We dont need meme C performance

you man-child kidos all day silling for an obsolete language like C, repent right now.

picture much related

Other urls found in this thread:

wiki.python.org/moin/PythonSpeed
github.com/magnumripper/JohnTheRipper
twitter.com/SFWRedditVideos

even better: use go and don't waste your time on hisssssss

Real problems begins pythons could be 50 to 250 times slow C.

>those mental gymnastics
wew

I've seen many C tards doing mental gymnastics but this one takes the cake.

runtime isn't all that matters. You would know this if you ever had a job
>C autist spends 3 weeks coding a project
>stays at work late every day troubleshooting bugs and getting chewed out by his boss for working too slow, while wife spends the night with Pajeet

>Python dev spends 3 hours by importing solution_someone_else_made
>gets a raise and is given the rest of the day off to go fuck his wife

>C autist boasts about how his code runs 0.09 nanoseconds faster
wew lad

Damn daniel, projecting this hard...
You should see a doctor, you must not be far from a total mental breakdown

>projecting
that's not what projecting is lad

yes Daniel listen to him and the meme magic, you will soon be dead if you keep that edgy path

Also those rdy to use libraries end up being a nightmare i saw that on golang with C libs, just fucking pathetic form of programming

jython, ironpython, and pypy are all fast and nobody will use them

believe it or not, not all programmers program shitty 2D RPGs.
Try making an optimized primesieve or 3D renderer then come back with your arguments.

are you telling me you never use any libraries?

>We don't need efficiency!
>WAHHH! Why is my browser so slow and using 4 gb of RAM?

>2) A way to blame others for your own negative thoughts by repressing them and then attributing them to someone else. Due to the sorrowful nature of delusion and denial it is very difficult for the target to be able to clarify the reality of the situation.

>1) Believing that someone else does not like you when it is indeed you that does not like them. By projecting this onto another you ascribe the negativity of the thoughts/feelings onto them so your ego does not have to admit the deficiency of your own thought processes.

Both sound relevant to the situation.

But he's right. What's the point of spending a week writing a tool in C when you can get it done in a few hours with an interpreted language when you won't be using it that often anyway? There are certainly times when interpreted speeds are unacceptable, but it's worth it --- for your own sanity --- to find out what's more important: getting the tool done fast or making the tool go fast.

If you are writing something where performance is needed a nanosecond can be extremely useful when something is ran tens of thousands of times(e.g a kernel function). Every language has a purpose, get over it.

no they don't

no man no, ofc i do but common linking to C and C++ libraries to a language that works in a different manner is just setting your self to a disaster, for the ones written in the native code... well if you abuse it your program will be eternally updating LE ruby cool much libraries.

It's probably more extreme than that.

The Python version would probably run a few milliseconds slower. Your boss still wouldn't care about that.

There are very few areas that actually require C speed. You would know if your work falls in one of those areas.

ok lets dish python for bash

kk boss

ehhh performance is overrate lmao

>0.09 nanoseconds faster
>than python
Try 9 minutes faster than python

>And all those run very slow
This is true

True story

>Usually it is not the absolute speed that's important
Sometimes it is just about absolute speed though. In a lot of domains, your software is better if it runs faster.
And those domains aren't just small niches. E.g.:
Embedded Systems are very common and they're very constrained systems.
Similarly mobile devices. Being efficient saves battery.
Data science. Data can be so big nowadays that your software is your largest bottleneck.
Operating systems or system libraries. If your OS is slow, then anything running on it / anything using it will also be slow.
Then there also are domains where you can't be fast enough, like simulations and video games. Having more fps or more detailed gameplay/graphics/AI is always better.
All of those examples get better, if they're running faster.
That's why you don't see any scripting languages as the main language in environments where you're dealing with efficiency constraints.

yes, and there are some cases where using python is perfectly acceptable. runtime isn't the only factor, yet this is all you autists worry about.

If you can't decide what language is appropriate for a task, you aren't competent enough to perform said task to begin with.

>Data science. Data can be so big nowadays that your software is your largest bottleneck.
>implying python is not used for data science

There are languages that are between the extremes of C and Python that are more ideal for balancing runtime speed vs speed of development.

Unfortunately
- you can have a better abstracting, more productive but still far faster language
- slower == more power usage for the same task which will be relevant in a mid term future
- all Python implementations are far below acceptable speed for lots of tasks

Why does it tell me what to think? What kidn of faggot wrote this piece of shit?

>all Python implementations are far below acceptable speed for lots of tasks

then write libwhateverbullshit in C to do the performance critical part, then call it in your preferred scripting language.

wiki.python.org/moin/PythonSpeed
They them self are writing this, you cant make this shit up

>They them self are writing this, you cant make this shit up
>They them self
yeah they should have hired you to write for them

writing with proper grammer doesnt mean what you write is worth reading

>some cases where using python is perfectly acceptable.
yes, of course. I just wanted to show that there are a lot of cases that aren't just small niches where python will not run fast enough.
Not everything needs the absolute best heavily optimized performance though.

Tasks like training even simple neural networks takes a few hours with python. If you have that time, that's ok. Then you can use python.
Obviously not all data science tasks require to have the best performance, but having software that scales to large data sets easily is what sets companies like google apart from others.

NO! I NEED TO ARGUE MY IDEOLOGY ON Sup Forums!!! YOURE EITHER WITH ME OR AGAINST ME!!! BURN THE GC BABBIES

true but writing with the skills of a 4th grader means whatever you wrote is not worth reading

kek

I guess someone should tell Larry Page and all the Google devs to re-write TensorFlow in C.

>le grammer falacy attack point
fuck off idiot i can write in 4 languages and english is not my native one.

The real fucker with taking this idea on board is when commercial programs written in Python slowly outgrow their old environment.
A payment handler near me did all their old work in Python because it was "fast enough" for the time, 10 years later their business has grown way beyond their original scope and the central pieces of their code are still crappy slow Python scripts. Bonus fuckery is that the guy who actually wrote them couldn't comment worth a crap and they still find bugs (some of them in security related things like the encryption algorithm) to this day.

As far as I know they still haven't scheduled a switchover to a sensible codebase despite the fact that their software falls over under load several times a day and could easily have leaked customer information if anyone knew where to look.

if you're just calling into someone elses C library it doesn't matter that much

if you're writing something you think is useful and will be widely distributed it should have a nice C interface

You can be fast to execute, fast to develop in, and not have a GC. See: Rust.

Sounds like bad software dev practices. I doubt there'd be less bugs if it was written in C. Goes to show that you can't let shitty software rot away just because "it works good enough right now"

I think you might have missed the part about "ideology"

Ideology needs to be kept out of programming.

I like compiled languages, because I don't have to install an interpreter on another computer to use it.

This.

(you)

>Rust
>fast to develop in

this.
try, for example, pyrit, a wpa hash cracker... you'll notice that, for some GPUs, it's even faster than competitors (hashcat, John the Ripper)

of course it's not all python, but pyrit still manages to be nice and fast

people still use John the Ripper?

github.com/magnumripper/JohnTheRipper

Turns out it doesn't work that well for anything remotely non-throughput oriented.
I rather write my stuff completely in D or [other faster but also more productive, statically typed, faster compiling than CPython is starting language] and enjoy the spoils of fast startup times, global optimisations and balanced latency.

To expand this a little, the only time where bindings are a good idea is when you basically have like 20 calls or so like in youtube-dl.

What is cython

incompatible, yet not fast enough to be worth it.

Fuck off Zed Shaw

that's a very obvious fact except to NEET autists, programming is all about making money niggers, it's romantic to think of it as an intellectual activity, but it isn't

there's 3000 lines of c code in there bucko

>rust
>fast to develop in

maybe if it didn't have the idiotic system of borrowing.
but with it, its just another one for the trash can

>running large, but simple multiplication calculations
>it's literally just multiplying twenty e-notation numbers

Python:
>takes 12 seconds to exit
>noticeable delay between each output

C:
>1 second
>compiled in 2 seconds

I'm not even autistic enough to care about: """python""" programming """language"""
but, this is ridiculous.

>what is cython
C with a very light coat of Python, and missing fingers and a leg.

exactly, you fucking retard. Python isn't just python.
try guessing the name of the most commonly used implementation...

did you even read the pic in the op? did you consider the time YOU SPENT WRITING THE C PROGRAM?

I did.

It took me roughly the same amount of time for a small program. I know it becomes more arduous the larger the project size, but it's easier to maintain and doesn't take a huge performance hit like Python does.

being fast can be important, but it mostly isn't, and assuming your language doesn't suck, you can just rewrite the bits that outright need to be fast in C
having code that doesn't fuck up is more important than anything else

I like writing in C, but the only times I've needed to are when writing shit for old systems (was writng a DOS program doing direct low-level hardware access not too long ago)

If all you're doing is multiplying floats there should be no descernable difference, not including the initialization of the interpreter. I'm not even kidding you, simple shit like that is extremely straight-forward to translate to ML.

Most likely you're using some expensive operation without even realizing, because the syntax is similar to C. Post examples of both

>up to 20x the memory usage
How can you justify using python?