OOPfags BTFO

Slightly inaccurate, you'd need an exception factory so you have a well structured way to shoot yourself in the foot.

>>56539985

I once wanted to make modifications to a large program OOP written in style, after some time I rewrote the whole thing in pure C. I guess OOP is good for databases and some other things

>>56541392
OOP is the only practical way to break down the programming to multiple plug-and-play pieces guaranteed to work together.

It's great for industrial-scale programming and very complex applications.

>>56541467

I agree fampiño

So, in a multi-paradigm language, should you forego the oop features altogether?

>>56539985
this is more like c# vs java

>>56540166
Underrated post

>>56539985
OOP is not bad, as long as you only use the features you need and don't go full enterprise pajeet mode.

>>56539985
Another anon said it's just like communism, great on paper horrible in practice.

OOP presents no problems whatsoever to excellent programmers. Unfortunately, the amount of (living) excellent programmers is rapidly approaching zero.

>>56544841

"Good programmers" don't exist.

>>56541815

No, you should use them or use a language which can exploit optimization caused by their absence.

>>56539985

Is it really fair to map out all interactions in a system? Obviously the animals on the left don't exist in a vacuum and therefore look like the right side in practice

>>56541467
>hey, anon, how about you add a feature XYZ which fucks the whole design? How many hours will it take you?

https://www.youtube.com/watch?v=QM1iUe6IofM

>>56539985
>leggable
>fleable
top kek

>>56544954

>Are you sure that feature is something you want? There are severe time and manpower tradeoffs that come with implementing it presently.

>Because the current system relies on the expectation that-

>[good explanation why your system made sense based on the requirements when you made it goes here, assuming you're bright and hadn't coded from the start a shit tier system and are now waffling in the face of work]

>So while an initial implementation will probably take

> (time it took to code all features contingent on the design * 2/3 because you have working sample code)

> You'll also likely be tying up the engineers working on other features, test, and QA to get back to the state we currently are in, rather than progressing.

>Knowing this, do you still want it?

Quit being a bitch who gets walked on and learn to talk to management.

Why are these threads still posted? I would think it would have gotten hashed out one way or the other, years ago.

What functional programming shits claim
>well I was here first so I'm better than u. muh 2k of memory. muh 80s.

What is actually happening
>Posting from lunch break at non-computer job due to unemployability

File: 1470804356900.jpg (256KB, 1280x960px) Image search: [Google]

256KB, 1280x960px

Oh, it's this.Thread() again.

>>56545249
>Posting from lunch break at non-computer job due to unemployability
You should've seen the face of my college profesor when he saw me working at a convenience store. I had top grades in my group, almost the best ones in my year. 7 year long gap in my CV after college due to severe depression made me unemployable.

>>56545271
AbstractThreadFactoryFactory() = TRUE

haha!!

>>56541467
Inheritance sucks in 90% of all cases and you should use something else. The rest of OOP is okayish.

>>56545271
who is this semen demon

>>56539985
I feel like nowadays OOP just means "couple data and functionality using mutable state". Any other feature of OOP isn't unique to OOP.

File: 1431562160572.png (640KB, 2900x950px) Image search: [Google]

640KB, 2900x950px

>implying i don't have the ebin counter-repost at ready

>>56547152
>Hey, we need to make a half-plant/half-bird hybrid for our game, that should be really easy, right?

File: confused europeon.png (209KB, 258x298px) Image search: [Google]

209KB, 258x298px

why do GNU/NEETs get triggered by OOP?

>>56547175
Because it is not their structure/way, and their structure/way is the best because it is theirs. They are lacking the social knowhow that having standards allows Interoperability, it may not be the best, but it is good enough.

OOP was created with a wonderful purpose, to approximate software abstractions to the real world. Then the enterprise saw it as a way to avoid the mess that plain structured programmers were making while writing giant monolithic software. For instance, it is really easy to write a giant piece of software inside a single .c file, however Java will bitch about it a lot, that way the code have at least some modularity, which is required for code maintenance.
The next paradigm to be fucked around by enterprise is functional programming, since all this existing massive software will need to be rewritten in a simpler way, the functional way.

>>56547175
Because they never worked on anything more complex than fizzbuzz.

>>56547068
Asuka Langley Sohryu

>>56548014
Eventually they'll move on to pure functional. I'm interested to see what bastardized monstrosity they make out of it.

Please tell me shortly what is functional programming and why is it better than OOP.

OOP IN LOO

>>56548391
Problem with these arguments is that no one agrees on any definition. 10-20 Years ago, LISP was the poster child of functional languages, and nowadays it's debated whether it's even functional at all.

Current definition seems to lean more towards pure functional programming which is characterized by explicit state/side effects.

Procedural programming was considered bad practice due to the pervasive usage of state. OOP improved the situation by hiding state. Functional goes further to improve the situation by either

>entirely eliminating state in large portions of programs
>making state explicit in the portions that have it

It also turns out functions compose much more easily than objects. If you have a multiple objects, you can't just add them together in a one-liner to get new functionality, but you can do that with functions.

As a side note, the biggest difference is the culture. Modern OOP languages can do things functionally; even C# has syntax for monads (from ... select ...) but since most libraries don't use this style, you tend to stick with OOP techniques.

>>56547175
>>56548029
this
they're just a bunch of fat NEETs. They never worked on big projects think that software architecture is a meme.

>>56548391
>>56548534
One more thing. There are a number of features that have nothing to do with functional vs. OOP but, due to culture, seem to only be available in functional languages.

Examples:
>algebraic data types/sum types/discriminated unions
>pattern matching
>extensive type inference
>dependent typing

>>56539985
I rarely use more than 1-2 layers of inheritance. If you use OOP like that it is quite neat.
Although it's certainly not the only way to solve every problem.

>>56545295
Add unfuckable to the list

>>56547168
What retard would want that

>>56548849
>my programming paradigm can't handle that so it's a dumb feature

Honestly, how do you even do enterprise tier CRUD applications without OOP?

>>56549055
The fags who make these threads have never even touched an enterprise application

>>56549055
Why do you think it would be difficult?

>>56549055
I work on one of those, and it's ancient C++. There's very little OOP in there, even though it's huge. They didn't even bother to use any ORM, too.
As you can guess, working on it is a nightmare.

This is a great talk about data driven design as opposed to OOP. It's worth a watch if you are interested

https://www.youtube.com/watch?v=rX0ItVEVjHc

>>56544929
>"Good programmers" don't exist.

The limit was zero, so this is still true.

>>56545151
and, management replies with the magic formula that fucks you in the ass since the beginning of your job in this industry:
"The customer said so".
And you, powerless to face such a counterstatement, get to work for the next half a year on a completely useless feature while cursing the day you tried to use OOP "correctly".

>>56546915
Composition > Inheritance

except for like 1% of cases, and those can be solved by other means anyways.

The truth is that programming languages should provide facilities for composition instead of facilities for inheritance.

>>56541392
>I can't modify anything without completely rewriting it
how to know you're a shit developer 101

>>56547175
because at a certain point it ties your hands impeding you to improve your architecture further. It's inflexible. Yes, even with the composition over inheritance and all the attempts to bolt aspect oriented features over it.

Inheritance as a standalone feature works very well when you have to create a particular tree-shaped data structure that needs reusability. The big problem that pure OOP has is - tying functions to the components of that data structure. If, at a certain point different branches of that OOP tree need similar functionality - you're fucked. All of the shit like Abstract/Partial Classes, Interfaces, Multiple inheritance - it has been created to cope with this lack of flexibility, and doomed the OOP to be the pajeet paradigm forever because of the mess it creates when used in really large projects.

So that's why people hate OOP. You don't hate it if you don't know it too well.

>>56544954
>not creating a god-tier architecture where adding any feature/component is a piece of cake

>>56545584
>how to spot a /dpt/ fizzbuzz warrior
that isn't valid syntax in any language

>>56548340
pure functional paradigm will suffer the end of the communism - no one will be able to implement it correctly because it's pretty hard to understand for the OOP pajeets who currently love to call themselves *insert OOP lang name here* gurus. So, a load of crap will come out and will damn the FP forever as hard to understand - while it is really not.

>>56547168
that's just a new branch of Protists though?

>>56549055
easy as fuck. Include the DB connector, and write a bunch of files for every object in the DB by composing the particular data you need with the default function. Boom, you're done.

>>56553627
Why would the developer curse anything?

It's the retarded businessman that's going to get fired for approving such a request and bottlenecking development for very little pay-off

>>56554081
because guess who will be blamed for the customer dissatisfaction? Yes, exatly - it will be you. And since you're the programmer - who works more with computers - and he is the manager - who works with people and knows how to convince them - guess who will be trusted more by the boss.

>>56554262
what an outdated mindset

in the contemporary agile work environment the manager gets shit on a lot more than you'd think

>>56547152
God/Evolution uses OOP
C fags BTFO

>itt: im too stupid to understand it, so i hate it

>>56554664
That's the mindset of a company still using OOP for important projects

>>56554664
>>56554917

I'm also with the pessimistic guy..

I heard the "the customer wills it!" often enough and sometimes you don't even realize how fucked you are before you start to implement some feature and suddenly..

>Dear maintainer:
>
>Once you are done trying to 'optimize' this routine,
>and have realized what a terrible mistake that was,
>please increment the following counter as a warning
>to the next guy:
>
>total_hours_wasted_here = 42

>>56539985
>humans are animals

>>56555341
>humans are not animals

>>56546915
At first I was angry that C# only allows for inheriting from one class at a time but then I realized it is a good thing.

File: 1473379897664.jpg (14KB, 321x312px) Image search: [Google]

14KB, 321x312px

>>56549004
>food analogy

>>56548014
Really easy and also really stupid, what's wrong with you?