C++ is for people who want to work

C++ is for people who want to work.
Python is for people who want work done.

Attached: serveimage.jpg (1280x720, 55K)

Thank You Kanye, Very Cool.

Please remove this meme, it's banned in the EU.

im not an eurangutan

C++ is for people who want work done.
Python is for people who want one single CPU core to work.

Parallel programming is a shitfest. Better to write optimised single core programs in python instead of being a hotshot.

C++ to sacrifice development time to gain computational speed
Python to sacrifice computational speed to gain development time

You really must not be a programmer

Parallel programming is downright unavoidable for a large number of programs

>i don't know what i'm talking about

Most devs can barely handle git, let alone thread safety for a complex parallel task.

C++ is for alpha chads
Python is for tranny redditors

That tells me that most devs should be doing something else with their lives.

Attached: 0.jpg (480x360, 10K)

>complex parallel tasks
the GIL makes python programs deficient even if the only use of threads is e.g. to make a single event loop.

Use multiprocessing. And node has the same problem, CPU-bound functions block the process.

now my large application which doesn't benefit in any way from being distributed is a collection of 2 dozen processes.
gee thanks

/thread

Found the manjaro desktop ricer

>Python is for people who want work done.
You're implying that getting work done in Python is easier/faster than C++ ?

python is a learning language.

please try again after you finish your homework

/thread

Question:

I'm a computational biologist, so I tend to use python over C++.

I analyze BAM files which are binary files that contain DNA sequence alignment.

There is a C library called htslib (High Throughput Sequencing Lib) and a suite of command line tools wrapped around htslib called samtools.

The python community developed a library called pysam which is the C library htslib wrapped in Cython.

Likewise the C++ community developed the library BamTools which is the same htslib wrapped for C++.

So if C++ is superior in speed than python, why is it that python/pysam is much faster than C++??? It takes longer to script out jobs in C++ than python, and in the end the job runs slower.

I think it has to do with C++/BamTools initializing strings that make it slower than python/pysam.

Does this make sense? I thought C++ was always faster than python since you can optimize the compilation and use defined types (short vs int)

Attached: 1526386978527.jpg (613x587, 11K)

>Does this make sense? I thought C++ was always faster than python since you can optimize the compilation and use defined types (short vs int)
remind yourself that you're talking about a wrapper

Python is for throwing together quick shit
C++ is for writing actual applications, systems programming, and embedded

>optimised single core programs in python
lawl

>most devs suck

sounds like you have found the problem

the actual code isn't in python, it makes requests to the underlying library which is written in a different language

At the end of the day, the developer is programming in Python and getting the job done faster (and in this case better) than the developer programming in C++.

Nah, it just means that the developers of the python library know their shit while the c++ ones used it to learn c++. Turns out they didn't learn shit.

Uhh yeah. If by getting work done you mean waiting 8 hours for that piece of shit to do what takes C 3 minutes that is.

This really. There are better C++ libraries for htslib now but I haven't learned the API. I'm trained as a biologist who went into programming so I'm not familiar with how to correctly use malloc and lower-level functions like that.

In my field, python is the common tongue. C++/C (and perhaps even Java) are the languages for high performance programs written by CS wonks who work in biology labs.

All of them suck big black fat cock. Once daddy Linus told anyone already that C is the only way to go.

So, you are a nigga or a shinese?

Nope. Even though both are supposedly wrappers of the same library they are completely different implementations. Of course you can write suboptimal program in any language, and it's also true that's easier done in C++ than in python, which is probably why so many on Jow Forums hates that language.
Besides, grabbing a few resources doing some high level juggling is what python excels at. It's moot to argue about python performance when almost all of its most acclaimed packages (numpy, pandas, tensorflow etc) rely heavily on libraries written in C/C++.

If I go for a run and my friend takes me for a ride in his car, why am I going faster than someone who is cycling?