go-zoo/bone

More comprehensive benchmarks.

andrewchambers opened this issue · 6 comments

Your benchmarks are comparing the fetching of a single url - This is a misleading benchmark which is heavily biased towards bone because of its simple algorithm.

Please add some benchmarks where the number of routes is medium and large with varying degrees of shared prefixes.

Bone is not built to handle big application, it aim the small one for normal people. So please if you're not happy with it just use something else !

Also, I am currently working on a way to automatically modify the algorithm of bone depending the size of the application. Linear search is really efficient for small application, but if the application is bigger it could be a problem, so i'm thinking switching from linear to tree when the number of routes comes bigger.

Currently it looks like you are telling lies to make people use bone. I'm sorry for being harsh, but it is what it looks like at the moment.

Anyway, better benchmarks will only help you make it faster.

Yeah i know, but don't be harsh for that. Bone arrived on github 2 or 3 days ago, let me time to work on it. And as far as i know, github is a place to develop things, not to be mad at people for weird reasons ... If you have any useful comments and you express them politely, i will probably be more prone to check them seriously. So let me some time, and i hope the next update will make you feel better about it :)

Ok, Im sorry for being annoyed. Your marketing is very good, your benchmarks are not.

...