/dpt/ - Daily Programming Thread

Old thread: What are you working on, Jow Forums?

Attached: 1555627922958.png (714x1000, 469K)

Other urls found in this thread:

yaldex.com/game-programming/0131020099_toc.html
twitter.com/NSFWRedditVideo

nth for Nim!

2nd for buttraping bugs

Just move the keyboard up a bit then your huge tits won't be pressing down on the keys for fucks sake woman.

no these nice preggo tits need a good ol' milking

>preggo

Attached: sweaty.jpg (225x225, 9K)

Working on translating a Lua script I made a while back into C

Oniichan look! I found this thing JavaScript and it rocks!

Attached: 20171208-18.jpg (500x750, 266K)

Why are OS development resources so lacking, bros?

I haven't found ANYTHING to help me with the details other than the OSDev wiki, some outdated textbooks from the 90's, and a handful of random blog posts.

Where do you guys go to get more info?

cute doll but FUCK javascript

my dad keeps trying to teach me php basic and c when I want to learn python and haskell

dolls are creepy

No idea, but how is it possible to use GL on bare metal?

I found a nice asp core 1.0 app how do I make it 2.2
I created a new 2.2 project and am copying the code from mvc classes etc hope it works.. it probably wont

Noone cares, doll website

web programming is real programming

>os development
>implying

you'd have to make your own library that translates opengl calls and does the necessary transformations to rasterize that stuff to hardware graphics

pls

You are absolutely right — web, noweb and cweb were created for real programming.

Attached: knuth_news.jpg (600x400, 76K)

Watching Handmade Hero, trying to absorb that guy's knowledge and making it my own without it getting lost in the depthless abyss that is my mind.
In fact, I just watched the Handmade Hero guy trying to demonstrate something (related to Windows memory pages) by dividing the memory size by 4096, multiplying it by 4096 and getting all excited over it coming out the same value.
Maybe this stuff isn't meant for mortal men.

Going to sleep and waking up every day, hoping that one day I'll wake up and suddently OOP will make sense to me.

that series is just a meandering waste of time that's going to fill your head with useless information
not a good tutorial

oh no it's retarded

Just think about OOP as containers where you put code that makes sense for that container so for container animal you will have eat() sleep() and its attributes like...fur color, size, etc. or container...Paladin where you will have attack() castMagic() and attributes strenght, faith etc. etc.

Attached: 367373467.jpg (477x477, 41K)

tutorials are for late-diagnosed ADHD sufferers.

What if we didn't get it diagnosed yet? Does that make tutorials a good choice until we decide to get our ADHD properly diagnosed?

yeah i agree but if thats what you want, dont watch handmade hero, its several thousand hours of what amounts to programming trivia

how unfocused.

You likely already understand OOP but fail to see valid use cases for OOP, which is common amongst frosh college students.
OOP is an attempt at teaching critical thinking and problem solving, which at its core is simply the pattern of breaking down larger problems into smaller problems which can be solved and put back together to solve a greater puzzle.

But because you're in college, you have never seen a "greater puzzle" in your life so you fail to understand the purpose.
OOP will come naturally once you try to solve a larger problem, one where you will find yourself reusing code.

There are countless open source operating systems to look at for help/inspiration.
Personally I try to figure things out as I go. Just remove "OS" from the problem description in your head and approach it like any other programming task.

because OS development is an ongoing research problem and the OS that you're using is the culmination of that research.
If you want to understand what goes into developing an OS, pick an OS component, any OS component, and try to figure out what you don't like about it, how it could be better, and what you lose in exchange

>ongoing research problem
thats a bad way of describing it

Because millions of dollars go into OS research and the guys certainly aren't going to share it with you.
I mean, do you really believe writing an OS plus all the drivers needed to interface with hardware is going to be just laid out in a nice book or something?

I call it as I see it. Just because you're not using an OS written in a memory-managed language doesn't mean that your OS isn't on the edge of innovation.

>tfw no rust gf

Attached: 1413777090510.png (1920x1080, 1.03M)

boomer vs zoomer ladies and gentlemen.

Attached: Annotation.png (1276x718, 1.27M)

Redpill me Intel® VTune™ Amplifier

it's free now apparently

OS development isnt a "problem", we have functional operating systems, most OS research is just fucking around at this point

>R is best girl
Should I learn R?

everything is a "problem" unless you're an autist and think that resolved problems aren't problems.
OS research is always trying to solve a problem that we've had before, otherwise there would be no funding and we'd still be using XP because it just werkz
Nobody would bother with Wayland and Purism wouldn't be dedicating any effort to mobile shit on Gnome.

horrible taste

Attached: _program_girl__notepad_by_reef1600-d8ki8px.jpg (1024x640, 198K)

Win 3.1 was totally functional too, does that mean it should have stopped there?
Fucking faggot.

problems have solutions, undirected research is good but you aren't solving a problem
and windows hasn't really improved since XP, yeah

Lemme guess, you're a rubyfag.

>windows hasn't really improved since XP
That's the kind of idiocy that would warrant you killing yourself as to improve the average intelligence worldwide.

> scrubs arguing about 'challenges'

I was a DEC Developer for RSTS/E systems in the 1980's. MACRO-11, ASSEMBLY, DCL.

kids these days. even scrub languages like Python are too hard for them.

Attached: Cava6CL.jpg (512x512, 61K)

Does anyone have screenshots of the 2ch desktop browser application?

It looks like a good project to clone to finally learn XAML properly with.

feel free to prove me wrong, id still be using it if I wasnt forced to update
oh yeah I suppose the default color scheme got a bit nicer

I'd say that ASLR, better window compositing, unified hardware drivers, the newer DirectX APIs, Windows Subsystem for Linux, Windows Sandbox and more are fairly hefty improvements.

Thanks for chiming in, Grandpa gray pubes

It's not undirected research if people are bitching and moaning about these problems.
And there are only two labs that allow undirected research, that's IBM Research and Microsoft Research.
Product divisions don't get to do what they want.

only two places in the entire world allow undirected research on OSes huh
you better tell the guy who keeps posting his in the thread

I have yet to see him solve any problems in an appreciably different way from other OSes, so it's not research, and he's not being paid to do it either, which means it has no impact (on those who would develop OSes)

the point is that anyone can do research

Working through SICP before learning a programming language, yea or nay?

its a huge meme

nay

Does anyone know how big the speed penalty if any is for shared memory between processes in OpenMPI versus direct access with multithreading in C++ with std::thread?
And what kind of system should I use? I the documentation is large and confusing so it would be nice if someone pointed me in the right direction.

"labs"

unrelated question hoping to not start a shitstorm: I often hear the term "code reuse" outside of OO. For example stuff related to linux drivers (like the "regmap" thing for i2c and spi).
Would an OO language be good for kernel drivers to improve code reuse and generally simplify development?

*dabs*

Burn that shit and don't look back.

You can reuse code in any language you don't need OOP to do it

Just learn C at the same time, you will be well prepared for advanced languages.

>to improve code reuse and simplify [...]
no.
you can reuse code in any language and in any paradigm.

OOP is just a natural way for people to think and is easy to get brainlets accustomed to doing so.

Don't get too attached to OOP.

silberschatz's dragon book is an incredible resource

Thanks for your answers. I worked my way through 1.1 and half of 1.2 before other stuff got in the way. It's fascinating stuff but it seems like a long-term project kind of thing and I'd like to get programming sooner rather than later.

Kernel drivers deals with hardware which are obviously objects so any driver development is OOP.

Stop objectifying hardware!

Rust gf would be very controlling, only let you do things with her allowance

>tfw no yandere dominatrix rust gf

Web development unironically seems like a difficult thing to be good at.
>shittons of frameworks to learn
>tons of possibilities for malicious user input

Attached: reigen horny and lonely.jpg (918x592, 96K)

homu

Attached: 1555425428131.jpg (563x712, 148K)

Does one even "get good" at it?
How do you "get good" in a field where everything changes daily depending on framework makers' whims?
Isn't everyone just average and doing their best?

there's a world beyond JS frameworks you know.

In Haskell: is it possible to run different versions of GHC in parallel using stack? I want my own project to use 8.6, but a tool requires 8.4.

What is a good book that can walk me through making a small rpg start to finish?
Don't care about language, I found most C++ books end up going for the nth asteroid clone focusing on language rather than actual game mechanics.

what tool requires 8.4?

brittany

run the 8.4 tool in a docker container and have your main 8.6 program communicate with it over an HTTP REST api :^)

it might be possible, if you only need the executable then definitely

All right so I have some gui apps that when I compile on linux run smoothly and barely use any cpu .But I when I compile them on windows with minigw I get pretty bad performance , I have tired both qt and gtk3 so its not the ui libs it has to be something else. Surely? Any ideas?

You get good by understanding the fundamentals. Everything is just doing the same HTTP requests and displaying the same HTML as it has been for decades.

well I don't like OOP, but everyone where I work insist on wrapping literally everything inside a class. I am struggling to tell apart the cases were I am just biased and OOP makes sense and the cases where it should be avoided.
Talking about hardware I once saw a "device" class with a "configuration" class inheriting from it all to set some dividers value in a PLL. There was also a third class managing the link to talk with the device inheriting the correct addresses and other stuff from the "device" class. The data packages themselves sent by the "link" class were generated by a small "data package" class.

Still haven't heard a valid reason to use C++ over Rust. Nobody cares that it's made by trannies. If Hitler himself made it while murdering my grandmother, I'd still use it over the heaping sack of garbage that C++ is.

Attached: oldman.png (1033x765, 1.21M)

your best bet might be to run the program with profiling. i dont know if mingw/cygwin make gprof available on windows, but if they do then try compiling the program with debugging symbols and profiling, then run your programming and examine the profiling results.

If you want to make a game use RPGStudio or something. Do not write an engine because you want to make a game.

To be fair, if Hitler made Rust, I'd also use it.

It's been a while ago, but I believe the program crashed while using GHC 8.6.

Eh, it's really not that hard these days. Lots 2-5 man indie groups do it. Not that you have to or should if you're just making something fairly standard but if you want to and have a really particular feature in mind that'd be hard to do without total control, it's really not that difficult.

maybe yaldex.com/game-programming/0131020099_toc.html

no forget that. start by writing down your requirements and/or desired plans, and estimate priorities ("requirements" as stated are not always possible)

Imagine if Hitler were born in modern times. Maybe he wouldn't have tried to kill all the jews and instead make a beautiful replacement language for C++. You never know, then maybe Hitler would be a name like Stroustrup.

>an OS specifically aiming to copy good stuff from other systems isn’t big on original concepts

Attached: 4C458544-8E57-4E48-805C-33EC68767A79.png (706x765, 857K)

I want to make a game not for the sake of only making games, but also learning how it's all tied together in an actual programming language instead of above layers upon layers of abstraction and scripting.
I'm more interested on the concepts than the actual game, as games come and go but the foundations remain.

based boomer

FUCK library package managers, FUCK soiboi configuration systems, FUCK weakly-typed languages, FUCK electron code editors, and most of all FUCK jannies

stroustrop is definitely a name like hitler after he killed 6 million good programming ideas

C is leaps and bounds better than Rust could ever be.

But rust is supposed to replace C++ not C. C will always be used by autists around the world. Including me.

Then you really don't need to care about the engine. Game mechanics are not tied to platforms.
Make the same game on a computer and for physical boards.