4 ms·
If OP actually cared about the execution speed, he'd be able to have the .NET code running often almost two orders of magnitude faster than Ruby: http://shooto
by acqq 15y ago
If OP actually cared about the execution speed, he'd be able to have the .NET code running often almost two orders of magnitude faster than Ruby:
http://shootout.alioth.debian.org/u32/benchmark.php?test=all&lang=yarv&lang2=csharp http://shootout.alioth.debian.org/u32/benchmark.php?test=all...
If he says that Ruby fits him better I can just conclude that he trades the execution speed for ease of the development, which can be a reasonable trade-off. I'd still like to hear him talking in a year if he actually worked on really comparable problems in .NET and in Ruby.
- rbanffy 15y agoI'm not very familiar with Ruby, but with Python, I can profile my app, map the hotspots and easily rewrite them in C. Probably, I'll be done with the rewrite before the .NET version reaches RC quality. There is no reason to assume Ruby would be much different. In reality, I very rarely have performance problems. I have heard Ruby got much faster with the 1.9 release.
- acqq 15y agoIf you "rarely have performance problems in reality" that explains a lot. Speed differences between Ruby and NET (see the link in my previous post) translate to having one server instead of twenty. If you make low traffic solutions, that's OK. If you move bottlenecks to something outside of your responsibility (like for example, your app waits on sql server because no caching is done) OK, not your problem. If somebody doesn't have problems in buying more hardware, OK. Still it's approx. 20-times (see the link before, it is for Ruby 1.9) trade-off between development convenience and the hardware involved.
- rbanffy 15y agoIf you really need a rack full of Ruby servers to do the job of a single .NET server, your Ruby programmers are doing something (hint: most probably everything) wrong. Language choice excepted. Server performance is only very rarely tied to CPU performance. There are a whole lot of factors other than that that impact performance.