Remember "Optimization is the root of all evil" from Knuth ? Why optimize something if it isn't needed ? if you show me something that clearly won't solve a problem due to it's performance, it's time to optimize otherwise it's just wasting time. "Uhh but this could be faster" yeah, and gnu ls could have less than 3131 switches and beer could be dry.
Cool I found a *huge* openbsd bottleneck, solve the bottleneck forget all the rest. Honestly if someone is that concerned about performance go to linux, they will sacrifice anything in their performance crusade. We could also, all shut up and code.
