Open yehosef opened 8 years ago
Gin is newer and was only added to the list of benchmarks in a subsequent round, after the initial writeup. Hence, it was not mentioned in the conclusions.
In fact, much of the writeup is no longer accurate -- in particular, Phoenix has evolved significantly, and is now actually probably closer to Rails than Sinatra (and the other minimal Sinatra-likes) -- while still maintaining excellent performance characteristics I've been told.
I don't have an intention to continue to update these benchmarks in perpetuity as I don't feel they are necessarily that useful anymore -- so my current thinking is I will probably make a TODO to mark the README as reflecting a certain period of time, no longer open to updates, and provide links to more up to date resources...
It seems that Gin had the highest thoughput (except maybe play), lowest latency and lowest deviation (highest consistency.) It seems that this would all be very positive indicators (within the confines of the test) but you don't mention it in the conclusions.
Is there some reason it's not relevant? If so, perhaps it should be removed from the results.