Closed Kabingeran closed 4 years ago
Hi @Kabingeran,
Could you please develop your opinion ?
Regards,
You are comparing frameworks with different functionality. Naturally, frameworks with more functionality will be slower. This means that you cannot compare everything with everything. You can only compare frameworks with similar functionality. Maybe you need to make categories of frameworks. For example, Yii, Symphony, Laravel, etc. But this is just my guess ;)
Yes, I Accept. But what you have mentioned above is only for PHP (Incase Laravel is a complex framework build upon Symphony). As per your idea we can't have Symphony and Laravel in the same category.
Between languages you are out of option to compare. Maybe methodology and functionalities same but Implementation is different. They are all different and unique in their own way.
When comes the term bench-marking, you should let it run all them together as one. Even complex frameworks should perform well in the basic tests.
Finally we are not blaming the least performing or praising the besting performing. Just a idea of comparison between A to Z. If need you may categorize the results which is same.
I suggest average the results by different test cases (math, string, ...etc) Ops with threading and concurrency.
Your Idea?
@Kabingeran
You are comparing frameworks with different functionality.
We are NOT comparing, but doing some benchmark => measuring some indicators (actually performances ones)
Maybe you need to make categories of frameworks.
Sure, but I'm focused on other features, but it will come. Feel free to help :heart: Possibly, by categorizing each frameworks here
@sakthisaran65 not opposed ... I'm more focused on making a decentralized tool, I mean having various parts, mostly a (GraphQL) api to store/query the results and a frontend to display various graphical representations
Like @Kabingeran i think the frameworks should be divided up in to three or four categories based on rough feature set and/or framework focus: http servers, http servers with routing, (or combine the previous two categories), micro-frameworks, and full featured frameworks.
The results roughly break down along these lines anyway, but you have to know what all of these are to know that, which is less than ideal.
We can see that the fastest are the http servers, then http servers with routing, then micro service/api focused micro-frameworks, then the whole hog/kitchen sink frameworks bring up the rear. :)
There are frameworks that perform better than their category, but they don't get the ranking they deserve when surrounded by lighter weight alternatives. There may be some that perform worse than their category. It would be nice to see how they all perform in their appropriate "weight classes" by default.
Hi! Is it appropriate to compare the response speed of frameworks with different functionality? I think no.