When I heard about that new Lumen framework, I said "Great, they decided to do something about the poor performance of Laravel".
Lumen advertize itself (frontpage) as "The stunningly fast micro-framework by Laravel." or "a framework dedicated to micro-services".
Yeah, you got it, the proper buzzword.
Then I saw some reddit post arguing that indeed it is faster than all others (ie Slim).
Then I went over this article from the author himself
http://taylorotwell.com/how-lumen-is-benchmarked/
(I will not get into the discussion on how wrong is this environment setup for benchmarks).
So since I know from my personal benchmarks that there is a huge gap between Laravel & Slim2 (Slim being 9x faster), since I noticed that Slim3 is slower than Slim2, I decided to setup a fresh machine to see by myself.
Quickly, my conclusion: it is pure sham, Slim3 is 2 times faster than Lumen.
Notes :
- command used: ab -k -t 10 -c 10 url
- "composer dump-autoload -o" executed as stated in the link above
Why do I think it is intellectual sham?
Remove the buzzword and you end up with yet another framework that by its existence aknowledges the poor raw performance of laravel.
However, when you look deeper it is far better than Laravel because a simple hello controller will only load 52 files (while laravel is 127).
So it equals Silex - And still it is bloatware.
A true microframework loads below 20 files,the lowest I seen is 6 files.
Why is it important? we all know the loaded files number has prime impact on performance.
Conclusion v2
Why creating a new brand? Why not just calling this the new Laravel version N+1, a complete rewrite, that tackles its pathetic performance?
Lumen advertize itself (frontpage) as "The stunningly fast micro-framework by Laravel." or "a framework dedicated to micro-services".
Yeah, you got it, the proper buzzword.
Then I saw some reddit post arguing that indeed it is faster than all others (ie Slim).
Then I went over this article from the author himself
http://taylorotwell.com/how-lumen-is-benchmarked/
(I will not get into the discussion on how wrong is this environment setup for benchmarks).
So since I know from my personal benchmarks that there is a huge gap between Laravel & Slim2 (Slim being 9x faster), since I noticed that Slim3 is slower than Slim2, I decided to setup a fresh machine to see by myself.
Quickly, my conclusion: it is pure sham, Slim3 is 2 times faster than Lumen.
- Lumen --> Requests per second: 212.89 [#/sec] (mean)
- Slim3 --> Requests per second: 526.94 [#/sec] (mean)
Notes :
- command used: ab -k -t 10 -c 10 url
- "composer dump-autoload -o" executed as stated in the link above
Why do I think it is intellectual sham?
Remove the buzzword and you end up with yet another framework that by its existence aknowledges the poor raw performance of laravel.
However, when you look deeper it is far better than Laravel because a simple hello controller will only load 52 files (while laravel is 127).
So it equals Silex - And still it is bloatware.
A true microframework loads below 20 files,the lowest I seen is 6 files.
Why is it important? we all know the loaded files number has prime impact on performance.
This is not a microframework (158k SLOC!)
Why creating a new brand? Why not just calling this the new Laravel version N+1, a complete rewrite, that tackles its pathetic performance?
Conclusion v3
They advertize about performance: it is pure sham.
To me It is just like creating a problem (overengineering a simple thing) and then being proud of yourself because you finally been able to solve it ;-)
They advertize about performance: it is pure sham.
To me It is just like creating a problem (overengineering a simple thing) and then being proud of yourself because you finally been able to solve it ;-)
