Hacker News Reader

Comment by Fargren | Hacker News Reader

If you send a small amount of bits over the network, the amount of responses you can get is commensurately small. So if you aren't sending a lot of bits, it would be possible to simply have a local mapping of inputs to outputs. There is a certain size the message has to have for it to be worth it to send it instead of solving the problem locally; obviously, I don't know what that size is or whether SimCity's packages where smaller than that size. But the claim of the OP isn't totally absurd.

View on HN

> If you send a small amount of bits over the network, the amount of responses you can get is commensurately small.

Oh, come on! You can entirely specify a cosmically hard problem in just a few kB. Prime factorization, anyone? Use discrete logarithms in finite fields, and you get down to handfuls of bytes.

Your conclusion is probably right, but your theoretical basis for it leaves a lot to be desired.

If a response is too small to describe the changes to the state of your city at that time, clearly it's a thick client which mostly knows how to simulate the city.

That sounds much better -- less like some pseudo information theory or complexity nonsense.

That said, I would put some small calculations on a server if I thought cheating was an issue. This wouldn't necessarily apply to single player games, though.

You don't have to transmit many bits to perform an internet search.

When you perform an internet search are you all by yourself consuming more computing resources than the i5 processor in your computer? That's unlikely, but you'd also have a difficult time replicating the functionality of Google with your CPU alone and only the storage on your own laptop.

Google is matching your query against a humongous database. In Maxis' case, the only relevant data is some aggregate data from the other cities in your region.

True, Maxis probably isn't doing much computation. rz2k is simply pointing out that Fargren's argument is incorrect. Same as what stcredzero said, "Your conclusion is probably right, but your theoretical basis for it leaves a lot to be desired."

Reply on news.ycombinator.com