Sadly, this video compare orange with apple. For example, the SQL insertion is made in one side using ORM ActiveRecord, and in another using raw INSERT query (unsanitized btw). Use ActionController::Metal, use PG driver and do the tests to compare orange with orange. Add YJIT, check for CoW for the workers (because it's not normal that it is using twice the ram for two workers btw) and then, yes, node will still be faster, but not by this crazy factor.
@@anykeyh I met so many people in the past couple years who recently migrated from Ruby and PHP to Go. Is Ruby really no more popular? Feels like it faded away momentarily after Heroku’s demise.
@@Nick-yd3rc Ruby faded away with SaaS gold rush. In 2010, building as SaaS was one of the best way to get rich quick. And RoR was praised to its quick development process, where all which matter is time to market. Ruby is a great language because of its syntax allowing DSL, and its awesome standard library. It's super useful for web, but also for scripting and devops. It is still well maintained and active, noticeably thanks to Shopify work later. The main competitor of Ruby is Python, and it gained a lot of traction because of AI and general academic use. I personally prefer Ruby for plenty of reasons. I think people should rationalize their need of performance in a web application. HTTP and Web has never been designed with performance in mind. While it's super great to have something going very fast, it's often a non-functional quality that you don't need. Most web servers are running application which are not exploited to their full potential, and if I spend 1 millions a year in cloud services, usually the part related to servicing web request is 10 to 20% max. Of course, if you plan to create a trading platform or a B2C platform and your revenue per request is extremely low, you might want to use faster language. I would never use Node as server-side language. Any serious application would ship with 1000+ NPM packages, which is insane and will cause you headache during maintenance and migration to newer versions. And javascript/node quirks and terrible std library (in my opinion) doesn't help to produce clean code, even using Typescript for example. Go is a great alternative, but the language is dull, by design (allowing easy maintenance). While I understand perfectly the approach of the language creator, I rather not spend 8 hours a day dealing with it. It's a question of personal choice here. Go is great, and if you like it, go for it.
@@Nick-yd3rc i'm not from the United States but i used to follow their job market there around 3-4 years ago, idk about now but in the past there was such a labor shortage of Rails devs that you can get quite a bit higher salary than Laravel dev for example. Where i live if you know .Net or Spring, you can get job pretty much any time (most of the full stack/backend job offers are for those 2) after that it is Node+Express or Php with Laravel or symphony and very rarely Go (albeit we do have slow increase in job offers).
@AntonPutra really appreciate your videos however for tests like this (and 209 - Django vs. Go) there's a bit of an apples vs. oranges comparison. Both RoR and Django start with more than a server like NodeJS & Go. E.g. middleware, Django has CSRF protection, XSS etc. all enabled by default. This functionality is not enabled in Go/NodeJS without additional libraries or function calls so it creates an imbalance in functionality and skews the test. Would be great if you could either a) remove middleware and other niceities built into frameworks (although that is not how they are used) or, b) alternatively expand the Node & Go code to match the functionality. This way if a dev was trying to evaluate a framework in different languages on performance alone it would be a fairer comparison. Thanks & happy hacking
@@ha5anmukhlis I would love to if I had time. Unfortunately I don't but I wanted to flag it incase someone else did have the time as these benchmarks are very helpful.
@@AntonPutra People only start to care when they see their favourite framework/runtime gets devestated in a public benchmark 😆 And only THEN they will bring their Sith-mastery optimization techniques with them.
@@AntonPutra maybe you're right, it's easier to watch the video and criticise than review first and wait for the video. But honestly I've been watching your tests for a while and haven't noticed dedicated segment with an announcement of next test (what exactly will be tested, here link to repo, ideally after intro). On the other hand I haven't paid attention to it, and if I'm wrong you can disregard my thoughts :)
Why do you compare large framework with millions line of code with pure runtime code without any dependencies? Rails can be compared only with Django, though it still has less functionality. Compare apples to apples, use pure Ruby API to compare with Node.js
For each video comparing two interpreted languages I’d love to see a go, zig or rust baseline to drive the point home about interpreted languages vs compiled.
That seems a bit pointless because we know those are faster. Also, the graphs would become unreadable. On the contrary I think he should only be comparing things expected to have similar performance, so that we might actually learn something new if they don't.
you compared the warm with the soft: - Rails is a framework with a full-fledged ORM, a large number of abstractions that simplify development - Node.js is just a JS runtime with a web server that does not even know how to work with cookies, there are no routes, you will not process the data from the form Perhaps we will see a comparison of Rails vs Nginx 😆 But let's compare Rails vs Express (the popular Node.js framework, which is not even close to rails in terms of functionality) and it turns out to be slower, see techempower
Yeah, my mistake, but I got two PRs with pure Ruby implementations, which I'll test next, as well as Rails 5-10 improvements, which I'll test alongside Django soon :) PR 1 - github.com/antonputra/tutorials/pull/330 PR 2 - github.com/antonputra/tutorials/pull/335
This is an incorrect comparison. Compare Rails vs something like Sails - that would be fair, and more "real life" - you would compare 2 full -fledged frameworks . Once you do - you will see your benchmark is going to be basicly the same for both.
Love seeing your vids pop up on my feed. Do Django vs Rails. A year ago I was getting similar throughput with a simple json response. This test was a bit unfair since node side was light weight, would have been more fair to compare against a basic ruby library like Hanami.
comparing a highly complex framework like Rails with nodejs is not fair. At least you should choose pure rack and write the code to see how different between ruby and runtime nodejs
This is a fantastic comparison between Ruby on Rails's 'full-featured framework' and Node.js's approach of 'picking your own small frameworks and libraries. This is why I like JS ecosystem. Thank you!
I am not sure that it is fair to compare rails framework with bare bone nodejs, but still the difference is too high to blame it on that Maybe next time pure ruby vs pure bun, just to make it different from the current video
Anton! love your performance test videos, give us a lot of insights and a lot of questions as well. have you think about doing some test over cache services, Redis, valkey, dragonfly as they are "forks" of the original redis. let me know what you think about this one, also if there's a close example of a similar environment on your repo I could do a contribution on this one.
Hi, many people are pointing out that RoR has features that the Node.js implementation does not have, instead of trying to replicate those features with libraries use an equivalent framework, I think NestJS (Nest with S, not X) or AdonisJS would be the right choice. Or maybe someone knows better alternatives.
no, it's in milliseconds. prometheus uses seconds for measuring histograms and summaries, but ruby on rails active record events emit duration in milliseconds. i just missed it. Here is the source code for that event - github.com/antonputra/tutorials/blob/main/lessons/220/rails-app/app/subscribers/active_record_prometheus_subscriber.rb#L67
i got a PR with a pure Ruby implementation, which i'll release in a few days; it makes a huge difference. also, i got a PR to improve rails by 5-10x, which i'll test alongside django. PR 1 - github.com/antonputra/tutorials/pull/330 PR 2 - github.com/antonputra/tutorials/pull/329
How about: SurrealDB vs PostgreSQL vs SQLite ? I know that they in different weight categories, but anyway it will be nice to see! SurrealDB seems nice, but not sure about it perfomance on a long run...
Wow, didn't expect that big of a difference. It's interesting to see the progress in these ecosystems, both in terms of performance and developer experience. 💚Node.js, 💛Bun, 🖤Deno, 💙TypeScript.
Would've been more interesting to see rails contrasted against a full-fat node framework like next.js, nest, or sails. Why even bother comparing a framework with that many levels of abstraction against a basic request router executing raw sql queries?
dude you should use nodejs framework too to make it apple-to-apple, it can be sailsjs or adonisjs if you use bare nodejs then you should use bare ruby, rails are heavier than the minimum ruby server
totally agree, i made a mistake in this video. i just got a pure ruby implementation, which i'll test with the nodejs standard library as well in a couple of days to make it right PR - github.com/antonputra/tutorials/pull/330
@@AntonPutrajemalloc can help keep memory pressure down in a long lived Rails app but the benefit tends to be fewer instances of long lived processes going OOM and crashing, and a drop of memory usage overall of about 10%. It’s an easy win if you’re hosting a Rails application but it wouldn’t have changed this test substantially
That brutal, but tbf rails is much more feature complete and some of them would slow down node too. Would love to see Ruby on Rails vs Laravel benchmark (vs Laravel Octane would be even better if you could)
Dis you really compare a full-featured and battery included project using Ruby on Rails and ActiveRecord with a barebone skeleton and minimal NodeJS? What could we expected from such comparison? What about oranges and apples next time?
Bun vs Tokio Tungstenite (maybe with tokio-postgres) vs uWebSockets : PostgreSQL Test PLEASEE SIR PLEASEEE (and maybe Go Gorilla) ❤ uWebsockets have 2 lang Node.js/C++. I guess C++ is better.
This comparison is totally absurd. It's just to stir up the discussion. To be fair, he'd have to implement the server using only Net::Http, without using Active Record (maybe only the pg gem), and thousands of other things that Rails provides. That project in Node.js only uses native runtime features while Rails is a full-fledged framework that provides a vast array of integrated tools and conveniences out-of-the-box. That was the most nonsensical comparison I've ever seen.
dam I thought RoR would be faster by 2024 since Shopify and GitHub threw millions of engineering hours trying to speed it up 😮 looks like a horse vs car race
They have far more complicated systems than simply relying on a single framework. Most of the system in modern world are always bottlenecked by databases and storage. You can scale rails if you have enough money, but dealing with databases, caching and other storage related stuff is going to be the most complicated process for any big software.
Ruby is so bad. 50x difference? the dev productivity has to be at least 500x for me to even consider it, it should automatically write my code when I create new file otherwise I won't even bother.
RoR isn't the ideal framework in any case, anywhere, anytime. It's awful. Aside from bad performance, it's filled with antipatterns. The only reason people use RoR is if they don't know better or refuse to replatform to a serious language and framework. The "developer happiness" is because ignorance is bliss
🔴 To support my channel, I'd like to offer Mentorship/On-the-Job Support/Consulting (me@antonputra.com)
Elixir vs Node
SQLite vs Postgres
Elixir vs Go vs Bun
Mind giving Swift Vapor a spin?
Prisma vs Drizzle
I would love to see a benchmark comparing traditional runtimes against Beam based ones. Maybe Node vs Elixir?
ok i'll take a look
@@AntonPutra If you have time, it would be interesting to see Gleam with Wisp for web framework in the test. Gleam is a BEAM language.
Ruby on Rails vs Laravel
i'll do php soon
@@AntonPutrafpm , swoole, webman
Rails vs laravel vs Django!
Hahaha. Anything PHP is extremely controversial! Get ready for the comments section of that video
@@greendsnow 😂😂
Just want to say thank you, really helpful for me ❤
Compare Rails vs other frameworks like Hanami if you have time
I think it can only be fair to compare Rails with Node framework like Express, Nest, Fastify, etc. Love your channel btw
thanks you! make sense, initially i wanted to compare with django
@@AntonPutrai would love a comparison between Django Rails and Laravel
@@AntonPutra I would be interested in django vs rails
@@Bob-qz5yj A comaprison between all major mvc framework would be great!
Whenever I see a new video, I check it if its Elixir vs something else. :) Do you have a plan to have a stress test elixir against golang or node?
i'll do elixir very soon 😅
@@AntonPutraamazing, thx 😅
Sadly, this video compare orange with apple. For example, the SQL insertion is made in one side using ORM ActiveRecord, and in another using raw INSERT query (unsanitized btw).
Use ActionController::Metal, use PG driver and do the tests to compare orange with orange.
Add YJIT, check for CoW for the workers (because it's not normal that it is using twice the ram for two workers btw) and then, yes, node will still be faster, but not by this crazy factor.
thanks for the feedback
@@anykeyh I met so many people in the past couple years who recently migrated from Ruby and PHP to Go. Is Ruby really no more popular? Feels like it faded away momentarily after Heroku’s demise.
@@Nick-yd3rc Ruby faded away with SaaS gold rush. In 2010, building as SaaS was one of the best way to get rich quick. And RoR was praised to its quick development process, where all which matter is time to market.
Ruby is a great language because of its syntax allowing DSL, and its awesome standard library. It's super useful for web, but also for scripting and devops. It is still well maintained and active, noticeably thanks to Shopify work later.
The main competitor of Ruby is Python, and it gained a lot of traction because of AI and general academic use. I personally prefer Ruby for plenty of reasons. I think people should rationalize their need of performance in a web application.
HTTP and Web has never been designed with performance in mind.
While it's super great to have something going very fast, it's often a non-functional quality that you don't need.
Most web servers are running application which are not exploited to their full potential, and if I spend 1 millions a year in cloud services, usually the part related to servicing web request is 10 to 20% max.
Of course, if you plan to create a trading platform or a B2C platform and your revenue per request is extremely low, you might want to use faster language.
I would never use Node as server-side language. Any serious application would ship with 1000+ NPM packages, which is insane and will cause you headache during maintenance and migration to newer versions. And javascript/node quirks and terrible std library (in my opinion) doesn't help to produce clean code, even using Typescript for example.
Go is a great alternative, but the language is dull, by design (allowing easy maintenance). While I understand perfectly the approach of the language creator, I rather not spend 8 hours a day dealing with it. It's a question of personal choice here. Go is great, and if you like it, go for it.
@@Nick-yd3rc i'm not from the United States but i used to follow their job market there around 3-4 years ago, idk about now but in the past there was such a labor shortage of Rails devs that you can get quite a bit higher salary than Laravel dev for example. Where i live if you know .Net or Spring, you can get job pretty much any time (most of the full stack/backend job offers are for those 2) after that it is Node+Express or Php with Laravel or symphony and very rarely Go (albeit we do have slow increase in job offers).
@@AntonPutraUse the ORM Drizzle for Node.js
@AntonPutra really appreciate your videos however for tests like this (and 209 - Django vs. Go) there's a bit of an apples vs. oranges comparison.
Both RoR and Django start with more than a server like NodeJS & Go. E.g. middleware, Django has CSRF protection, XSS etc. all enabled by default. This functionality is not enabled in Go/NodeJS without additional libraries or function calls so it creates an imbalance in functionality and skews the test.
Would be great if you could either a) remove middleware and other niceities built into frameworks (although that is not how they are used) or, b) alternatively expand the Node & Go code to match the functionality. This way if a dev was trying to evaluate a framework in different languages on performance alone it would be a fairer comparison.
Thanks & happy hacking
Agree, he should have used something like Sinatra. But Ruby would still lose big time here.
make a PR on the repo
make sense
@@ha5anmukhlis I would love to if I had time. Unfortunately I don't but I wanted to flag it incase someone else did have the time as these benchmarks are very helpful.
Exactly this. He compared a full-featured project with a minimal NodeJS skeleton with raw SQL query.. omfg
you could consider announcing next test in advance: prepare the repo, give people some time to review and then run test with feedback applied
i tried that, but it didn’t really work. i get the most valuable advice only when i release a video, people like to criticize, and i don’t mind 😊
@@AntonPutra
People only start to care when they see their favourite framework/runtime gets devestated in a public benchmark 😆
And only THEN they will bring their Sith-mastery optimization techniques with them.
@@AntonPutra maybe you're right, it's easier to watch the video and criticise than review first and wait for the video. But honestly I've been watching your tests for a while and haven't noticed dedicated segment with an announcement of next test (what exactly will be tested, here link to repo, ideally after intro). On the other hand I haven't paid attention to it, and if I'm wrong you can disregard my thoughts :)
Why do you compare large framework with millions line of code with pure runtime code without any dependencies? Rails can be compared only with Django, though it still has less functionality. Compare apples to apples, use pure Ruby API to compare with Node.js
😂 because of us, his audience wants to see
true, it was mentioned many times, so i decided to test it
For each video comparing two interpreted languages I’d love to see a go, zig or rust baseline to drive the point home about interpreted languages vs compiled.
noted
@@AntonPutra I'm loving these videos - keep up the great work!
That seems a bit pointless because we know those are faster. Also, the graphs would become unreadable. On the contrary I think he should only be comparing things expected to have similar performance, so that we might actually learn something new if they don't.
C# (minimal api) vs Node.js (with instance per core) would be nice
ok will do
These videos are very interesting thanks for all your hard work!
Spring Boot Vs NodeJs
noted
would be interesting to compare ror with django
yes, those two frameworks are much closer to each other
you compared the warm with the soft:
- Rails is a framework with a full-fledged ORM, a large number of abstractions that simplify development
- Node.js is just a JS runtime with a web server that does not even know how to work with cookies, there are no routes, you will not process the data from the form
Perhaps we will see a comparison of Rails vs Nginx 😆
But let's compare Rails vs Express (the popular Node.js framework, which is not even close to rails in terms of functionality) and it turns out to be slower, see techempower
Yeah, my mistake, but I got two PRs with pure Ruby implementations, which I'll test next, as well as Rails 5-10 improvements, which I'll test alongside Django soon :)
PR 1 - github.com/antonputra/tutorials/pull/330
PR 2 - github.com/antonputra/tutorials/pull/335
This should've been a Rails vs Adonis comparison.
Express is died framework
@@KravetsFamily What are Node.js folks using these days?
This is an incorrect comparison.
Compare Rails vs something like Sails - that would be fair, and more "real life" - you would compare 2 full -fledged frameworks . Once you do - you will see your benchmark is going to be basicly the same for both.
I like this comparison. For me, its a statement of how "good enough" for 99.9% of websites Rails can be.
Do you have videos on how to setup those kind of metrics on Prometheus/Graphana?
yes, i have a bunch of them, just search for prometheus on my channel
Love seeing your vids pop up on my feed. Do Django vs Rails. A year ago I was getting similar throughput with a simple json response. This test was a bit unfair since node side was light weight, would have been more fair to compare against a basic ruby library like Hanami.
Any plans on RoR vs SpringBoot or Go?
well maybe in the future
Thank you for the video 🙏
my pleasure! i'll update this benchmark with pure Ruby implementation in a couple of days
@AntonPutra it would be great!!
PHP 8 vs Python pls
ok will do soon
PHP will win so hard haha
@@AntonPutra Don't forget to fine tune PHP as well ;)
I Hope for every developer that your app will get to the point, when it gets hundreds of request per second.
😊
What about Scala vs Go?
in the pipeline 😅
comparing a highly complex framework like Rails with nodejs is not fair. At least you should choose pure rack and write the code to see how different between ruby and runtime nodejs
i understand, i'll do ruby on rails vs django soon
@@AntonPutra Django on PyPy xD
interesting would be RoR vs Laravel?
maybe, but i don't think people care about RoR anymore
Could please make some benchmark about aws lambda with different runtimes? Thanks.
This is a fantastic comparison between Ruby on Rails's 'full-featured framework' and Node.js's approach of 'picking your own small frameworks and libraries.
This is why I like JS ecosystem. Thank you!
You didn't close the second single-quoted string.
But good point.
Take Agoo, Roda or Hanami - and your Ruby app will run circles around Node.js
Балдежные видосы, удачи
spasibo :)
Love to see benchmark on PHP with database connection ❤
Pure php vs pure python . Laravel vs lumen. ❤
ok noted
Can you please compare Rails to Roda and Hanami (other more lightweight ruby frameworks) as you did with node/bun/deno ?
ok maybe in the future
It would be nice to see performance comparison of JWT tokens (their creation and validation) in different languages such as NodeJS, Bun, Go, Rust
Do you think doing SvelteKit vs NextJS is possible? Or would that be going too deep into the JS ecosystem for you?
I want to know Python - Fast API vs Other🤤
🫡
Wonder if jruby would change results much
I am not sure that it is fair to compare rails framework with bare bone nodejs, but still the difference is too high to blame it on that
Maybe next time pure ruby vs pure bun, just to make it different from the current video
true. i'll see if it makes sense to test ruby with anything else
Next, Elixir vs Go vs Bun
ok soon
Anton! love your performance test videos, give us a lot of insights and a lot of questions as well. have you think about doing some test over cache services, Redis, valkey, dragonfly as they are "forks" of the original redis. let me know what you think about this one, also if there's a close example of a similar environment on your repo I could do a contribution on this one.
Hi, many people are pointing out that RoR has features that the Node.js implementation does not have, instead of trying to replicate those features with libraries use an equivalent framework, I think NestJS (Nest with S, not X) or AdonisJS would be the right choice.
Or maybe someone knows better alternatives.
yes, pure nodejs was not the best choice for this comparison, my mistake
Have you tried to test python and php? I am currently working for a company uses those langs
python, yes. i'll do php soon as well
If the p99 latency is 17 seconds, then how can availability be 100% when the time limit is only 1 second?
no, it's in milliseconds. prometheus uses seconds for measuring histograms and summaries, but ruby on rails active record events emit duration in milliseconds. i just missed it. Here is the source code for that event - github.com/antonputra/tutorials/blob/main/lessons/220/rails-app/app/subscribers/active_record_prometheus_subscriber.rb#L67
How about Spring Boot vs Go?
I wonder if JRuby helps Ruby on Rails perform better
i got a PR with a pure Ruby implementation, which i'll release in a few days; it makes a huge difference. also, i got a PR to improve rails by 5-10x, which i'll test alongside django.
PR 1 - github.com/antonputra/tutorials/pull/330
PR 2 - github.com/antonputra/tutorials/pull/329
How about: SurrealDB vs PostgreSQL vs SQLite ?
I know that they in different weight categories, but anyway it will be nice to see!
SurrealDB seems nice, but not sure about it perfomance on a long run...
Wow, didn't expect that big of a difference. It's interesting to see the progress in these ecosystems, both in terms of performance and developer experience.
💚Node.js, 💛Bun, 🖤Deno, 💙TypeScript.
And in terms of DX, I personally ❤NestJS.
i had to test rails at least once
Could you please try next, sveltekit and nuxt? Thanks and amazing content ❤
noted!
@AntonPutra Thank you
Node js vs .net ?
ok noted
PHP Symfony next ?
i'll do php soon
Surprising results.
Please do dotnet vs node/bun
interesting, noted
We need more rust videos.
haha ok
next, Elixir vs Gleam
noted, i'll do Elixir in near future
Elixir and Gleam both run on same Beam VM so it's not so interesting comparison. Much more impotant between them would be the dev experience.
I'm interesting in Node.js VS Python FastAPI test :)
ok soon :)
Laravel vs Rails Please
ok i'll do php soon
Would've been more interesting to see rails contrasted against a full-fat node framework like next.js, nest, or sails.
Why even bother comparing a framework with that many levels of abstraction against a basic request router executing raw sql queries?
makes sense. initially, i wanted to compare it with django
dude you should use nodejs framework too to make it apple-to-apple, it can be sailsjs or adonisjs
if you use bare nodejs then you should use bare ruby, rails are heavier than the minimum ruby server
totally agree, i made a mistake in this video. i just got a pure ruby implementation, which i'll test with the nodejs standard library as well in a couple of days to make it right
PR - github.com/antonputra/tutorials/pull/330
Was this running with the jemalloc setup on the Rails?
no, i don't think so
irb(main):001> RbConfig::CONFIG['MAINLIBS']
=> "-lz -lrt -lrt -lgmp -ldl -lcrypt -lm -lpthread "
@@AntonPutrajemalloc can help keep memory pressure down in a long lived Rails app but the benefit tends to be fewer instances of long lived processes going OOM and crashing, and a drop of memory usage overall of about 10%. It’s an easy win if you’re hosting a Rails application but it wouldn’t have changed this test substantially
I’d also be interested to see the difference with the various jit’s in use.
That brutal, but tbf rails is much more feature complete and some of them would slow down node too.
Would love to see Ruby on Rails vs Laravel benchmark (vs Laravel Octane would be even better if you could)
my mistake
No problem - they are still “close enough” for this benchmark to be interesting and useful to people
Great video overall
Python std vs flask vs fastapi
it maybe interesting
Dis you really compare a full-featured and battery included project using Ruby on Rails and ActiveRecord with a barebone skeleton and minimal NodeJS?
What could we expected from such comparison?
What about oranges and apples next time?
my mistake
Would like to see rails vs Django - more accurate framework usecase
agree, it was my original idea... not sure why i used nodejs
your blurring doesn't work mate. I can clearly see which is whic, even if you switched the colors :D
i know 😊
I don't think this is a fair comparison, on the node.js side, I believe Nest.js will be a more closer comparison
Adonis, I would not cast nest in framework category as it does not provide even quarter of what Rails offers.
@ijazkhan3335 hmmmmm that's true, Nest isn't batteries included like Adonis and I think Sails
probably true; i'm unlikely to test rails with anything else
@@AntonPutra with django thats the all batteries included framework and fair comparision
Its not fair to compare framework vs pure language. You should compare ruby vs node.js or frameworks vs frameworks like django sailsjs phonix
agree, my mistake
Sir, Express.js vs Next.js please
the ideal would be rails vs another similar framework that runs on nodejs
makes sense. initially, i wanted to compare it with django
maybe u can try Elixir to counter the power of Nodejs
yes i'll do Elixir vs go soon
@@AntonPutra with Rust too for get more experience 😃😁
I am proficient in both languages. I like javascript but I love ruby.
1200 requests per second is good for me even though I don't know anything about ruby.
😊
Bun vs Tokio Tungstenite (maybe with tokio-postgres) vs uWebSockets : PostgreSQL Test PLEASEE SIR PLEASEEE (and maybe Go Gorilla) ❤
uWebsockets have 2 lang Node.js/C++. I guess C++ is better.
ok, i'll do it sometime in the future 😊
Thank you very much, I am looking forward to it
a bit weird comarison, rails is really slow there is no secret in that, try ruby iodine
ok, will do. rails is just the most popular framework in my mind, at least that's why i used it.
☝ Please, C# with .NET 9 vs (Java with Spring Boot and Python with FastAPI), all of these using PostgreSQL. Thanks in advance! 🙌
thanks for the suggestion, just added to my list!
php vs python vs node is still here :p
Sinatra vs Flask vs rails vs django
noted!
fastapi when? :3
haha soon :)
I don't think anyone actually uses ruby on rails. You promised PHP swoole bruh, yet you give us something nobody cares about.
seems that way, but i used it myself many years ago and decided to try it out. so far, it's the slowest thing i've tested 😊
Fun fact gitlab (and I think GitHub) is ruby. Not sure if they’re rails though.
Wow, ruby is weak!
In a recent language perf test among all the major production languages, 72x slower, and python was around 50x
But it gets the job done with Rails compared with Java and Spring, the productivity in Rails is up normal.
Its not ruby, its rails that is a heavy framework
it's mostly rails so slow
Some how an event loop based framework for Ruby never really took off, the thread based sync approach really hurts it in tests like these
Spring boot vs jakarta ee
PHP Swoole vs Node
i'll do php soon
@@AntonPutra PHP has many runtimes, FPM, Swoole, Franken, similar to Ban, Deno and Node, maybe a comparison between runtimes first would be nice
I think all the scales should be logarithmic for this one 😅
Laravel vs Ruby on Rails
i'll do php soon
Phoenix Framework vs Ruby on Rails
Why are you comparing a framework with a runtime, it makes no sense
my mistake, just released the updated benchmark!
@@AntonPutra Awesome, quality content
This is not so surpsiring .. I would love to see Elixir 😅 It would maybe loose on one PC but win on many.
i'll do Elixir vs go soon :)
The only fair comparission would be Rails vs Django since none of this js frameworks have everything those two have
agree
ultimate-express vs spring boot
doesn't even make sense to compare a framework to a runtime. Try RoR vs NestJS or ExpressJS
This comparison is totally absurd.
It's just to stir up the discussion. To be fair, he'd have to implement the server using only Net::Http, without using Active Record (maybe only the pg gem), and thousands of other things that Rails provides. That project in Node.js only uses native runtime features while Rails is a full-fledged framework that provides a vast array of integrated tools and conveniences out-of-the-box. That was the most nonsensical comparison I've ever seen.
ror vs laravel will be best
ok i'll do php soon
Ruby on Rails vs Elixir Phoenix vs Python Django vs PHP Laravel
ok noted!
I add NestJS and Java Spring
interesting to see gleam :)
compare maybe with go / nodejs
seems I got answer why RoR "died"
i'll take a look
HTML5 vs BASIC vs Django
forgot assembly?
dam I thought RoR would be faster by 2024 since Shopify and GitHub threw millions of engineering hours trying to speed it up 😮
looks like a horse vs car race
They have far more complicated systems than simply relying on a single framework. Most of the system in modern world are always bottlenecked by databases and storage. You can scale rails if you have enough money, but dealing with databases, caching and other storage related stuff is going to be the most complicated process for any big software.
yeah, rails out of the box is quite slow
Ruby is so bad. 50x difference? the dev productivity has to be at least 500x for me to even consider it, it should automatically write my code when I create new file otherwise I won't even bother.
there is nothing about Ruby in this video, it is Rails
it's mostly rails that's slow
🍿 Benchmarks: th-cam.com/play/PLiMWaCMwGJXmcDLvMQeORJ-j_jayKaLVn.html&si=p-UOaVM_6_SFx52H
once again node reigns supreme
😊
Ruby should not be used in production systems. Use any other language
i mean, it was popular some time ago
RoR isn't the ideal framework in any case, anywhere, anytime. It's awful. Aside from bad performance, it's filled with antipatterns. The only reason people use RoR is if they don't know better or refuse to replatform to a serious language and framework. The "developer happiness" is because ignorance is bliss
make sense 😊