Comment Re:Why bother? (Score 1) 421
I agree with the majority of your post here, just wanted to comment on this:
"(why is Apache still spawning processes for every request that comes in... don't they realize the overhead of that??)."
It's because they don't control the entire software stack, as MS does, so they can't guarantee that a multithreaded model will work properly -- all it takes is one library in the 3rd party module chain to have been buggy (or simply not work) under multithreading, and then the server is unstable.. Apache has supported a multithreaded workers forever.. well back to the 2.0 days (at least a decade), but it's never been default. It also supposes a hybrid mode, where there are multiple workers and each worker spawns multiple threads, in an attempt to alleviate some of the issues caused by the library chains.