Computer Science == meme

What is it about computer science that makes it a meme?
Why should/shouldn't it be classified under a "science" term?
is it more like an engineering field?

As a side note, a lot of brainlets in school thought making a powerpoint or being good at using windows meant they are destined to be an expert at computers and then some meme teacher tells them to do computer science and they ultimately fail heavily. Hence why at many low-tier universities comp sci degrees have a high failure rate.

Other urls found in this thread:

ideone.com/c697Lt
blog.computationalcomplexity.org/2003/03/theory-and-theory-b.html
cstheory.stackexchange.com/questions/1521/origins-and-applications-of-theory-a-vs-theory-b
money.cnn.com/2017/04/03/technology/h1b-visa-fraud/
twitter.com/NSFWRedditVideo

>Why should/shouldn't it be classified under a "science" term?
because it doesn't follow the scientific method

Comp Science is at it's very core and pure level a better science than other fields. It has better fundamental questions than in any other field.

So you're implying that every field that doesn't utilise the traditional scientific method in every single problem isn't science?
I would disagree that computer science doesn't utilise the scientific method.

>I would disagree that computer science doesn't utilise the scientific method.
go on...

>What is it about computer science that makes it a meme?
People that think web development is CS.

== is a meme.

remember when := meant assignment, and we didn't need == ?

What bothers me is the linguistic act of borrowing bits of formal language and integrating them into natural language in hopes of gaining an air of authority. It's clumsy and pompous.

Back in the day people thought it was a new kind of science, only later we realized it is just a extension of mathematics, but the name was kept.

...

Yeah, that's what science is. What part of CS uses the scientific method in any way?

why you keep posting this

MechE major here

Here's my code from a connect 4 project. I've bee told I should consider switching majors by my professor but know damn well this code sucks, although the game works 100%.

ideone.com/c697Lt

bump

The real question here, is... Why do you care?

Are you here to be productive and ask serious questions, or are you here to make fun of a major that doesn't follow your preconceived notions of convention?

>ideone.com/c697Lt

I'm laughing so fucking hard, is this for real, or its just trolling?

its just where I store my code

fucking /g/ YLYL material

>1600 lines of garbage is still garbage
>Could have been done in less than 500
>Talk so easily about doing CS

You're being too obvious when it comes to trolling.

NIGGA, YOU CAN DO THAT UNDER 50, TOP FUCKING KEK

No, I'm not a compsci fag. The CS professor suggested I should consider CS as a major because of that program. I said I hate this programming class and the program is shit

wew lads

it's a branch of mathematics. it's STEM for sure. whether or not you want to call it "science" is more of a language/semantics issue. in non-english languages "science" translates a bit like "the art of knowledge" and there is no issue there. but i can understand why native english speakers and especially americans don't want to call it a science.

tl;dr it doesn't matter it's just jealous brainlet spergs nitpicking about the meaning of a word

CS major here. I wrote an entire discrete-event network simulator in less code than this. This is terrible.

Did you write a script to generate that code? Be honest user because otherwise that's a lot of effort to get (You)s

Theory B is pretty comfy and has strong ties with mathematical logic

What the fuck is theory B

It's a subset of mathematics when done right.

I know it is, its why I respect you CS guys so much. I hate CS, I went into MechE to stay away from anything programming related.

>It has better fundamental questions than in any other field.
Like what?

No I wrote that from scratch.

lots of ctrl+c/ctrl+v

It's expected if you don't program on a regular basis, but I'm wondering what in the world is wrong with your CS professor that he thought your spaghetti indicated promise in the field.

Do you do particularly well with algorithmic problem solving? My desperate hope is that at least your educators have an idea of what it takes for somebody to show promise as a computer scientist.

>1700 lines
Wew lad.

go away /pol/

It's more of a language syntax exploration than actual science.

Assembling word puzzles basically.

She told me for someone who has never had any programming experience, the fact that my program has absolutely no bugs impressed her.

The reason that program sucks is that I fucking suck at functions. C sucks, I think I'd have a much better time in a more developed language desu

C is fine. I think you just needed to consolidate your logic a bit. A lot of your current logic is redundant.

>C sucks
It has nothing to do with the language and everything to do with you being absolute trash at it

Yes, that is is exactly what it means to be science

One could also argue that math in general is a task in puzzle solving. I disagree with your statement that computer science is simply 'syntax exploration.' How would such a definition apply to the computational study of cryptography (as in the field of provable security)? Computational geometry? Computational complexity? People on Veeky Forums like to argue that these don't truly belong to the field of computer science for one inane reason or another, but the fact of the matter is that these are all genuine sub-fields and areas of concentration within computer science.

I know I could've used functions to reduce the redundancy, but that's my weakness in the course and I was on a time crunch. Really, some extra for loops could've reduced the length significantly, but every minute of that code was painful to write.

I agree that I'm trash at programming, its why I never wanted to program and why I want to pass this class with a B and never look back.

== is logical equality you literal retard

'Theory A' is computational complexity and analysis of algorithms. 'Theory B' is stuff like programming language theory, logic, type theory, etc.

blog.computationalcomplexity.org/2003/03/theory-and-theory-b.html
cstheory.stackexchange.com/questions/1521/origins-and-applications-of-theory-a-vs-theory-b

In American CS departments, A is more common. In Europe there's a mix of both.

Computer Science isn't a meme necessarily, but unfortunately most Computer Science programs are more along the lines of an automotive technician programs, practical / lab oriented courses with minimal theory involved. These programs are designed to churn out vaguely competent programmers, not the next generation of computer scientists who are going to make the newest fastest algorithm or prove P=NP (although there will be slapsticky arguments over those kinds of topics, they won't amount to anything).

Frankly, I see programming becoming a trade field in 10-20 years, along with welding, automotive technology, plumbing and electricity. This will probably siphon off most of the brainlets (not talking shit, just being honest) and leave Computer Science as an odd, but scientifically rigorous field with a paper being published every now and then by a few passionate professors.

I mean, there's Chemistry and Physics, as legitimate sciences, and then there's Chemical Engineering and Mechanical Engineering, and then there's the automotive technician certificate programs. Computer Science fields will inevitably fall into a similar hierarchy:
>High level theory work, experiments, data collection and statistical interpretation
>Theory-dependent practical application work with a high level of precision required to solve mid-level problems
>"Everyman" job of using the engineers work to implement solutions in everyday situations (average pajeet programmer, "technician" type job)

I mean, math is a computing language on its own with specific syntax, procedures and results. Really, Computer science is more of a field of mathematics and than a science.

Your first programming class in C may be a bit jarring, but the sort of command over programming methodologies you would need to implement Connect 4 in a reasonable manner isn't anything to write home about; it's not really something that should be beyond the reach of a novice who has little practice.

Are you actually following along with the units in your class? Did you attempt to put in the time to figure out how to manipulate functions in C? You shouldn't be so defeatist about things. If you can't put in the time to learn simple things like C syntax and design patterns, how far do you really believe you will make it in your own field?

Nice effort (Y)

Veeky Forums is a cesspool of retards. The opinions here don't matter

Interesting, I've never heard of the topics being divided like that in the states

That would normally seem to make sense, but you lack the perspective of your average room temperature IQ guy, whom the basic logic of loops and if statements frustrates.

You all need to understand as elementary as calculus seems to you all, 80% of Americans will never do it. Computer logic is similar. You spend your life around those of similar intellegence so you conclude anyone could do it.

I like my field of study, I'm a junior at my uni. I've thoroughly enjoyed all my classes so far except compsci. I just don't like programming. I don't have the patience for it. I've put in the effort for functions, but I struggle with implementing them and debugging them so much that I avoid them as much as I can.

I kind of enjoy writting programs to solve math related problems, but compsci is just not my thing.

>That would normally seem to make sense, but you lack the perspective of your average room temperature IQ guy, whom the basic logic of loops and if statements frustrates.
Possibly true. During one of the lower periods in my life, I was studying to become an automotive technician, and christ above, the guys you encounter there are just not that bright. Nice, hardworking guys - they go to church, they keep their head down and all, but the mental capabilities just don't quite seem there for a lot of the abstract thinking stuff.

They all get to pass the course because every course had the majority of its points based on the "labs," which were just common technician procedures. You could not do any of the tests or quizzes and still pass with a C (although you still had to do the written homework.)

I think some of them could hack it. Auto diagnostic flowcharts are easy enough to grasp, so I'd imagine most of them if they put the work in could get past if / else and while loops. For loops are a little tricky, but they're essentially extended while loops.

So I disagree, I think programming is for total brainlets and any moron could be taught to be a pajeet level java programmer in 2-3 years.

>in non-english languages "science" translates a bit like "the art of knowledge" and there is no issue there. but i can understand why native english speakers and especially americans don't want to call it a science.
German here. For "science" we say "Wissenschaft" (literally "knowledge shaft"). We do consider Computer Science a sceicen, but we distuingish it from Physics, Chemistry and Biology, which are Naturwissenschaften (natural sciences, or "nature knowledge shaft").

plow me with your knowledge shaft omg

Programming is already a trade field in Germany. Not even kidding. But people still study Computer Science because it offers you more opportunities, you can get into more fields besides just programming (computer engineering and computer science are usually a combined degree here), and you earn more money, you are more intelligent, etc.
It's basically like being an electrician by trade vs studying Electrical Engineering. Or being a dentist's assistant vs being a dentist.

No wonder I see so many Germans doing low-level stuff

this

a pajeet is almost worthless. the only reason companies will compromise on competence is because there aren't enough competent programmers to go around, because programming is hard. programming beyond the most basic tasks is something the average person simply cannot do. if programming were to get "dumbed down" to the point where anyone can do it, we would have invented general artificial intelligence and AI + automation/robotics would render almost ALL human professions obsolete.

You can get a programming job without a degree. It does really demand a lot if math

>is it more like an engineering field?

No it's more like mathematics. Things are built up from axioms and studied formally. There isn't really experiments/hypothesis/etc like there are in empirical sciences (besides maybe in certain subfields of CS like machine learning)

>I think some of them could hack it. Auto diagnostic flowcharts are easy enough to grasp, so I'd imagine most of them if they put the work in could get past if / else and while loops. For loops are a little tricky, but they're essentially extended while loops.

Yeah but that is following a flow-chart vs taking some specifications and turning it into an algorithm (flow-chart) yourself.

>Engineering
CS is not software design.

CS is about the study of computation.

You're right, It uses mathematical proofs. Which is superior to empirical science in every single way.

Insecurities about his own choice of major and inability to find a job.

It should be called "computational mathematics" in my opinion.

cs more like CShit

Looks rather comparable to me.

>Which is superior to empirical science in every single way.
Why do you think so?

I agree that comp Sci at its core is a science. The stigma comes from the fact that the field relies heavily on math but most programs only go to Calc 4. Some newer programs don't even require Calc 1. Without proper logic and analysis, comp Sci is reduced to nothing more than vocational training for monkeys.

Wow, the amount of crap you get is unbelievable. The fact is that if you had a problem and knew how to solve it by any means necessary, you're a winner. At school you're not usually given many tools (at a classroom teaching standpoint) to work with so that kind of code is expected. You know how the game works and how to set it up. I think you learned from your project and you definitely should continue making programs, as you're not afraid to work harder if you come across an obstacle. When you learn how to execute that with functions defined by you alone, you're looking at a much more efficient code. Hell, the reason why I like this code so much is because it's EASY to follow. I know exactly what you're doing and you have sufficient documentation, which is a good practice. I wouldn't switch majors though, stick with ME and use your problem solving skills there.

My uni is #5 in CS program in the US and includes a lot of rigorous theory-heavy classes. Would Veeky Forums still consider that a meme?

CS is a sub field of mathematics. Whenever sci says computer science they mean web dev, game dev and other areas that aren't "pure" CS theory.

Pure CS theory is just a sub area of mathematics. Most CS majors stay away from that

>Drawing numbers in circles n drawing arrows
>science

Why is CS the most pretentious of all the sciences?

I think this has to do with Germany's big automotive industry.

CS != Software engineering

I'm CS, I see the core of computer science as not being computer related:
>A bit of maths
>Advanced management skills
>Advanced deduction skills
>Logical thinking
>Ability to consume A LOT of technical documentation.

Likewise - I design circuits now and then for work, and I find my proffessional buying skills are more usesful than my basic (hobbyist) Electrical Engineering skills.

Why do you hate functions?

>Why should/shouldn't it be classified under a "science" term?
>is it more like an engineering field?

I agree. I think Computer Science would be better thought of as an engineering discipline.

> Hence why at many low-tier universities comp sci degrees have a high failure rate.

That's because most people have no clue what Computer Science is. People use Powerpoint or Excel and they think that's what goes on in a Computer Science classroom. There is a huge, huge amount of confusion out there.

>the next generation of computer scientists who are going to make the newest fastest algorithm or prove P=NP

We have very, very little need for this -- relatively speaking.

For every computer scientist who works on pure theory, we need 20 others doing development, or doing research that directly supports that development. And, sorry theorists, but trying to prove that P=NP is not the kind of research that is going to be "supporting" any development projects.

HOLY FUCKING WHAT?! WHAT ARE YOU DOING?! WHY WOULD ANYONE DO THIS?!

i have a feeling that portions of this are going to be featured in those cs graduate memes pretty soon

I believe the other guys lost perspective. Yes, you could have done it shorter, cleaner and better. Perhaps you could have done a must robust modelling, but the code is clean and clearly you cleaned the code and debugged it.

I think it is above average for a standard non-comp sci project, you could do a lot more if you expand your knowledge in a systematic way . If you enjoyed it I would suggest at least getting a few comp sci classes.

Knowing how to code (and do it well) can do a lot for you no matter what you do.

Unfortunately, a lot of places teach it as just coding.

as long as it works it's not bad for a newb but it's nowhere near what would be expected from a professional programmer

>""""""professional"""""" programmer

money.cnn.com/2017/04/03/technology/h1b-visa-fraud/

computer science functions under layers and layers of abstractions, and uses very little physical, chemical, or quantum sciences in it. It is more akin to computer engineering, but without the engineering degree.

>It is more akin to computer engineering
Only if you don't know what computer engineering and computer science are.

I agree. But learning should always encouraged and the shape of his code suggests that he is interested enough in logic, structure or programming in general to get a few courses.

But he definitely must get better.

>being this new

Thank you.

Thank you, but this is the only compsci course required for me to get a MechE degree. I'd rather deal with thermodynamics, dynamics and material science than compsci.

Yes I suck at this but its the last time I have to deal with it besides Matlab

Please provide an example wherein computer science uses the scientific method. I'm genuinely curious.

Stop it. It's not funny

Proposing a algorithm to find a approximation for a multi-armed bandit

>these are the people talking shit about CS majors on Veeky Forums
k

>Software Design process

Isn't that more SE than CS?

>In American CS departments, A is more common

I dunno about that. Seems like at an undergraduate level most schools are just turning CS into Software Engineering with 1 or 2 mandatory algorithms class.

As a dude in the field, I think it should be called "Computational Mathematics" instead.

It's just a matter of phrasing, at least for anything applied. Observation "it might be possible to write a better algorithm for this problem", hypothesis/prediction "this algorithm X will perform better than existing algorithm Y", experiment by running them both, the algorithm becomes the new theory, new observation "we might be able to improve algorithm X by doing Z".

>#5 in CS program in the US
what uni user?

>Pure CS theory is just a sub area of mathematics. Most CS majors stay away from that
Yep, because it's pretty dull and very un-sexy. The only advantage is not having to keep up with new programming languages or concepts.