Python: keeps driving while pitstop people are running EDIT: Golang car being a garbage truck that weighs 10 tons with a 500 horsepower engine EDIT 2: C car driving at 1000 mph and skipping pitstop, occasionaly catching on fire
For real. I’m out here compiling UE4/5 for like 2 hours. At least it only takes that long the first time you build, and every time after that it’s like a minute at most…
@@that_guy1211 so u know that if u have like more than one (1) template it immediately slows down a lot? lots of templating in a c++ project can make it as slow as rust
@@meninoesperto2773 Average leetcoder mindset, theres a reason why python/js cannot be used to power any realtime app of large scale. And interop with C doesn't count, thats just writing C and then attributing all the performance to JS/python
Rust glazers are gonna learn soon 😂😂 in 2 years everyone will regret wasting their life learning/using rust and c++ will prove superior to Java and rust and python and all goofy ahh languages
Python is like 4 people on top of a race car changing the tires while the car goes at two km/h, but sometimes the 4 people randomly can’t figure out how to screw in the nuts of the wheel.
Nobody uses Python for performance (or lack thereof), it is famously slow. It does have some very useful features and fields where it excels, but speed (at any stage) has never been one of them.
@@j.r.r.tolkien8724 You do realize that I am not the one who brought up the "speed of Python" right? The entire video and title is about compilation, so yes, I fully realize this, and was merely replying to another's comment.
@@ZihadJoy python is interpreted, and interpret time is pretty much the same as runtime, because interpreting is going line by line through the program and doing what it specifies, which takes as long as running the program does.
@@rule6elur still compiled. Machine code or not. The byte code could then be ran by either a JIT compiler or an interpreter. The actual process of running python still involves compiling one way or another.
Coding in Rust, I don't think compilation time is the worst experience about it. But given the single factor you consider on using Rust or not, I assume that you also don't know the best things in Rust.
once you get past the initial compilation of your dependencies, rust's incremental compilation combined with debug releases tends to make it a very tolerable experience
@@mr_sauce_cooks whether u wait 5 seconds or 10 seconds for a compilation is the last concern when choosing a language rust’s compile time is high not cos of the compiler itself but its inherently more work to do the borrow checking and all the other safety things that rust has to do
Nah we busy pitying the c/c++ devs who don't have proper toolchain tech and have 50 years of duct tape tech debt slapped together and kept running by hopes and prayers.
@@fredesch3158 that he does. For curiosity sake just checkout any meme on rust. There will be a joke about butthurt rust devs and a huge amount of replies from actually butthurt rust devs doing exactly this, it gets funny how they try to actually refute a joke.
Python is a interpreted language, we're talking about compiled language. Interpreted means it runs the code directly (That's why Python is slow), compiled languages convert your source code to machine code before running it, that's why C++ and other languages that are compiled are so fast.
@@GDT-Studio Kind of incorrect. Python is an interpreted language, yes; however, it is primarily just the bytecode that is interpreted. Initially, the Python code is compiled to bytecode with a JIT compiler and then is sent off to an interpreter that then runs the bytecode. Python would technically fit in this race, it's just that the pit crew would be trying to change the tires while the car is driving on the track. xd (referenced from susamogus11111's comment)
Never seen a comparison which values compile time and nothing else. Thats like the least valuable virtue. Especially since many times you don't have to recompile the symbols if they haven't changed.
@@pruthweeshasalian3688 yep, its funny. Could have done the same thing with C++ performance in comparison. But I guess its funnier putting Go as the fast one.
@@macicoinc9363 really? It's a one time cost. My big tech has massive server farms that run multiple builds in parallel CI/CD. Consolidating stable prebuilt libs solves having to build everything from source. While it may take a long time to build one instance, development can also multi-task. This is fundamentally what you call DevOps. I will always compromise compile time for production performance.
I haven't worked with java , but i have tried kotlin recently and its awful in terms of speed , so i thought java compile time should be close to it@@theshermantanker7043
Nothing prevents a programmer against writing buggy code: 95% of bugs (c++/c#), which I fixed were related to business logic - no language can solve that. Also there is no problem to make a lot of errors in Rust related to the deadlocks.
c++ depends. If you use Clang in combination with Make as Generator you wait for ages. Also if you have programms with a lot of meta programming its becomming worse... With g++ and ninja it's better...
@@user-uf4lf2bp8t the thing is that python code is mainly interpreted so that's why it compiles fast(interprets fast) but actual code is slower than a snail
Both have their use cases. Comp speed is kind of trivial though, as you only compile the finished project once. Also, with rust, the second compile is much much faster because it doesn't recompile dependencies.
Nope, I program in Go, Rust, C and now Zig, anyone with that tought of "They all feel incomplete" are just wrong because they feel so good being baby sit by the compiler, unable to deal with memory management in other languages
You shouldn't compare Rust compilation time with other languages, you should compare Rust compilation time with the time you average developers taken to debug in other languages.
For those who don’t know anything about programming like the user who posted this video this is the kind of youtube crap that u should forget u ever saw it
It's true. Of course, rust has advantages in that it's faster at runtime, safer in terms multi threading smd also the compiler does a lot more thorough error checking.
@rico_1617, do you know the llvm ecosystem? Clang, as well as the rust compiler, are just syntax variations with other rules. They ultimately get optimized by llvm. Clang and Rustc just parse the language (do eventuallchecks) and feed llvm the tokens needed for compiling.
@@luigidabro It's actually even worse for Rust. Compiler frontend MUST do optimizations as well. How else would C/C++ be able to optimize out entire memory allocations? 20 lines of basic self contained code? Maybe you'll get the exact same result in C, C++, Rust and Zig, but it doesn't scale much further. It will eventually diverge and language + your language proficiency + compiler frontend starts playing the main role.
@@logicaestrex2278 You do you. I prefer python for many things and people sometimes judge me for it. Use whatever gets the job done that you need to do. That is enough for 90% of people.
That's objectively wrong. If I want to debug and each time I want to launch the program I have to wait a minute when it could just be 5 seconds then this is a very real disadvantage of the language.
@@sayven if you are waiting a whole minute for your build to finish you did not set up your build pipeline correctly. Even in compiled languages that run on native hardware (like C, C++) your build times would be reasonable if you used precompiled headers, modularized your code properly, used dynamically linked libraries where appropriate, etc. Aside from that, a project will get built a limited number of times, but will need to execute many many more times on countless units of hardware, so in cases where power usage and/or speed is a concern, compilation times is a minor downside. If you are working on a project where execution speed is not of critical importance (for example, if you are blocked by network I/O or if hardware resources are overkill for the problem you are trying to solve) then you can use high level interpreted languages (like JavaScript or Python) and have a better developer experience. you don't use C/C++ because of developer experience, you use it for resource efficiency and speed. comparing and ranking statically typed compiled languages based on build times is a silly exercise. similar to ranking airline companies for their cuisine. you don't get on a plane to eat, if gastronomy was your concern you would go to a restaurant.