Er, that doesn't seem like a good reason to downgrade the UI of something important. I don't know how much money it takes to keep HN running, but would just buying ten times as much computing power really be a significant expense for YCombinator, compared to the dealflow from HN?
I suspect the bottleneck isn't cost so much as the mental bandwidth to switch hosts, etc. It would be a context switch for PG/RTM/bringing someone else up to speed on arc. HN's just naturally lagging other priorities.
---
Farther in the future: HN's arc webserver uses simple flat files for storage, so moving from one host to two is probably a big context switch as well. I wonder what the current server's specs are, and how long we have until that kinda-sorta-Y2K-like barrier.
Comments
Er, that doesn't seem like a good reason to downgrade the UI of something important. I don't know how much money it takes to keep HN running, but would just buying ten times as much computing power really be a significant expense for YCombinator, compared to the dealflow from HN?
As far as I know, the current server has close to the fastest processor available.
RAM might matter more for news.
or rewrite it in a more efficient language/runtime would likely have more effect.
I suspect the bottleneck isn't cost so much as the mental bandwidth to switch hosts, etc. It would be a context switch for PG/RTM/bringing someone else up to speed on arc. HN's just naturally lagging other priorities.
---
Farther in the future: HN's arc webserver uses simple flat files for storage, so moving from one host to two is probably a big context switch as well. I wonder what the current server's specs are, and how long we have until that kinda-sorta-Y2K-like barrier.