Is that actually true that PHP is slower? Say something like Symphony2 vs rails for a reasonably complicated app. Is there much out there in terms of benchmarks?
(I think we are both coming from that it really doesn't matter most of the time side but would be interesting to know if PHP is actually slower).
I don't know about Rails, but as a Symfony2 user, it's actually a bit on the slower side compared to other PHP frameworks.
But ~300ms server response times is nothing if your end user has to wait 5-6 seconds for the page to render due to suboptimal frontend code. Plus, throw in varnish/nginx microcaching and requests drop to ~10ms regardless of your framework.
> But ~300ms server response times is nothing if your end user has to wait 5-6 seconds for the page to render due to suboptimal frontend code.
True for a single request, but once you start getting lots of concurrent requests all those ~300ms server response times start adding up quite quickly.
When Twitter started out, their use of ROR wasn't a problem. Once the service started to become popular, however, it was.
Yes, the VM itself is not fast. However, a lot of it is just a thin shim over a C call, so whether you're going to be bottlenecked by it is going to depend on what you're doing.
(I think we are both coming from that it really doesn't matter most of the time side but would be interesting to know if PHP is actually slower).