Again, a totally useless benchmark. I really feel all these benchmark tests are useless, if they are not in the realm of any real world app. So why not create a crud app mimicing benchmark, test writes and reads, and calculations instead. In the real world it comes down to the ease of development, deployment and maintenance. This is where the traditional web languages still win, and ill want to bet you can ship software faster than ever these days when using a proved web framework.
Speed is not the only factor, and usually not even a factor that has to be accounted for, lets be reasonable, 99% of any web apps dont reach the state that they even need to scale to 10000s of user actions per second.
Id be interested to see how hhvm/php whould compare though.
Useless? Not at all, but it's just data. It has to be interpreted, and how you do that depends on your needs. If nothing else, it's interesting from a technical standpoint to see the huge difference it performance.
To think that a benchmark like this tells the whole story of the "best" framework is wrong. Of course it is. It's also wrong to think that it tells the whole performance story. Of course it doesn't. But that does not make it useless.
Id be interested to see how hhvm/php whould compare though
Aren't you assuming that the benchmark is targeted at all web-apps? They are just creating a benchmark. If an app you design doesn't need to scale, then you can ignore the benchmark.
To the rest who are building apps that need to scale, the benchmarks are definitely useful.
Speed is not the only factor, and usually not even a factor that has to be accounted for, lets be reasonable, 99% of any web apps dont reach the state that they even need to scale to 10000s of user actions per second.
Id be interested to see how hhvm/php whould compare though.