/dpt/ - Daily Programming Thread

What are you working on, Jow Forums?

Last thread:

Attached: comfy lain.jpg (1300x1244, 809K)

Other urls found in this thread:

jukedeck.com
svgur.com/s/AAz
svgur.com/s/A9_
gitlab.com/ryukoposting/weblisp
twitter.com/NSFWRedditGif

using neural networks to generate music

Attached: lain14.jpg (720x540, 37K)

First for webdevs are worse than pajeets.

Dear CS majors who shit on self-taught devs,
It’s great that you paid thousands of dollars to learn about binary trees and understand the complexities of discrete math and science. You know what I did? Googled that shit and am now living my best life, debt free. But continue...

I already created that user jukedeck.com

I am self taught, and I'm in college to get better
Enjoy webdev, seething codemonkey faggot

>2019
>Zhe doesn't know Reason

Attached: Girls.png (449x401, 490K)

Sperging over which distro to pick.

Tfw want minimalist bloat-less freedom distro to be used for development

I'm writing my own web browser from scratch. I plan to get full support of HTTP protocol then move onto making rendering html, then parsing css and rendering html+css and then work on js support last.

...

> from scratch
Just use chromium, goy!

I'm creating a new blockchain that will end bitcoin and make me rich

Chromium's source code is an absolute mess and it's bloated.

Is Python 3.X finally worth using instead of 2.7?

Just use arch. Stop being a faggot, and choose something that works.

It always has been, what have you been smoking for the past decade?

newfag

What do I do when I want stars on my Github but my projects don't get any attention even when I shill them all over?

Attached: k-kawaii.png (1280x727, 1.05M)

>It always has been
Can I get an answer from someone that uses python libraries?

C master race reporting in

The scope of the features added in Python 3 is breathtaking, the unwillingness of a bunch of autists to update their code is irrelevant.

C is one of the worst languages I used. No OOP, no FP. What the fuck is the point? It's a dinosaur. Legacy code. Just shut the fuck up already.

How much are they paying you to say this?

Nothing, I'm just experienced with the language.

engine

t. brainlet

Does arch support muh freedoms

Underage, ban

Look man, I just want an unbiased opinion.
I'm language agnostic and I just don't want to have to swap between two versions of a language just cause a specific library I need doesn't work for one for the version I'm working in but does in the other one.

People are refusing to use python 3 for no real reason and it's hampering development and progress. Use 3 and be patient and/or creative. Otherwise, you're part of the problem

Take it easy, RMS. You have to be realistic, the use of non-free hardware and software doesn't make you part of the problem. You use what you have.

Elixir web app

suck my dick concurrency cucks

nobody cares

You have a containment thread you white pajeet.

I'm working on my lisp web framework thing again. Have someone from IRC helping me now, too, so that's fun.
He's adding support for a ton of the HTML5 tags at the moment. I'm currently adding support for variable assignment.
cons doesn't have a keyword yet, but the logic for it is done. list is also done. Precompilation is verified clean at the moment. In retrospect, this might have been a lot easier to just write in Racket or something, instead of writing an entire lisp interpreter, but whatever.

I know exactly two (2) webdevs who are really fucking good at programming, and it seems like neither actually gained their skills from webdev
imagine a webdev pajeet, though

Attached: weblisp0.png (811x805, 42K)

>you use what you have
Sure, but there’s no restrictions on choice of software/distro

> imagine a webdev pajeet
The thought alone gives me an aneurysm

Cute logo.

I can vectorize it / do any work if you want to make a request

>but there’s no restrictions on choice of software/distro
Not sure what you mean.

cute

Attached: lain_218.jpg (1520x1080, 189K)

Pheonix still gets cucked on by java based web apps heh

What exactly is your framework for? It this another compilers to javascript meme?

You don’t have to “use what you have” if what you can have is anything. It’s not about being part of the problem as much as it’s about not having to deal with the problem/be subjected to it

Question, what would the advantage of me using this framework be over compiling lisp to webasm?

The david and goliath story is bullshit. You and a few million autists aren't going to accomplish shit.

You dont have to topple companies to avoid it for yourself

>You dont have to topple companies to avoid it for yourself
Then enjoy the wonders of free and open hardware.

I was asking about software

I'm thinking about writing my own eHealthcare system. I currently work for one of the biggest in the market but it's a shitshow and I feel like I could rewrite a better one at this point in my life. My only issue is that it's a conflict of interest and many other legal issues.

I'm thinking maybe about doing it entirely anonymous, but gaining users would be almost impossible then due to government legalities and laws for customers.

I need to think.

What is Chaiscript

Attached: 1533859327004.jpg (958x1024, 124K)

then use arch! goddamn.

if you could vectorize it, that would be amazing. seriously, I made that thing in about 20 minutes in GIMP. It's horrible once you start looking at it a bit closer.

It does not compile to JavaScript. It does not support JS. I refuse to support it because I am strongly opposed to the following things:
>offloading computation on users' machines that can be done server-side
>use of convenient tools over the ones that do the job best
>forcing browsers to not just be a document viewer, but also have an entire fucking interpreter and runtime system for a turing-complete language completely unrelated to the language of the document itself
I would at least consider support for a very, very limited subset of parenscript or clojurescript, but absolutely nothing more than the simplest basics. There also will not be syntax for
>cookies
>fingerprinting
>DOM
the former two for obvious reasons. There is no direct support for DOM because it is completely unnecessary when your syntax is completely homogeneous, instead of having shit spread across 3 languages that are each bad at their job in different ways.

So this is basically for backend rendering that spits out HTML/CSS for browsers to recognize?

But written in a Lisp? I'd also just like to say for all the bad rep JS gets it's very nice for dynamic pages that reload data in the background which is wonderful for monitoring tools in networks, etc.

On it. Be back shortly

My first thought about Lisp and other programming languages in the context of web development is that they're good for recursion. And because HTML is nested and semantic you could probably come up with some creative ways to represent things

trying gentoo

But why this over WebASM compiled lisp? I'm not trying to grill you, just trying to improve your pitch

I'm not the guy those are just my thoughts

because it isn't javascript. It compiles to HTML and CSS, and it supports CGI.

>written in lisp
the language is lisp, the compiler is C.

HTML actually translates quite nicely into a lispy syntax:
(img "i.imgur.com/YYKGVl2.jpg"
(width (px 100)))
;; or, equivalently
(img "i.imgur.com/YYKGVl2.jpg"
(width 100))

in precompilation, this would get translated into raw HTML wrapped inside a (raw _) expression, which is then evaluated when a request is made for the page. precompilation is optional, of course, but it will improve performance for pages with a large amount of static content.

nobody cares

What does this do that existing CL webservers don't?

Anyone here work in Big Data? My company is switching over to Apache Spark this coming year and I want to learn as much as I can before hand.

You fucking idiot. Big Data is a fad and will die out soon. Get the fuck out of that field before it's too late. You're not a scientist or a mathematician sweetie.

Well, it isn't a web server, firstly. It's meant to slip into pre-existing and widely-used infrastructure cleanly (e.g. Apache or Nginx) You can think of it like a CGI scripting language on steroids. But, instead of a language like perl that was simply adapted to CGI/general HTML generation, this is meant for the specific purpose of generating HTML and CSS.
From what I've seen, most similar projects are unmaintained, and/or are lacking support for modern CSS, e.g. CSS grid. My project is practically designed around support for HTML5 and CSS grid.
It also handles string literals a bit differently, in an effort to make it more convenient to write long pieces of text (like blog posts, etc). Newline characters are permitted inside strings, and are treated as newlines in the HTML output. This would be bad in most programming languages, but here it's perfectly fine as HTML pretty much ignores user-added whitespace anyway. It allows for much more convenient editing of bodies of text than would be possible if I were using a standard lisp.

Alright here she is. Sorry that took a little longer than I thought it would

svgur.com/s/AAz
svgur.com/s/A9_

The first one is based on the official logo (the lambda is the same size and it's sans the border) and the second one is based on your version

The half-lambda furthest to the right is just slightly different because the bezier curve wasn't quite right

Lemme know if you want anything else done, I'd be happy too

Attached: v5icckq11fy01.png (720x646, 413K)

(img "i.imgur.com/YYKGVl2.jpg"
(width (px 100)))
;; or, equivalently
(img "i.imgur.com/YYKGVl2.jpg"
(width 100))


What does this transpile to?

FUCKING BASED
thank you so much, user! should I cite you somehow on the repo?
If you find the project interesting, I could link the repo to you but I don't want to start shilling gitlab links on here unless someone actually asks

Attached: 1542860901061.jpg (941x812, 87K)

Just throw my name in the credits - LemonPieKneeGrow.

I'm afraid of people knowing that I browse Jow Forums

Just drop a link to the repo or some kind of way to find you and I'd be happy to help however I can by DMing you

peacepeace

Attached: 1538231953805.jpg (570x396, 60K)

Alright, your acknowledgement is in the development branch for now, but it will get pushed up to master in the next few days. Thanks!

gitlab.com/ryukoposting/weblisp
a lot of the discussion happens on an IRC channel, I'll send you info over gitlab since I don't want to shill the channel here.

oh wait
I've been had
>knee grow
you son of a bitch, I knew something sounded weird about that

I hope i made it into your commit history

Attached: 1544500384018 (1).jpg (399x351, 67K)

you narrowly missed, dumb frogposter. I guess I'm a dumb animeposter though so I shouldn't be mean- we have a common enemy.

Attached: 1542523277131.png (1075x1518, 1.76M)

How is it?

I have a job interview for "lead engineer" tomorrow. They contacted me not the other way around, but I don't really think I'm qualified to be a "lead" so I feel a bit weird about it.

Is this "lead" thing just meme business mumbojumbo? I mean, I checked the company site and it seems like their team is too small for me to "lead" any type of subgroup. So what the fuck would I be lead of??

Do you expect at least a petabyte of data? If not, a few relational DBs separated by domain will do a better job for you. Big data is mostly a buzzword.

In your case lead just means you are the only one.

You're basically sub-manager of the team and get all the blame and responsibility of meeting deadlines. Unlike your manager, you'll know programming and tech and be responsible for good architectures.

If you're the lead on a team with no one under you i'd be worried.

even a small team of engineers needs a leader.

if that's what they're really hiring for, you can expect:

>writing tech specs
>breaking down stories
>attend more meetings
>assisting your team when they're struggling
>tracking story points
etc.

I'm completely unqualified for this shit.

I don't know why they would contact me. I didn't lie in my resume.

Is it for a startup? Have you researched the company?

Go into the interview with an open mind, and get a lay of the land. Don't let on that you feel underqualified. Be on the lookout for red flags, and ask about the size of the group you'd be leading.

whats better a switch of ~8 cases or a lookup table of the same number of cases?

Attached: eeb161ed0103676b8657f90a21f519e8.png (451x619, 428K)

switch

frogposters > boobposters

switch will be optimized to jumps in assembly
lookup tables aren't and incur memory lookups

Also the switch optimization will only apply to typically somewhat close ranges of switch. Like 0, 1, 2, 3, 5, 6... If you have a switch that matches on like 0, 1, 1941, 1249129 then it's not going to optimize it and revert back to a series of compares/jumps(ifs). Also ARM and Power probably won't see it.

In the end it probably doesn't matter unless you have a shit ton of cases and the right one is usually near the bottom because it'll still be faster to do the compares than the indirect memory lookup

To add a point, if there are enough sparse cases, some compilers will emit binary searches

if you'd rather look at some dumb frogpost than boobs, something is definitely wrong with you

What about a frogposter with boobs?

don't you dumb frogposters usually have a collection of dumb frog images? post your dumb boob frog image and I will judge its merit

I programmed this image in Lisp.

Attached: file.png (1075x1518, 1.94M)

it sucks but it made me laugh, you're ok frogposter.

Lisp is the most powerful programming language.

Horrible. Why would you ruin an image like that.

What the shit is going on here?
class Test1:
__a=0
def __init__(self,num):
self.__a += num
def get(self):
return self.__a

class Test1:
__a={}
def __init__(self,num):
self.__a[num]=num
def get(self):
return self.__a

a=Test1(5)
b=Test1(10)
a.get()
>10
b.get()
>15

a=Test2(5)
b=Test2(10)
a.get()
>{ 5:10 , 10:15 }
b.get()
>{ 5:10 , 10:15 }

Why does class Test2 reference the same dict object on difference instances? How do i stop this?

Attached: tux.png (251x377, 10K)

>switch will be optimized to jumps in assembly
How

And also both cmp and jmp are 1 op cycle so it should be the same or am i wrong?

sorry, corrected the code i had it wrong
class Test1:
__a=0
def __init__(self,num):
self.__a += num
def get(self):
return self.__a

class Test1:
__a={}
def __init__(self,num):
self.__a[num]=num
def get(self):
return self.__a

a=Test1(5)
b=Test1(10)
a.get()
>10
b.get()
>15

a=Test2(5)
b=Test2(10)
a.get()
>{ 5:5 , 10:10 }
b.get()
>{ 5:5 , 10:0 }

I clearly create two instances of class Test2 yet they use the same __a dict variable. why?

hmmmmmm, onthe official chart seems to disagree so try again sweaty uwu

Attached: 418bf913274812dbe37f9f4ebd4e528e3e15b246570155a38046809c674dc236_1-1.jpg (583x622, 79K)

go is cute...

Mmmhhh Haskell is difficult and has a weird syntax