Julia

Pajeet names his programming language with a white girl's name.

Pathetic.

Attached: 8b3977e9fabd849c0d9e8a8bb923945bf983d073.png (1260x852, 36K)

Turnabout is fair play, right? Start your own language, name it Pajeet.

Julia is a shit language. It's literally just a shitty JITted wrapper for other code. The Julia "package manager" just clones shit off Github, the computations are handled by FORTRAN code (BLAS), and 90% of everything else is done by importing a Python library. There is literally no good reason to use Julia.

Guys what can be done about the pajeet problem? I'm sick of them. Why won't they just go away?

their automated forward differentials are pretty neat though. And you're quicker than writing FORTRAN code, while having similar-ish perfomance.

Why would I write FORTRAN when I can do Julia?

Not to mention the fake benchmarks (i.e. benchmarking tuned Julia code against shit code from other languages then claiming that tuned code is not allowed for them because it's "not idiomatic" or some other bs)

At least it's not food stuff.

>Julia is a shit language
data scientists and several businesses disagree

>wraps other code
oversimplification

>pacman clones githubs
so does Arch pacman, pip, and most community-focused pacmans.

>computation is handled by fortran
as it should be

>90% of everything else is done with python
????

>there is no reason to use it
1. Its great for scientific computing, because it
2. combines the simplicity of python scripting with the speed of c-compiled binaries, and
3. handles everything in a FP way that does not
4. impede on productivity.

>inb4 - but muh pooscript

I like Julia for scientific work. I use it as my main algorithm prototyping language. In that domain, I've only found a few issues (sometimes it optimizes poorly, and PyCall.jl is clumsy when it's necessary). Outside that niche, I wouldn't use it at all.

this. Also: multiple dynamic dispatch.