array). The effect on the macro benchmarks was undetectable.
When proxies are involved, the network overhead of sending remote
messages swamps the slowdown caused by invoking the normal sends.
So, not the test you were hoping for. :)
improvisational musical informaticist
Smalltalkers do: [:it | All with: Class, (And love: it)]
More information about the Squeak-dev