Does sci respect programmers?

>>16707245 (OP)
I do not respect any employed “people”

>>16707254
Scientists are employed, many of them are programmers, even those who are involved in fields beyond computer science. OP, why do you seek approval from this kind of person?

>>16707245 (OP)
Most branches of STEM have a computational spin to them: cheminformatics, computational biophysics, bioinformatics, theoretical and quantum chemistry. Most of the time those programmers aren't very good because they prioritize science of coding. Lack of documentation, a lot of hardcoding and use of only the most basic functions of the language, low performance. There are gems and competent scientist programmers, but the default is:
> PhD in bioinformatics
> Researcher at top 1 university in the country
> Writes a Perl script that parses Sequence Alignment File using a 120 character long regex
> it takes 1h to parse 1.5GB of data
For comparison, I wrote a python script that does the same thing but finishes under 3 mins.

A different angle is that pure programmers often make very bad scientific programmers. They lack necessary domain knowledge, make errors that you wouldn't expect, and write high performance code that you don't understand and can't verify whether it does what is scientifically valid.

i do. engineers have a level of mechanical sympathy that seems supernatural to me.

>>16707260

Computer science is essentially applied mathematics. We need it to build supercomputers to do important research, but it also develops technology itself. A commercial software developer is like a technician. The research domain is funded by universities, government labs, tech companies.

If you ask strangers on the internet what they "respect" you could be talking to time cube or flat Earth people. Some things are simply important regardless of their online popularity or publicity.

Programming as a skill or activity is not inherently a scientific endeavor. Conducting research in computer science is, and this involves high level math beyond what you typically see as a normal software engineer.

Take a look at some of the most difficult algorithm problems demanded by the top tech employers and you will see that it does not require the application of high level math. Not even calculus. Like 20 lines of code from a memorized algorithm.

Read some recently published papers on machine learning models if you want to see the cutting edge of computer science. It requires the application of high level math and is a fascinating subject.

>>16707245 (OP)
I do not respect anyone who writes C or C++.

>>16707268
This is a pretty narrow view desu.
>Read some recently published papers on machine learning models if you want to see the cutting edge of computer science
They are not even remotely close to the cutting edge of computer science. This field is stuck in the 1980s and thanks to NNs finally becoming large enough to encode a useful amount of information, it's all been reduced to basic statistics, integration and matrix transformations.
The real cutting edge of computer science is basically entirely centered in type theory, graph theory, computability, theorem provers, inference, and of course non-classical logics. Quantum computing is obviously the big one, but paraconsistent models of computation are catching on more and more.

>>16707245 (OP)
Only Hardware, Tool and robot programmers.

I am a researcher, i programme alot, i am not a computer scientist. Whenever we get to work with Students and they submit their bullshit in Python running some backend fortran library i outright puke black goo

About 90% of our students fail because they too stupid for lowlevel programming and developing algorithms from scratch without a library backend

>>16707292
this is why I code in c and NOT c++

>>16707275
it honestly makes me feel depressed knowing that we will never truly have terminator level AI intelligence in the next 100 years

>>16707245 (OP)
absolutely. they're wise and important people.

>>16707300
Wasn't this guy just a low-level QA dev at Blizzard who only got the job because of his dad's connections?

Bad science programming is harmful.

>>16707245 (OP)
i was one of the few people that had a physics AND cs background when i went to grad school for ee.
everyone knows how to program something in ee, but generally everyone is a dogshit programmer that hasn't cut their teeth on proper software engineering yet and are surprisingly resistant to learning anything other than what they first learned as a baby duck. granted, the nature of research is a moving target, but there still are ways of going about things that can save everyone a lot of pain that no one is aware of or adopts in the field

>>16707346
Amazing

>>16707245 (OP)
>He doesnt use ms notepad to write hit C++ code

>>16707332
well yes, but he also made some spying scripts and sold gay furry sex stuff in second life.

>>16707300
do you respect c or c++ programmers more?

I don't respect them in the post-vibecoding age.

>>16707380
not anon, but C programmers look like pic related

>>16707382
we are not all using chatgpt to code lol

>>16707354
>He doesn't write assembly in vi
NGMI

>>16707300
qrd? is this Grummz?

>>16707268
>computer science is essentially applied mathematics
do you mean actual computer /sci/ence or webdev slop?

>>16707245 (OP)
computer science is the only actual science nowadays

>>16707275

I am only providing a single example, and it is a valid example. The 80s was only the start of this research.

https://arxiv.org/pdf/1511.06434

Am not saying those fields are not wortwhile, am just providing an example. Neural networks are my field of interest and papers like these were published in the last 20 years. Statistics, integration, and matrix transformations are the basic building blocks of neural networks. That is the basic model from the 80s. Newly published models take this concept and expand on it in more complex ways.

A DCGAN is a variant of a GAN that uses a Convolutional Neural Network (CNN) to in both the discriminator and generator to generate more accurate results and rely on less training data. In other words, it is an improvement on efficiency. Generative Adversarial Networks (GANs) themselves combine two neural networks that work on those fundamental concepts as you describe them to generate realistic imitations of training data fed to the discriminator.

GANs were invented in 2014. New research is being published all the time to develop new models and improve existing ones. You understand the basics of it but the current research is in higher orders of abstraction and complexity. I find it is an exciting field as it leads to technological development and is much more challenging than most commercial software development.

>>16707419

Actual science. Basic web development is not "slop" but once you master the fundamentals it stops being intellectually challenging. That is what drew me to the scientific side of programming. You actually get to work with higher level math in interesting ways.

Neuroevolution is very fascinating and exciting. We can build neural networks that think, learn, and evolve like an organic living being. Some people are superstitious about this but it is really just multiplying matrices at the fundamental level.

Just like how everything that a computer does can be reduced to flipping 1s and 0s.

>>16707485

My CPU takes hours to train really basic models. Using GPU acceleration it us substantially faster but still limited. My interest in DCGANs was a project to generate high resolution images of galaxies from space telescope astronomical data.

I wish I had access to a supercomputer or one of these trillion dollar GPU datacenters. The computing power required for this is massive, matrix calculus in the thousands of dimensions. My desktop PC simply does not have the processing power.

>>16707406
>He doesn't manually short circuit the transistors in the CPU to make the computer do things
NGMI

>>16707494
>He doesn't carve his own computers out of wood
NGMI

>>16707274
Anon, that is how you end in a list and some files "magically" appear in your machine with metadata faking a typical predator activity. Who is going to believe you are innocent? In no time you will become an statistic of prison violence.

>>16707495
>He doesn't just do all the calculations and computations in his head
NGMI

science laws should be translated to programming languages, only real science is simulations, vid related

>>16707346
The messy code part got me. Anyone who can’t write proper clean code should be immediately shot in the head and strung by the feet to the nearest overpass for all to see.
t. Software developer

>>16707245 (OP)
I think that most people here don't respect programmers because most of them just do web related stuff which doesn't involve any kind of hard-ish problem to solve. But if you don't stop there, you will notice a lot of programmers doing cool, math/science based stuff. Things like, physical simulations, DSP, crypto, bioinformatics, etc.
Unfortunately the former group has been louder than the latter for about 15 years..

>>16708020
Do you think that FPGA programming can also be cool?

>>16708436
who the fuck cares what some anon thinks on 4gay
and yes, fpga is neato

>>16707409
He's a supergrifter that derailed a campaign for better consumer rights with regards to digital ownership. Grummz is just a regular, mostly harmless, grifter.

>>16708436
yeah, especially in aerospace and defense or HFT

>>16707492
This is the sad reality of trying to do any bleeding edge research. It's been true since the days of mainframe computers. 80 years from now people will be looking at the modern desktops/supercomputers of our time and marvel at what we were able to do with such limited compute power.

>>16708020
yeah I think programming is a good end game for math autists that actually pays if your not Terance Tao incarnate

>>16708605
Yeah, you either become a programmer or an high school math teacher and, honestly, I'd prefer to be a farmer than teaching a bunch of teens that x+x is not equal to x^2

I write all my code in microsoft excel then convert them to latex tables with comments in the footnotes and of course a references section where i have all my stackoverflow searches and AI conversations.
Then I build and ship the pdf over email directly to my boss.

>>16707254
you are nobody and your life has 0 significance for science or mathematics.

We do a little (modern) APL

>>16708436
I want into this. To me it is one of the most interesting fields

Yes

>>16707274
got curious about OP image and found the codebase

>//>===================================================>===========
>// Copyright (C) Intel Corporation
>//
>// SPDX-License-Identifier: MIT
>// >=============================================================
>
>// This is a simple DPC++ program that accompanies the Getting Started
>// Guide of the debugger. The kernel does not compute anything
>// particularly interesting; it is designed to illustrate the most
>// essential features of the debugger when the target device is CPU or
>// GPU.
no idea what this is, but..cool!

https://github.com/oneapi-src/oneAPI-samples/blob/main/Tools/ApplicationDebugger/array-transform/src/array-transform.cpp

>>16707274
ok tranny well, rust is dead in 2025, maybe you guys can go hang out with the survivors of the Haskell extinction of 2015

I respect a sci who can program

>>16707267
>mechanical sympathy
are you from /lit/?

>>16707245 (OP)
No