what does /g/ think about computer programmers who only scratch

>>60199983
You can say the same thing at every level of abstraction from machine code

>>60200026
Agreed. And I guess it makes sense that the easiest / outtermost layer of any given development stack will always have the largest audience. I just wish it wasn't this way.

I don't really care desu.

>>60199983
>I personally hardly consider them real programmers
nobody cares what you think

>>60199983

I consider them real programmers

I don't consider them engineers who know about computer architecture or computer scientists who know about math though

>>60199983
It has become easy for people to get by writing programs with high level abstractions. Not knowing how your code is compiled at a high level, no concept of computer architectures, etc.

Code monkeys

OP you will soon realize that there are dozens upon dozens of layers of abstraction. Everything is abstracted. Hell even C lang is many layers above machine code. Maybe in 1974, one could have a near complete understanding of everything from the circuit level up, but it's not 1974 anymore.

What is a REAL programmer anyway? Why is a web developer any less of a programmer than an embedded systems programmer? My advice - stop spending so much time in /dpt/.

>>60200747

This.

This "real programmers vs. web devs" meme is so stupid..

Do you think a "low level" C++ dev writes his own graphical driver and sends the pixel colors one by one? Or does he just use some toolkit like like qt which has several layers of abstraction itself?

And a "low level" C dev that writes his own sockets (which is ridiculously low and unnecessary) doesn't write his own package protocol, but uses the default TCP/IP implementation.

In the end you can't get ANYTHING done if you don't use a big level of abstraction. Maybe it feels like glueing together libraries and frameworks isn't very "manly", but it's the only way of programming professionally.

It's the "midgets on the shoulders of giants" thing..


>>60200678

Math and programming have little in common, since programming is not a science, but a craftsmanship.

That's why so many brilliant physics or math guys are terrible programmers.

I've studied Math myself for a few years, and most scientists don't get how programming work. They think "oh, I understand each part, so the complete thing can be solved easily", but that's how math works, not programming!

In math you try to break something down into parts you know things about: "oh, it's a step function, so it's automatically lebesgue-integrable.."

So when you understood the parts and found the correct "tools" for breaking a problem down, solving it is not an issue. There are no "bugs" in math. Math needs no testing or tooling or deployment or logging like software.

Programming is exactly the other way around:
You can know something like the back of your and theoretically it SHOULD work, but is doesn't. Or you incidentally raised the complexity so much that you can't cope with it anymore.

Good software looks "easy", but it's not easy to break down the problem in the right parts. It's not easy to make software and consider all kinds of future problems. And here it's a lot about experience and not so much about high level science.

>>60201117

> Math and programming have little in common, since programming is not a science, but a craftsmanship.

Maybe the way you are doing it

Anyone can be a programmer. Doesn't mean they're computer/electrical engineers or computer scientists.

File: RMS.jpg (58KB, 850x400px) Image search: [Google]

58KB, 850x400px

>>60201214

Thanks for activating my trap card.

File: conway.jpg (79KB, 850x400px) Image search: [Google]

79KB, 850x400px

>>60201306

Anyone can be a scientist.

Can you calculate?
You are a scientist!

Can you make coffe?
You are a chemist!

Can you write a python program?
You are a programmer!

Not.

>>60199983
I assume they have a different job that they were better suited to, in a similar-ish area. I'd "personally hardly" consider an electrician an electrical engineer, but I sure as fuck know who I'm calling when I need electrical work done on my house.

>>60199983
>web developer
>html, css, javascript
Perhaps you ment web designers?

File: computer_science.gif (2MB, 320x240px) Image search: [Google]

2MB, 320x240px

>Where did CS go wrong?

Probably with people who think computer science is about programming to the metal.

>>60199983
the real world of business doesn't care about whats more fun or makes you feel good about yourself

You are programmer only if you study HoTT 24\7 and prove theorems in Coq.

>>60201117

Isn't that why some people prefer functional programming languages? Imperative programming isn't easy to break down, but functional programming reads like math.

Why is it always the undergrads flinging shit about how they're such hardcore programmers?

>>60199983
I make more as a javascript codemonkey than the CS professors I had in college. Why should I give a shit whether people on an anonymous imageboard that can't sum the first 2 million primes think of me?

>>60199983
>lucrative
there you have it, no consumer wants low level programs, they want to type in "program now" to their (((browser))) and have something ready to go. i.e. agar.io, and any other webapp

>>60199983
It went wrong by being too much effort for too little money. It's much easier to program basic things then actually understand computer science, and it's not that much less profitable.

>>60199983
>Where did CS go wrong?
People like you describe aren't cs graduates as REAL cs courses DON'T teach garbage like that

>>60199983
This thread is dumb.

>>60202619
Yes I had to learn all that on the job. Still better than being a poor PhD student working 80 hours a day on "research".

File: 4ixIXBa.jpg (188KB, 1024x768px) Image search: [Google]

188KB, 1024x768px

>>60199983
Sounds like you've been hanging around /dpt/ too long. The vast majority of internships/entry level dev jobs ARE webdev shit and that's where the industry is heading. Go look up on indeed or angel list, 80% of job entries are either "full stack", front end gigs with only a smattering of iOS, back-end and whatever else is left over.

The barrier for entry is lower than it's ever been, and people are coming in without any kind of college education or job experience, knowing only JavaScript and whatever stack is currently trendy. Can you blame them with all the money being offered?

Meanwhile most of /dpt/ is still living in flyover country saddled with crippling debt, arguing the merits of D, LISP, Haskell or whatever shit language they masturbate over and working for a pittance. Don't believe me? Ask them yourself, do it casually, and the truth will come out. It's pretty sad t b h

>>60199983
what does /g/ think about automotive engineers who only scratch the surface of the machine? e.g. Toyota employees who exclusively work on the Camry, squeezing the alternator in behind the starter. Maybe they to get to design the exhaust for the sports model but not likely.

I personally hardly consider them real engineers, but it confuses me as the only lucrative internship opportunities I can find as a student are at boring ass shitbox companies.

Why doesn't everyone work at Lamborghini and Ferrari?

>mathfags bitch about CS because "not theoretical enuff"
>/g/ayniggers and autists bitch about CS because "not bare metal enuff"

Just do what you want, whether that's working in webdev with Javascript, embedded with C and Assembly, or business software with Java. There's no such thing as "real programming," the only people who give a shit about that are unemployed NEETs on this board trying to fellate their own e-peens.

>>60199983
Just take nand2tetris.

>>60199983

nearly everyone with this view tends to not be very good at programming in any sense

people naturally tend to classify people into groups for the sole purpose of shit talking instead of actually being productive

>>60199983
>What is asm.js
Please kys.

File: 1491775093572.jpg (39KB, 457x361px) Image search: [Google]

39KB, 457x361px

>>60199983
>programmers
>web """developers"""

pick one and only one. Honestly no amount of JS fags with any amount of frameworks and raw processing power available will be able to replace a traditional education founded in actual CS; one that includes discrete math, algorithm design and analysis, systems programming, computer architecture, and networks. If you haven't ever had to program in assembly for any architecture ever, if you don't understand how activation records and call stacks work, if you've never used C, you are not a programmer. You're a script kiddie faggot who doesn't know dick.

>inb4 he mad

Goddamn right I'm mad. I'm sick of all these fuckboi retards calling themselves developers and programmers and coders. You don't know fucking fuck-all about shit. I would take pajeet over you.

>>60199983
Working within your business domain is fine and dandy, kiddo. Jacking off to x86 asm all day isn't going to amount to anything if you don't use it on the job.

>>60204111
>t. unemployed neet writing his 56th iteration of fizzbuzz in x86 assembly

>>60204183

>t. assblasted JS fag

I'm actually still a student but I have been working in the industry for the last 3 years and the sheer amount of mongoloids that can't do anything without "muh npm" is astounding, esp since they get paid the same as me. Perhaps thats why I'm so bitter about it.

>>60199983

Pretty stuff sells. Maybe not to you personally or /g/, but if it wasn't for a frontend dev taking the time to get that app or program pretty for the customer then you wouldn't have a job as a backend developer. It also goes both ways.

Well you're a moron. Who the hell are you to judge when you have zero understanding of higher-level vs. lower-level languages?

File: GettyImages-495486464.jpg (120KB, 1024x683px) Image search: [Google]

120KB, 1024x683px

>>60204183
>I'm still a student

So they get paid the same, you'll be starting your adult life in debt, 4 years behind, and the things you'll have learned will have no real bearing on your career, the only benefit of your degree being a marginally higher starting salary, and the possibility of being recruited through your university's network (I'm assuming you're going to a top tier university). Does this sound like a good deal to you?

>>60204231
>don't hate the player, hate the game

Plenty of embedded devs are dabbling in web dev, why? The pay is far better, the opportunities are far better, and an obsession with "SV culture" keeps H1-B poojeets out of the sector.

If you're making money, who gives a fuck? Not everything has to be a difficulty circlejerk.

>>60201117
Anon your example of a C++ dev is pretty terrible. Yes. You can certainly work at a high level of abstraction in C++. But these are the odd ones out. You normally work in C++ or C because you have engineering concerns like 'will this fit in memory?' or 'will this be fast enough on this platform?' or 'how many watt can I save in this server application?'. When you have these concerns you do have to work at a low level. But not always. There's more often than not some concessions in that stuff CS people call cross cutting concerns (how you pass data from one logically defined system to another). So a rendering engine person might just accept that he's getting his mesh in a specific format, he doesn't get to poke that format to increase efficiency without it being a team effort where he consults other parts of the team to establish a format that's more favorable for all. This is what work is done at high level in C++ or C with these engineers.

People who care about performance do not use Qt. First of all it's a side of programming that's rarely troubled by performance. And secondly it's a generalised system. You limit yourself drastically in the goals we established above if you're working with a generalised system like that. Especially multi-platform systems like qt.

The idea that people don't write their own low level stuff for speed purposes is ridiculous. If you have genuine concerns of this nature you will always do it in some way.

Now. There's a case for people who do not care much about performance to work in lower level languages. It's if they just want the headroom. In C if you write something that isn't all that stupid (which is easier in C because there's not all these features that let you do stupid things quickly) it's very easy to get pretty good performance without caring very much. If it becomes an issue you have the ability to change the characteristics of your and data code much easier than high level languages.
(small cont.)

>>60204378

>debt

only a few grand, easily able to pay it off in my first year, thats what working while going to school does for you. Also, the real actual job experience I have right now is tremendous compared to my peers that I will be competing against when I graduate, and my university itself has a stellar relationship and reputation with companies like Apple and Google. (Although I don't want to work for either of them). SE and CS grads from my uni were the highest demanded of all US uni's in 2016, and not far behind in previous years. Who knows if it will retain that status though.

If only I had a vagina I'd be hired before I even graduated making 120k my first year out.

>things you'll have learned will have no real bearing on your career

what? yeah fucking right, even my job I worked for 3 years proves that to be false.

>does this sound like a good deal to you?

yes. absolutely yes. There is no replacement for a full and comprehensive CS education. Fuck, my former roommate got hired before he graduated and was making 135k his first year including bonuses working for intuit. Some bitch who interned at my work got hired for 175k at apple before she even walked.

>>60199983

No one wants to hire anyone that doesn't have solid CS fundamentals in data structures and algorithms. You still use these concepts even if you are doing what you consider pleb web dev shit. Designing distributed systems that have perform a variety of operations that actually make people money isn't easy just because they are in high level languages.

>>60204468
(cont)
If you have this headroom you can do things like halving your simulation step size to make your system more robust. It's a cheap way to do it. You can opt to use more robust algorithms but this may add more cost than the alternative and still produce the same result.
Having that headroom is often ignored as a benefit in these languages. Because the in-crowd takes pride in making good solutions to the problems without compromise. People outside these low level languages don't get to experience these situations. They don't know how much room they're wasting because they work at a high level. So nobody talks about it. I find C++ programming convenient for this reason. Naive solutions can often be good enough. It saves so much time.
Now C++ has the drawback of you constantly having to check the assembly to ensure that your abstractions don't fuck you over. A missed inline can be catastrophic. If you use the standard library it may be constrained in ways you didn't need it to be (see sqrt(), it's 1 instruction on x64 and more often than not gets to be way more than that because of the constraints put on error checking in the function).
So there's these extra concerns. It's why I tend to use a subset that's closer to C because it limits what issues you may have.
But C++ programmers can certainly be just as frivolous with performance as web devs. They do work in an environment that assists that too.

>>60204500
What did you learn?
What have you used?
Take the ratio of these things.
Figure out how much wasted time college was.
Figure out why you spent so much money on an education that took you that much time yet didn't give you precisely what you wanted.

Imagine if school was like any other service. You ask for the mechanic to fix your engine and they style the car interior and fix the engine. And charge for a third thing you didn't get.
You may certainly be an exception but most CS students I face at my job feel just barely capable beyond a few months of education. What they have is a proven ability to learn. College garauntees that. But they didn't learn what we need them to.

>>60204500
I think your ideas of things aren't very accurate. First of all, assuming we're talking about Bay Area (You mentioned Google, Intuit, Apple). 100k is BASE salary, that's what some yokel that just arrived on a Greyhound bus makes doing entry level webdev work

So when you mention making 135k plus bonuses, that is bog standard compensation. You could easily be making that after working a few of years(leaving and negotiating for a higher salary). Second getting a job at Google is not exactly difficult, I know at least two former coworkers that got that jobs there with less than 3 years of experience. They were good people but not extraordinary gifted or anything. Apple is a bit trickier I give you that, but I guarantee you some pussy pass fuckery was involved especially with all the "diversity" programs in play.

If you still want to go down the University route, I will say this. Try to find a career path that uses what you have learned, look up job growth projections in your chosen field to make sure it's viable in the future. And most of all get used to unfairness. There is an explosion in webdev jobs right now and they are being very well compensated, enough so that many other people like embedded systems devs are switching over. It's not fair I know, but that's just how it is. I assume you want to be rewarded for your efforts and aren't solely interested in academic pursuits.

File: real_programmers.png (83KB, 740x406px) Image search: [Google]

83KB, 740x406px

>>60200747
Unless you're writing in machine instructions, I doubt you're a real programming.

>>60201117
what is linear algebra and discrete mathematics?

Math and programming have many things in common. Math is what makes the code work. Try working with neural networks without having a foundation in linear algebra. you'll struggle immensely

>>60199983

Oh these mother fuckers? I hate em'. They think they know everything all of a sudden about programming. It's annoying.

Anyone else experience this or is it just me?

>>60205618
Gee I didn't know every programmer needed to use neural networks.

>>60205618
NN and 3D graphics don't require a lot of math knowledge to use, just to be good at.
Even a poo\loo (or as I've recently taken to calling it, poo minus loo) could understand that gradient is a direction you go in to maximize a function.

You can do a lot while just treating all that shit as a black box and aping other people.
Sure, not developing anything new, but with modern libraries, even a woman could plug in data and get sensible results. That is, provided someone installed all the libraries for her.

>>60205853
>"Math and programming have little in common"

I don't agree. You do not need to be programming neural networks but you cannot deny that math and comp sci aren't intertwined.

>>60206570
>comp sci
>programming
One is engineering the other should really be called computational mathematics.
https://en.wikipedia.org/wiki/Computational_mathematics
See if this doesn't correspond much better to the idea of computer science the way it is taught than anything else.
Computer science is just wrong.
This lecture is very informational on why computer science is a misnomer and draws a clever analogy to geometry which has the exact same problem (though we don't think about it).
https://www.youtube.com/watch?v=2Op3QLzMgSY

Confusing programming and computer science is just as bad.

>>60206674
I agree the two are different things. But imo there is significant overlap depending on what you are programming

>>60205618
>liniar algebra
>math

We all use a bit of math in programming but its not like were doing crazy calculations of some new equation we made up, i think thts wht op ment

>>60206719
>depending on what you are programming
Yes if you're working with a very mathematical field you certainly will need the domain specific knowledge to do your programming job.

But you would probably not label programming as related to sewage treatment or mechanics just because you're programming a sewage treatment plant. That's confusing what programming is. It's a craft just like electronics or woodworking. Your goal determines a set of concerns you have. But that doesn't make it core to the subject.

Universities love to confuse the two where they shouldn't.

>>60205618
>>60206570
>>60206674
ANNs have made massive progress in recent years in large part because researchers have been willing to try things experimentally long before they are understood as mathematical models. asynchronicity is a good, concrete example of this. the mathematical models come afterwards, sometimes many years afterwards, after the result is shown empirically

another example: the entire reason that AlphaGo was impressive is that many people had tried to build a working Go AI and failed, so we all knew the problem was hard, and it's unexpected that AlphaGo can be shown, empirically, to beat 9 dan professionals. but there was no a priori "proof" that AlphaGo should be able to do this, that was an experimental result

>>60199983
web dev is programming and there is literally nothing wrong with it. i have no personal interest

for whatever reason some CS people are allergic to human-computer interaction even though it's a very obvious need with large economic value. if you are one of those people then you shouldn't be a "frontend" programmer. but don't denigrate there work, it should be clear to you why there's demand

>>60199983
Depends on the job

>>60206903
>for whatever reason some CS people are allergic to human-computer interaction even though it's a very obvious need with large economic value.
It's generally not computer scientists who view frontend developer work to be not programming. It's usually programmers who make the distinction.

>>60199983
>exclusively use html,
Last I checked they didn't know html, javascript does it for them.
>basic bitch web dev. companies.
Just make sure they don't force web2.0 dogshit on you. web1.0 is fine: have minimum markup on client and do everything on server.

>>60206951
>do everything on server.
That doesn't make any sense. People have fast computers now. It makes much more sense to let the server handle tasks that it's good at and let the clients carry some of the burden.

File: 1483981345432.gif (438KB, 414x532px) Image search: [Google]

438KB, 414x532px

>>60204934
> Asm babby can't into VLSI

>>60199983
It comes down speed of development, the lack of education to the general public and the lack of Complexity the public want to handle.
Web devs don't know anything about the underlying machine, but don't have to nor need to. And they can pump out projects month after month using their millions of frameworks.

The Same goes for music. Rock is a genre full of talentless hacks who know little about the underlying theories of music and their music ends up being really simple, repetitive, boring and unimaginative.

Abstraction effects all crafts. Just deal with it and except you'll never be the most popular, just an essential part of the machine.

>>60206903
>some CS people are allergic to human-computer interaction even though it's a very obvious need with large economic value

is front end really human-computer interaction?
I don't think being front-end is equivalent to acquiring requirements from users/clients, or designing an aesthetically pleasing GUI, or planning on usability, this is the real user-computer interface. Programming it can be done by any code-monkey.

>>60207219
>Rock is a genre full of talentless hacks who know little about the underlying theories of music and their music ends up being really simple, repetitive, boring and unimaginative.

here's something ironic, pop is boring unimaginative simple and repetitive, but there's a fucking lot of effort put towards producing those hit songs by ppl who know lots about music theory, pretty amusing.

> tfw listen to experimental music

File: Curriculum 68 Recommendations for academic programs in computer science a report of the ACM curriculum committee on computer science.png (156KB, 887x1128px) Image search: [Google]

156KB, 887x1128px

>>60199983
>Where did CS go wrong?

When it stopped being this.

>>60207299
>linear algebra required for analisys 2
Fuck yeah.
>OR
fuck no.

>>60207392

How else will you do analysis on manifolds?

>>60199983
> web developers who exclusively use html, css, javascript and maybe a little bit of a backend language (php, java, etc.) but not likely.
It's OK, somebody have to deal with frontend.

>>60207412
I don't know. They decided to get it all confused by making us learn some linear algebra in Analysis. It was horrible. Got through it though. But it's dumb to rely on self study when you can simply reorder the courses. The next course I took was linear algebra so that was all really pointless.

>>60207275
lol.
I had no idea.
https://en.m.wikipedia.org/wiki/Pop_music_automation

>>60201214
>>60205618
It's like you didn't read his post fully. Of course you need maths for a lot of programming problems and for most of the interesting ones. But his main point was that in programming, especially designing large systems, managing the complexity and making choices about how to subdivide problems is critical. A lot of people from maths or even competitive algorithmic programming don't really understand this at first. They feel that if they understand all the details, then it shouldn't be a problem to solve the larger problem. This often leads to spaghetti code that one can get lost in. And indeed when you look at code that is written by scientists, it's very often just one long blob of code without any structure whatsoever, repeating code, etc.

No need to drag the op into the whole "do you need maths for programming" debate. That is old, annoying, unproductive.

>>60199983
people have always done the bare minimum