/dpt/ - Daily Programming Thread

What are you working on, Sup Forums?

Old thread:

Other urls found in this thread:

medium.com/@bryanedds/functional-game-programming-can-work-95ed0df14f77
strawpoll.com/9xw1bd8
en.wikipedia.org/wiki/Intel_iAPX_432
twitter.com/SFWRedditGifs

Anyone have that image that was like a program idea roulette with a shit ton of ideas listed?

卐Python卐

better-dpt-roll.github.com

thank you gcc. beautiful code as always.

Reimplementing Python in JS

an AI that shitpost in daily basis and is capable of baitposting and sauceposting.

The more I develop, the more I realize that CS is an endless pit of knowledge

medium.com/@bryanedds/functional-game-programming-can-work-95ed0df14f77

How many of you guys have Caps Lock rebound to Ctrl? Why do keyboards still have Caps Lock?

Should make a ML baneposter

If you work with SQl or VB.NET and BASIC you would know

>Add to that the fact that many engineers foresee GC becoming part of the computer’s hardware—at least, once we settle on a universal set of GC primitives.
Really?

>SQl or VB.NET and BASIC
what is this

strawpoll.com/9xw1bd8

calloc when you need it, malloc otherwise.

>Currently, the Nu Game Engine can process up to 25,000 on-screen entities at 60 FPS before saturating the CPU

Does he mean 25k quads or what?

What are some GOOD programming podcasts/audiobooks???

SE radio is fine, but 90% of the new episodes are about nontechnical shit like hiring, agile memes and culture.

What are some good audio resources about technical topics?

They've been trying to do that for ages.
en.wikipedia.org/wiki/Intel_iAPX_432
that was designed back in 1975.

So I upgraded from my Debian system's gcc 4.9 to gcc 6.3. Then since I figured I ddin't need the system's gcc anymore I removed it with apt. But now I pretty much can't build any C++ program anymore. For example when i do
#include

int main()
{
std::cout

How do you get past the captcha?

GTK and Python 2 vs 3 is an absolute nightmare

GTK made me hate C with fiery passion

Unfortunately writing php, not out of choice but because I need to eat.

Debian has the headers and shit split off in their own packages, for some reason.
Is libstdc++-dev installed?

Not him, but a Sup Forums pass can be used for botposting and posting from non-js browsers like w3m.

a file monitoring program in C that monitors files and directories recursively, detecting additions, changes, and deletions.

I want to tear my hair out.

try
always works.

So, git?

so git?

I want to write a simple notification server on Linux. I honestly don't care what language it is, it just needs to receive notifications and maybe forward them to another program if I have to, I just want to know what would be the least painful way to make this happen?

wipe it all then do a clean apt-get
Its amazing that package managers still cant resolve dependency complications.

why not inotify?

what editor is this

So, git?

go.

So, git?

libnotify

it's supposed to use stat() and getdents(), and it is not a file uploader dammit

I'm learning dr. racket in college, but I want to make games in a functional language, what should I learn now that is similar to dr. racket?

colombiANO ajaja

Common Lisp

notepad++

hola choriboludo.

que haces aqui boludin?

Managed to solve it thanks to the first reply. I tried to find whether the packages were installed somewhere or not and dpkg -i them. However it turns out clang builds against g++'s libraries on Linux. So when I saw the following line in clang++ -v
Selected GCC installation: /usr/local/bin/../lib/gcc/x86_64-pc-linux-gnu/6.3.0
I just had to look into the folder and notice that it was basically empty.
It turns out when you build gcc, it only builds and installs the bare minimum. So the solution was to copy and build the content of libstdc++-v3 in gcc's folder.

I thought about doing that, but a shitton of packages depend on the libstdc++6 so I kind of kept it as a last resort.

Actually sounds like a nifty idea.

>Nu Game Engine
Had to check that it wasn't a Sup Forums meme

Rate my ray tracer
#include // card > aek.ppm
#include
#include
typedef int i;typedef float f;struct v{f x,y,z;v operator+(v r){return v(x+r.x,y+r.y,z+r.z);}v operator*(f r){return v(x*r,y*r,z*r);}f operator%(v r){return x*r.x+y*r.y+z*r.z;}v(){}v operator^(v r){return v(y*r.z-z*r.y,z*r.x-x*r.z,x*r.y-y*r.x);}v(f a,f b,f c){x=a;y=b;z=c;}v operator!(){return*this*(1/sqrt(*this%*this));}};i G[]={247570,280596,280600,249748,18578,18577,231184,16,16};f R(){return(f)rand()/RAND_MAX;}i T(v o,v d,f&t,v&n){t=1e9;i m=0;f p=-o.z/d.z;if(.01

awful

>operator+
And you fucked it.

Not a language

Does Go have reasonable DBus bindings?

>59646343
>59646343

Let me guess. git?

>Not a language
even fucking batch is a language at this point

What the fuck are you trying to do?

Find how many times matrix n appears in matrix hs.

Windows batch scripting has always been a language. libnotify isn't a language, and I don't think it even helps you set up a notification server

I have never used git

fuck I'm all over the place. I can't decide whether I want to invest most of my time and energy into mastering C (I'm already decently proficient), Rust, or another newer language such as D or Go.
I'm sick of C# and Java.

This feels written by someone who only has a vague idea of how computers actual work. In todays modern CMPs with multi-level cache systems, it's hard enough to just keep them coherent. Just implementing a GC system on top of that would be an incredibly complex undertaking (requiring probably decades of research, and even more time to put into production), and would probably not even net you an advantage. Accessing cache (i.e. miss) or god forbid memory is already on the critical path for virtually any program.

That doesn't even take into account the maturity of compilers, programming languages ect, that would need to be re-assessed, and the fact that all consumer level HW manufacs would need to sign on to some standard, and the fact their would be no backwards compatibility (i.e. forced into original situation, if run-able at all) if it did happen, ect ect

How long did it take you?

pick the best option.
pick c++17

pick up a functional one.
imperative programming really was a mistake.

Do this honestly

> delivery

C++17 is not some panacea, the only people who feel good about C++17 are the people that don't use C++

For your input of [[1,1], [1,1]] to the larger matrix of 1's and a single 2, it should be 3. What's the problem?

But for what purpose?

At least you're not like me: still doing fizzbuzz-tier programming in Python after 7 months

I'm doing the Danmaku engine right now. Does anybody have an idea how many bullets should be supported simultaneously? I'm already designed a structure-of-array style architecture to make it fast but I'm not sure what the benchmark should be.

are you saying that Windows scripting is superior linuxshit

Some of these are impossible to write from scratch.

It's not meant to be taken seriously

google golang

It shows "4" instead of "3". Like, one more submatrix, which is wrong.

The bootloader one deserves its "fuck you" tier.
t. someone who tried doing hobby osdev

soa is literally a meme. What matters is access pattern vs cache contiguity.

I never said that it was any good, just that it is a language

Nothing says you have to do them all from scratch

What is good and bad about C++17?

How the fuck can that tiny cluster fuck of code output something that nice? I can't even do that in opengl ffs.

Nothing substantial really happened, they just added some extra features from Boost and called it a day.

>What is good
stealing good things from better langs
>and bad
as always those features are bolted, not natural

Usually nothing but this week been tripping all over C trying to make a program to serial communicate keystrokes to an arduino mega. Part 1 of ridiculous LED keyboard project which I could probably do so much easier in a number of different ways.

is there an accelerated track for learning C++ for those that area already seasoned programmers in other major languages? I don't need an entire fucking chapter dedicated to variables and looping constructs.

Do you have a better idea how to move many objects (I'm thinking 10k? Maybe it's a bit too high) and display them with a more traditional approach? The thing is I also have to account for collision checks, thus I figured not having to change the player's coordinate for each test was better.

Ray tracing is at its core extremely simple. The trouble is in making it efficient.

...

yea ya do

the render equation is an integral

Ever heard of that magical thing called a quadtree, nerd?

Can't it be done in parallel tho?

All he's saying is that you need to optimize for what your actual access patterns are and not just make things structure of arrays willy nilly. You're probably correct to do so in this case, that being said.

Optimizing collisions is mostly about doing a gauntlet of increasingly precise tests. How you structure the data is likely not significant.

So?

That would make the code a lot more lengthy, but yes, of course.

k-d tree

You can use numerical methods for solving integrals: Monte Carlo, trapezoids and so on.

You are incrementing count whenever the two overlapping matrix positions are equal, correct? So, if this was working correctly, it would do 16 comparisons, and only 15 should come out to be correct. But you only ever do 4 comparisons (obviously), and they all come out to be 'true'. With that, you should be able to tell what the problem is.

Rate my matmat function!
#define max(a, b) (((a) > (b)) ? (a) : (b))

typedef struct {
int **mat;
size_t size;
} mat2;

int matmat(mat2 *hs, mat2 *n)
{
if (n->size > hs->size)
return 0;
size_t i, j, k, l;
for (k = 0; k < hs->size; k++)
{
for (l = 0; l < hs->size; l++)
{
size_t matches = 0;
for (i = 0; i < n->size; i++)
for (j = 0; j < n->size; j++)
if (max(k+i, l+j) < hs->size) /* bounds */
if (hs->mat[k+i][l+j] == n->mat[i][j])
matches++;
if (matches == (n->size * 2))
return 1;
}
}
return 0;
}

So?

Last night I spent an hour trying to get a bluetooth ODBII dongle to connect to this scantool program that I found. I got the dongle to pair, but I couldn't get it to connect to scantool, because scantool is some old DOS tool that was modified to see USB tty devices, /dev/ttySX iirc, as COM1-8. The hackiness of the solution appealed to me.

I was apparently able to bind the dongle to a new device, /dev/rfcomm01 using some tool whose name I am forgetting right now, but I couldn't find the place in some scantool source where the COM ports were mapped to these USB devices.

mapping the dongle to the USB devices didn't produce an error when I ran the command, but also didn't work when I ran scantool.

I also wasn't able to get it to connect with bluetoothctl, but apparently Bluez5 needs some work... I'lll play with it further when I have more time.

Do any anons have any resources?

It's the core of the efficient algorithms for calculating ray tracing.