/dpt/ - Daily Programming Thread

operator overloading is dumb
for a properly designed programming language, see Go

imagine what you could accomplish if you stopped caring about what other people were doing

doing math without operating overloading is dumb
who wants to do vector1.add(vector2)

opinions on hackerrank? I have to brush up (read: actually learn) sepples for several job tests soon (now), just going through hr problems.

op overloading gets abused (even in the sepples standard, see >), but i really do like it and agree with you

This is the proper way to declare an array in Assembly.
table0 dq 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
[\code]
am joking, pls no bulli

all power can be abused, doesn't mean you shouldn't have it

too bad, you are getting bully for [\code]

agree

>doesn't mean you shouldn't have it
that, my friends, is how C++ came to be

>for a properly designed programming language, see Go

Attached: 1540183613198.jpg (682x761, 324K)

lol Ruby
ranks = (2..10).to_a + %w(A J Q K)
suits = %w(hearts diamonds clubs spades)

Card = Struct.new(:rank,:suit) do
def to_s
"a #{rank} of #{suit}"
end
end

deck = suits.flat_map { |s| ranks.map { |r| Card.new(r,s) }}
deck.shuffle!

Except he's right though. Power is only bad in the hands of retards, and retards shouldn't be programming in the first place nor should there existence be acknowledged or considered.

Can someone please tell me I am not wrong here. This was a question on my OS midterm. My answer was "it depends on which job the scheduler decides to run first. Since all threads arrive at the same time, the FIFO scheduler can choose any one." I got it marked incorrect. I talked to the TA and he said that you had to pick one possible schedule and calculate the turnaround time. But the question said what IS the average turnaround time not what is one of the possible turnaround, implying there was only one. The fact that there is only one solution to the question on the answer sheet and the fact that there is an example in the book which is almost exactly the same except it says " Imagine three jobs arrive in the system, A, B, and C, at roughly the same time (Tarrival = 0). Because FIFO has to put some job first, let’s assume that while they all arrived simultaneously, A arrived just a hair before B which arrived just a hair before C. Assume also that each job runs for 10 seconds. What will the average turnaround time be for these jobs?" It's obvious someone messed up writing the question and can't admit their mistake. Now I have to lose points on my exam. There was another instance of this on the exam too where he wouldn't give me points back. And the practice questions we got were full of mistakes like this. Please tell me I'm not going to have to deal with non English speakers this much in the job world. These foreign assholes are everywhere in academia.

Attached: Screenshot at 2019-03-29 21-39-10.png (820x204, 34K)

here's my impression of Ruby
class Fag:
belongs_in_mouth: dick
end

Man.fuck(Fag)!
elsif gay:
gay!
I'm gay! {I'm | super | gay! | always!}
end

elsif!

Getting into .net stuff, VB since we are migrating out of VBA to it. Hate VBA, but VB is comfy enough that I forget it's VB sometimes. Shit has lambdas yo.

Also finally getting the hang of unit testing with mocks. Setting it up is hard becausr of having to rewrite things, but the end result feels better.

Life at work is pretty okay.

Now back to my @home sepples "linkage router" project.

not your homework help board

You can design a powerful language without making the mistakes of C++. Power is not proportional to the number of language features.

I didn't ask for help with my homework dildo

What is the best way to represent booleans at the low level?
>Only 0 is false
Is bad because bitwise NOT doesn't work on true values.
>negatives (excluding 0) are false
Is good because bitwise NOT works like boolean NOT. Is bad because true 0 is not common convention.

Are there any representations that have false 0, bitwise NOT working like boolean NOT, and are zero cost?

I agree, but operator overloading isn't a mistake.
In fact the reason why C++ abused > for streams was because it actually wasn't powerful enough to implement better formatting syntax. But it's getting closer to that and should be possible in C++20 to do print(x, y, ...) while retaining the same upsides of standard streams and none of the downsides of printf.
C++ needs to be _more_ powerful.

>Haskellers are trannies
>seppleshits are mentally ill
>Ctards are retarded
>JavaScript rocks
What other memes?

C++ needs to die, its too bloated

Why would you use bitwise not on a boolean? Use logical not you baka.

I'm using Assembly.

And?

all booleans are represented as 0 being not
bitwise logic is not and is not supposed to be equivalent to normal boolean logic
and and or are not equivalent, so why should not be?

>you have th experience we desire
>you're personable on the interview
>we rly impressed by your GitHub projects
>NOW WE ROLL 2 D20S TO SEE IF YOU MEMORIZED THIS LEETCODE PROBLEM OR NOT

fuckkkkmkkk oofffffffddd

Attached: 1547509208230.png (1000x1184, 616K)

pythonfags are babbies

>print(x, y, ...)
this a joke, right?

C++ code is actually getting less bloated.

Assembly doesn't have a 1 cycle logical not. Doing it the C way is slower.

the language itself is bloated
it has too many features and caveats
the syntax is awful
the build system is awful

It's the only way to get compiled and type checked format strings without introducing constexpr parameters into the language.
No, it's not a joke, and it's better than both C++ streams syntax and printf.

I think he actually meant open-std.org/jtc1/sc22/wg21/docs/papers/2019/p0645r7.html

Assembly is not a CPU architecture. What shit filled CPU are you programming for

Intel.

Operators in general are a mistake. Overloads are just the workaround for the underlying design failure.

>the language itself is bloated
Can all be learned in ~2 years
>it has too many features and caveats
Can all be learned in ~2 years
Each standard release, it takes like maybe a few days to learn the new features.
>the syntax is awful
Improves each standard.
>the build system is awful
What build system?

The advantage you get out of it is convenient performance and zero overhead abstractions.
Worth it if you actually care about performance and efficient code.

That's an accurate impression of Ruby on Rails.

Yeah and you can learn any other language in two weeks
the syntax doesn't "improve", you can't change the syntax without changing the language
You don't need to have a language as shit as C++ to have zero overhead abstractions
you're as brainwashed as a haskelltard

why do i end up spending so much time doing jack shit? i've been a programmer for years, but i still find myself sitting there with my mind blocked up and i can't write anything. eventually something clicks and i can get a lot done, but a solid 50% of my time just isn't productive and it's very frustrating.

No, it's not that. The snippet I posted relies on non type class instance template parameters being introduced in C++20.

What would be the advantage of it over the std::format specification in the document I linked? I'm actually curious.

Not him but I just started learning C++ and I'm not finding it that difficult.

You're a mistake.

>Yeah and you can learn any other language in two weeks
With shit performance, code gen, etc. So who cares? May as well just stop programming at that point. There's no point in writing code that isn't efficient.
>you can't change the syntax without changing the language
The language changes every standard. You don't even know what you're talking about. Here's just one example: fold expressions.
>You don't need to have a language as shit as C++ to have zero overhead abstractions
I agree but C++ is currently the only one that doesn't have a borrow checker.

You can not learn a language in 2 weeks. You will be able to write stuff but you will be far from perfect.

>The language changes every standard. You don't even know what you're talking about
So with each new standard your old programs won't work? Because that's what 'change' is, anything else is an addition, not a change

Depends on your prior experience

>the syntax doesn't "improve", you can't change the syntax without changing the language
You can introduce syntax changes that don't break backwards compatibility just fine.

so to create a socket for interprocess communication do i just have to set the host to 127.0.0.1 and choose whatever port?

I haven't even read that specification. But looking quickly over it it looks like they're passing the format string in as a regular parameter which means they can't do compile time format string compilation or type checking. So mine would probably have better codegen in more cases.
Unless that proposal proposes a new language feature that does allow them to do that, in which case I'd appreciate if someone who actually read the whole proposal point me to it.

Fuck off with your semantics. Addition is change too. I'm am not going to argue this dumb shit with you.
The committee changes the language all the time in a way that doesn't break backwards compat with previous standards. Sure, you can program in the old standard if you want but the new standards provide better and cleaner syntax for some things that you can and should use.

The distinction is important because throwing more shit on the pile is what makes it so bloated, and you can't fix shitty design decisions that have been in there since the beginning without breaking everyones programs

>just started
that's why

See the following anchors:
open-std.org/jtc1/sc22/wg21/docs/papers/2019/p0645r7.html#Safety
open-std.org/jtc1/sc22/wg21/docs/papers/2019/p0645r7.html#CompileTimeFormat
open-std.org/jtc1/sc22/wg21/docs/papers/2019/p0645r7.html#BinaryCode

>and you can't fix shitty design decisions
They don't matter anywhere near as much you'd like (to support your opinion).
All your complaints are a consequence of C++ being old and being based on C. Being based on C was definitely a mistake don't get me wrong. But you can't do anything about C++ being old.
That said C++ still has important performance advantages and the language gets better to use every standard.

"Just one more hack, and it will be perfect" - the mantra of the C++ addled mind

>But you can't do anything about C++ being old.
yes you can
you can make a better language and stop using C++
you've gone from "C++ doesn't suck" to "well it's old and based on C"

I've got a job writing c++ lined up after school. I didn't know c++ when they offered me the job, but I've been learning as much as I can in my free time. Anyway, the job is "software engineer" but one of the guys who interviewed me told me I can expect to be doing tests for the first 3 months until I learn the code base and get up to speed. Is this normal? I do not want to be a tester, and I have read stories of people getting roped into testing and support jobs with the promise of software engineering roles.Should I be concerned about this?

>it depends on which job the scheduler decides to run first
It's FIFO, they are run in the order listed (A,B,C)
I don't think exams are a good place to be pedantic.

>open-std.org/jtc1/sc22/wg21/docs/papers/2019/p0645r7.html#CompileTimeFormat
>It is possible to parse format strings at compile time with constexpr functions which has been demonstrated in the reference implementation [14] and in [18]. Unfortunately a satisfactory API cannot be provided using existing C++17 features.
>requiring a user to create a new type either manually or via a macro for every format string is not acceptable
tldr you can't actually do compile time format string compilation with this proposal.
As with the last link, they just got lucky with the optimizer, that's all. It's good they have the type safety at least, though it's unclear wether that type checking is being done at compile time or not.
Mine still looks to be better, performance wise.

That doesn't make sense. FIFO does not choose based on alphabetical order, it chooses based on arrival time, and they all arrived at the same time. The TA even told me that if I scheduled them in any order (didn't have to be ABC) the answer would have been correct.

>you can make a better language and stop using C++
That's not doing anything about C++, retard. Nice post faggot.
>you've gone from "C++ doesn't suck" to "well it's old and based on C"
No read my posts again.

who said anything about "doing anything about C++"? I said it needs to die, not that it needs to change. As in better languages need to exist so people can stop using it

Writing unit tests? Doesn't sound too bad. Or do you mean clicking through the gui like a monkey?

In many companies everything moves so slowly you would not believe it. My first 2 weeks I did nothing, then 2 weeks working on some internal app crap.

As long as your title is software engineer and you get paid like one it seems ok.

>who said anything about "doing anything about C++"?
Me, the one you replied to with that post.

The depends whether you consider your goal to be "making a good programming language" or "mantaining C++ at all costs"
it's pretty clear what camp you belong in

>As in better languages need to exist so people can stop using it
I'm not disagreeing with you on this, I don't know why you think I am. But if such language does come into existence, it should uphold the same principles of C++ to an even stronger degree to be considered an actual C++ replacement. Rust comes pretty close to that but it completely ruined usability with the borrow checker, and lacks meta programming.

i just fizzbuzzed in C. are you proud of me

Attached: 1525104783584.gif (640x480, 1.26M)

post code

>tldr you can't actually do compile time format string compilation with this proposal.
That's only for the standard version - which is a good thing, because if they introduced shit API now they wouldn't be able to change it in a future release, so it's better to hold on for now while C++ gets the required feature, though you can just simply add fmtlib to your project which has an implementation with the less-than-satisfactionary API which does in fact provide compile time formatting. The proposal branch in the repo has also more elaborate examples that also get optimized fairly decently. Really though I think it's overall great, you have to consider the fact that they're working towards something that will be in the standard that has to cover a lot of specific cases and various standard conforming APIs as opposed to implementing something that is tailored to your specific needs.

>They don't matter anywhere near as much you'd like
They don't, until you have to work with someone else's code, and they insist on abusing every obscure and poorly designed language feature they can think of.
>Being based on C was definitely a mistake don't get me wrong.
No, being C is one of C++'s few good choices. Most well-written C++ code only uses an extremely small subset of the language that a C library can't replace (which has become much rarer in the last decade). The only exception is video games, where C with OOP is a necessary evil.

That depends what you consider the principles of C++
zero-cost abstractions, managing your own memory, sure
throwing every feature and the kitchen sink in there, terrible syntax, header files, not so much

>Writing unit tests? Doesn't sound too bad. Or do you mean clicking through the gui like a monkey?
I don't know. I guess I probably should have asked.

>As long as your title is software engineer and you get paid like one it seems ok.
I want to write code though. It also won't look very good on a resume if all I can say I did at my last job was write tests. I understand that I can't just walk in there and start rewriting all their code, but I want to be contributing as soon as possible.

I'm proud of u user

My goal is to write efficient code with the least amount of effort. You're a dumb retard with blind anti C++ goggles on.
And yes, it is true that C++ as a language is improving in terms of power and usability. If you're not a dumb retard who can't hold all the features in your head then this doesn't become an issue.

>where C with OOP is a necessary evil.
how is it evil if it's necessary?

#include
int main()
{
int i;
int j;
for (i=1; i

Attached: thank.png (746x512, 105K)

Your goal is to defend C++, which makes you the retard, not me
I have no problems using C++, I just don't think it's a very good language, objectively, it's too old and bloated

dumb frogposter

Java newfag here doing a course in the basics of programming, but I have a question:

Why would you ever use the basic array syntax over the curly bracket way?

For example, why would I do,

int array[1] = 1
int array[2] = 2
int array[3] = 3

over doing,

int array[] = {1, 2, 3}?

Attached: Questioning.jpg (294x171, 8K)

You wouldn't do the former because that's not valid code.

>Your goal is to defend C++
Because it allows me to achieve my actual goal I stated. I think it's a fine reason to defend C++. If you're not interested in performance then you can just fuck off from this conversation because you're wasting both our time and energy.

good frenposter

"defending" anything makes you a brainlet because it means you have an emotional attachment to something instead of viewing the situation objectively. I use C++ too, but I can fully admit it's failings instead of making excuses for them and it's long overdue to be depreciated. Performance has no reason to be exclusive to C++

Yeah ok nice job missing all my points, you're just making yourself look retarded now. Not even going to bother.

It is, though, aside from the missing semicolons but I'm not in an IDE right now so cut me some slack.

That is the syntax of a basic array. Square brackets differentiate a variable from an array.

you don't have any points

ok retard

It absolutely is not.

imagine being this insecure

who?