This is the current state of web development. https://github

pajeets

>>59609146

what's your criticism, exactly?

This is a result of the web being a terrible platform to develop for. So you need a bunch of polyfill libraries to support legacy browers and enhancing the underpowered styling language and a "framework" to gloss over the details of the shitty way the browser renders it's layout.

plus a live reload server so you can save and see your results.

>>59609146
Don't worry, if you're too dumb to get past the buildstep you'll never learn any of the technologies it contains anyway

This babel trend shit has to stop.
Just to support the fucking ES6 style imports these fools import like 130mb of dependeces, what the fuck.

Just use the fucking requires! Node already supports most of the ES6 shit, fucking god.

I've been playing with freecodecamp and similar things to get away from IT and maybe towards "development". Then I start into platforms and I spend more time reading than building anything that works. Am I dumb, or is the "web" field of work just a giant mindfuck? I don't know or work with any devs, just IT people.

>>59609146
What the fuck am I looking at exactly

>>59609676
Dependencies on a modern programming app's environment file

>>59609305

Why not simply not support legacy browsers? If the site is rendered in IE6, it should just display a message that says, "Go to hell and stop using Internet Explorer 6."

>>59609799
This is how I can tell you've never held a real job in your entire life, you academia weenie.

>>59609799

>what is forced legacy usage

Loads of companies are tied to ancient webapps that still require legacy browsers. Forcing updates would literally kill productivity for a whole bunch of companies purely because one developer is pissy about supporting anything but the latest and greatest browser of his opinion.

File: 1453242449111.gif (565KB, 246x205px) Image search: [Google]

565KB, 246x205px

>>59609183
>Art majors and college dropouts that took a bootcamp class that now call themselves engineers

Pajeets can at least do Calculus

>node

Not even once

>>59609146
It's retarded.

No one has the balls to write a vanilla JS library or project. It's actually not hard.

>>59609645

it's a giant mindfuck, mostly because there are so many ways to implement something. if you keep on top of things it isn't too bad.

>>59609146
Babel, React and Webpack purposely split their modules into individual dependencies, it makes your package json look long but it doesn't matter at the end of the day.

No one is forcing to use any of these technologies, there's no requirement to using ES6 / React / Build tools etc. Just stick to script tags and jquery if you want.

>>59609146
Whoever set this up doesn't know what he's doing.
All that babel shit etc. goes into devDependencies, only stuff in dependencies should be react & react-dom as that is the only stuff actually required to run the app.
You then configure webpack to build a dist.js or whatever. All that other shit is only needed for *building* the app.

>>59609146
Can someone tell me what's wrong in simple terms? I've only done a small amount of embedded programming.

>>59610898
Yeah because it makes total sense to write code for the same shit you're facing building any SPA.
Routing, keeping state and DOM in sync, ...
Any armchair coder will beat the performance of React's virtual DOM for sure ;-)

>>59609305
"live reload" is so 2010
HMR is state of the art. You save changes, the whole app state remains as is, and only module changes get loaded via sockets.
It's a very good dev experience once you went through the whole webpack setup meme desu

>>59611090
live reload and hmr is a meme

>>59611040
A web page is written in HTML.

HTML dictates what ressources are loaded, the browser interpret the ressources.

In modern web dev environments, you need a bunch of external tools to make you type as little HTML as possible. So a simple web project will have a file at its root listing dependencies. When building the project, the build tool will download all the dependencies, compile them into a single huge file. It will also cross compile you're incompatible javascript into browser friendly javascript further increasing the web page size, load time, and run time. That's why you see more and more fucking webpages with a loading screen. Each framework comes with its own set of tools, cross compilers, and even the workflow.

What's wrong? It's incredibly inefficient. Everything changes too fast. It's whole paradigm shifts every time you have to use some framework a bit above average in size. People start developing a framework, everyone move to it, and later they realize the thing is a mess and abandon ship, move to something else and repeat until they find something that last more than a few months. Starting a new project from scratch proves difficult especially if you don't know from the start what tools you'll need. Most of the time you'll have to copy paste a template project and work it up from there.

Calling people out on how there's no need to support IE6 in 2017 will get you hordes of pajeet calling you a jobless NEET. But they really are just scared of losing their worthless slave jobs.

I can't stand people who do this. This is also a problem in mobile development - where nobody really knows what the fuck they're doing so they add yet another needless dependency

>>59611138
if you are developing babbys first SPA maybe
once things get complicated you do not want to lose your app state every time you change sth in your src

>>59609146
that's a baby easy package.json file. basic basic stuff.

you're a retard.

File: ruby dev.jpg (3MB, 1366x687px) Image search: [Google]

3MB, 1366x687px

ruby is even worse OP. here is an empty project

I can't stand web developers who complain about shit like this. You are coding/programming at the end of the day for a system that is a document parser for an end users machine that isn't supposed to be able to execute arbitrary code. Of course its going to be curmudgeoned and hamfisted. If you want to program like a normal person, write a normal program.

Time to take up Gopher.

HTTP is clearly dead.

>>59611465
what did they mean by this
even a simple concat into a single file would speed up load times by a factor of at least 10x

>>59609928
I fucking HATE this. At least CS / SE graduates had exposure to some math and theory.
If you 'graduated' from a bootcamp or graduated with a degree in finance or some shit and you think you're special because you built a shitty web app in rails from a tutorial, you are a code monkey and have no goddamn business calling yourself a software engineer.

>>59611465
JUST

>>59611525
ask the guy who designed Rails. I hear even ruby "devs" don't like him

>>59609146
Yes, people use various tools to catalyze their workflow.

What is your "criticism", that there's a lot of them?

>>59609146
The actual code isn't even the problem. Configuring your build tools is.

Good luck trying to learn soime deeply nested webpack object structure by heart.
I don't even think that it's possible, and I fucking HATE configurations I can
t replicate without having to look up the documentation or copy from a previous object.

>>59609799
A million times this.

>>59609146
I said this in the thread you found this image in but you must have ignored it because you didn't want to pass up the opportunity to shoehorn your opinion into another thread and make yourself feel profound:

JavaScript is inherently a dinky piece of crap and people are building on top of it to solve the problems it has. This leads to a confusing mishmash of different tools doing different things on different levels and it might not be ideal but it's a step forward. Having poorly organized tools doesn't mean it's a stupid idea to have tools, I guess that's what you're supposed to be saying because your post is conveniently ambiguous enough to be exempt from saying anything concrete.

Hopefully it will be more cohesive in the future.

>>59611426
If refreshing the page sends you somewhere different, then your SPA is providing a worse user experience than NCSA Mosaic.

>>59611606
>Read extremely limited documentation
>Get weird bug
>Google around for an hour
>Finally run into a stackoverflow answer that goes "Oh the instructions sometimes don't work, try this <code>"
I spent 10 hours today getting my build tool to do something entirely trivial. It's really hard to express the amount of loathing I have for this retarded process.

>>59611663
I hope you're right, but I think we're too far down into webshit hell

>>59611535
every person i've seen come out of a bootcamp 6 months later has retained a mere fraction of what they were taught
i'm pretty sure these things are a scam, it's amazing how pricey they are.

>>59609928
>people that don't have anything to do with operating or maintaining engines getting upset that "their" title is being stolen
pottery

>>59611287
Thanks mate.

>>59611672
If the URL contains your whole app state you're probably 'developing' Wordpress templates

>>59611465
I just hope you realise this is done only when running in development.

File: froggy.jpg (16KB, 225x225px) Image search: [Google]

16KB, 225x225px

>mfw I'm exactly the type of brogrammer /g/ hates
I haven't written a program in my life yet my job title has "Engineer" in it :^)

>>59611891
me too senpai
I don't even have my B. Sc. in compsci yet and I'm already making some good cash with frontend memes :^)
probably more than the average /g/ NEET shitposter will ever archieve

>>59611705
Yeah. And in the end, it takes longer to build your stupid web project than it would take to compile some Million LOC C++ projects with a lot of template metaprogramming.

And then there's the scanario where your application stops being super trivial and you have an actual backend and you can no longer use webpack-dev-server. In fact, I can't imagine any situation where you don't have a backend. So for whom is webpack-dev-server even there?

>>59611968
? you're missing the point of webpack
webpack-dev-server is just a nice addon for a nice dev experience.
what you are actually using webpack for is bundling your app for different targets, most commonly bundling everything into a single file, minifying stuff (webpack is actually really smart about this, check out tree shaking), applying css-postprocessors etc.

>>59611723
I think we're finally just going to start anew. Webassembly and whatnot

Like we've known this forever but I think we're finally just gonna do it over

JavaScript is basically a monument to technology progressing faster than it can get a grapple on. Atleast we can learn from our mistakes and have some funny stories to tell our kids

>>59609850
or maybe he doesn't work for a company/client that still lives in the 00's

Enterprise Java is annoying, but hipster JS is worse.

>>59612002
I know that, but automatically realoading your web browser is an important part of the development experience and you can no longer do that if you have some faggit apache serving the page instead.

>>59612027
actually webpack exposes an API and you could run the whole HMR thing in your production server if you wanted to
but why would you? it's better to develop a feature locally using dev server, push it to master and hope your DevOps monkey set up CI so it gets deployed automatically

File: 1485072007979.png (596KB, 800x800px) Image search: [Google]

596KB, 800x800px

>>59611939
I know right? A bit of management and soft skills, a bit of frontend, a bit of automation, and you can earn decent money merely by walking around the office, attending meetings, and giggling with female colleagues like a retard.

>>59609799
I recently updated a customer website for bootstrap and modern CSS. The website is a search engine for auto parts, and guess, a lot of people in this industry still uses XP and even crt.
Since the site was updated we get 1-2 nails every day of some asshole complaint that he can't use the site on IE. One fucker sent a print screen on do and the asshole had chrome but was using the fuck ie.
Man just force people to update browsers. Really. Chrome runs on almost any os. There is no reason to stay locked because the shitty IE that never adhered to standards. IE is a fucking hell.

>>59609146
The purpose of Redux is to force developers to do things atomically as much as possible.

For small projects, it probably adds too much verbosity for it to be worth it, but for large projects, added verbosity is an acceptable trade-off for having easy-to-trace code execution path for debugging.

>>59609799
>Why not simply not support legacy browsers?

Because something you consider a legacy browser today is not IE7 vs 8, where you have 2 years and a full level of css standards compliance between versions. It's WebKit 537.11, vs 537.17 where you have 2 months between versions, and each version adds a new experimental feature while also fixing support for a 4 year old experimental feature that became a standard 2 years ago.

Today, two browsers with the same version number, user agent, platform, and everything else, can end up rendering a page in two different ways.

Microsoft and Mozilla had to add support for -webkit- prefixed CSS because they are so fucking widespread.

It is simply not possible to not support legacy browsers today, because there's no line between modern and legacy browsers.

>>59611050
If you need to do all that shit on a web page, then you are already doing something very, very wrong.

>>59612111
trips of truth.

>>59611287
>In modern web dev environments, you need a bunch of external tools to make you type as little HTML as possible.

In return, you end up typing a fuckload of extra JS, and the same amount of resources will get loaded anyway (in fact, probably more).

>>59612071
kek that part about giggling with female colleagues hit close to home
I'm literally the only dev here that gets along super well with the female designers.

Frontend is super versatile. Can do some design work if I feel like it, code some elegant abstractions, write tests, optimize build process, or just develop new components, it just never gets boring.
meanwhile the C guru gets a hardon because he could shave off a clock cycle in some for loop

>>59612073
>bootstrap
>modern CSS
that shit has been looking dated for a good 2 years now and putting in the time to make it look good/different would be better spent somewhere else desu

>>59612111
That is a good reason to stay the more you can true to html bare standards than JScript magic.
People have horrors thinking about reposting a page and how every shit has to be dynamic and imediate.
For a great majority of applications posting is better than infinite JScript msgic. Look big sites, Microsoft, Intel, Reddit. All employ posting while some retards insist in doing a full post less Ajax site that less than 500 people access per day.

>>59612126
>what is a SPA

>>59612174
Not a website...?

File: Screen Shot 2017-03-27 at 8.59.20 PM.png (84KB, 549x683px) Image search: [Google]

84KB, 549x683px

I don't get the whole "do CSS in JS" thing that's trendy right now.

https://github.com/cssinjs/jss

>>59612172
POST is just plain bad from a UI and UX standpoint (e.g. history)
The reason those huge ass sites don't use a more modern approach is a) to ensure compatibility across ancient devices and b) because they have such a large legacy feature base that migrating to a SPA approach would require a massive investment.
If you wanna do it right you can't just change shit here and there and add some jQuery pajeet tier code, you gotta develop it from the ground up.

>>59609146
most of that should be in devdependencies you noob

also these are just build tools, it doesn't bloat the actual script that's compiled.

all that babel shit just "compiles" react and allows you to import modules. all that sass shit just compiles sass into css, autoprefixer automatically adds browser prefixes etc.

>>59612241
Because CSS in JS doesn't make any sense at all.
HTML in JS àla JSX does because if you want to develop dynamic components you're always gonna need both.

CSS is not a solved problem by any means but I'm really liking the CSS modules approach.
It allows you to write 'local' CSS code that will make your stuff look the same no matter where you use it, pretty great stuff.
https://github.com/gajus/babel-plugin-react-css-modules

>>59612172
I honestly don't mind ajax submissions, it's pretty neat, and has its place. In fact most modern crap can be neat and have a place.

But, people who did not learn the basics end up making a navigation menu in 250kbyte of php + bootstrap + angular + 100kb css animation library, with two or three completely separate list elements for desktop/mobile/tablet designs, hardcoding all links one by one.

The single fucking line of JS you need for that is an onclick class toggle to trigger the menu between open and closed states. All the rest can be done easily via css. If you want more sub elements to have open triggres, then you need another class toggle for every child menu as well.

That's the shit I had to deal with the past week; I eventually figured that I couldn't be arsed to (nor had the extra day required to) rewrite it so I just changed two animations to hide the giant glaring design bugs in it.

>>59612294
I've been just writing one big CSS in Stylus and having Webpack inject it onto the page.

This modular approach is probably better for larger projects.

>>59612354
The cool thing is, you can still use CSS preprocessors if you want to. Just need to add the appropriate loader to your webpack config.
For my needs, I'm totally fine with not using a preprocessor with CSS modules, because they allow variables/imports natively and an autoprefixer in webpack takes care of the stuff I'd previously use SCSS mixins for.
But to each his own, I love that we have so many options.

File: file.png (78KB, 464x803px) Image search: [Google]

78KB, 464x803px

>>59609146
looks like my current project

>>59613093
how many megabytes do you load per refresh?

>>59613185
Do you mean gigabytes?

File: file.png (30KB, 582x243px) Image search: [Google]

30KB, 582x243px

>>59613185
>>59613202

>>59613247
nigga. you need a manifest chunk so that your vendor bundle only changes when you add a new dependency and not on every fucking build

>>59613247
>>59613381
even looked up the docs again for you senpai
https://webpack.js.org/guides/code-splitting-libraries/#manifest-file
step up ur game thx

>>59609850
This is wht i do and i get paid 180k/yr as a sr react dev at the age of 23 without a degree,

>>59613442
How does it feel to be a one trick pony?

>>59610861
this

>>59613442
lol no.

>>59613569
>become react dev
>all languages become lock, cannot learn all skillpoints used

Does g honestly know anything about coding besides knowing how to fizzbuzz in punchhole cards

>>59613721
Becoming a full stack webdev is basically giving yourself mental retardation.

>>59613721
I make $300k/yr as a software architect.

This is how e-penis comparing works, right? Because I would be winning.

>>59611771
You can say that about cs majors too lmao
so much useless filler in a cs curriculum.

>calls himself an engineer
>doesn't even know basic college level physics
why is this allowed

>>59613790
Post a timestamped photo of your w2.

>>59613809
explain why you need physics to write software without trying to justify your expensive degree

>>59610861
why not?

>>59613790
U ddnt post ur age or ur degree

man idk how you JS people do it. This looks like such overengineered garbage, but at the same time it looks like the desolate wild west. Some new sheriff (framework) comes into town and you gotta roll with it and end up in the back of the saloon getting tied to a horse.

nobody would ever start programming if this was what it entailed. Fucking webpack?? Manifests??
And this isn't even the dynamic page shit like python and php is it? All for some shit crud app

stop doing web development, it's not real programming... take my advice and just stop now, you will thank me later

>>59609146
suggestions?

>>59613790
i make about $110K as an architect doing godless java EE. we average about 100 mill transactions through our system a month. that seems cool to me. i'm barely following this thread. front end seems painful. we don't have women designers to flirt with but i get to develop substance abuse problems with my fellow backend graybeards.

>>59609799

Yeah, I'll just drop support for 95% of my customers, this will work out well.

>How the fuck did this happen?
When you tell everybody they can code, they get into webdev because they "used to do their own CSS for myspace xD!"
Then they realize they don't know how to do shit, but this library does it for them.
Repeat that about 15x.

Stack overflow writes 90% of web application code nowadays.

>>59614738
Only 15 times?

>>59612111
How are the version numbers of each different browser even related?

File: 1460718961184.png (465KB, 497x544px) Image search: [Google]

465KB, 497x544px

>>59609418
literally retarded, those dependencies don't go into the final output

you don't include your compiler into your programs, do you?

>>59609901
then have them use two browsers.

This entire thread is reeeeeeeeeeeeeeeeeeeeeeeeeee


I spent four years getting a degree and some slut bimbo went to a camp for three months and now makes the same as me.

>>59612022
Enterprise Java is (slowly) getting better.
JS is not, unfortunately.

>>59615089
Well most people start out with just a simple static webpage

>web devs
I wonder what they actually look like while they're "working." I always imagine them blankly staring at monitors while they do the 21st century equivalent of an ape pounding on a typewriter.

>>59609799
>Why not simply not support legacy browsers?
Because you don't want to get fired.

>>59609146
>left-pad

>>59611595
>catalyze their workflow

>>59616174
k

>>59613442
why you lyin bro

This is the current state of sane backend web development

package main

import (
    "io"
    "net/http"
)

func index(w http.ResponseWriter, r *http.Request) {
   io.WriteString(w, "Hello world!")
}

func main() {
    http.HandleFunc("/", index)
    http.ListenAndServe(":80", nil)
}

I feel for frontend guys though.

>>59616113
is he OK?

>>59616546
>golang

File: frontend.jpg (126KB, 700x918px) Image search: [Google]

126KB, 700x918px

>>59611040
>>59611287

pic probably related

File: life-of-the-buddha-30.jpg (85KB, 600x718px) Image search: [Google]

85KB, 600x718px

>>59609645
Webdev is just as the whole world right now : a giant clusterfuck.
Nobody really knows what he is doing, but everybody tries his best.
Normies web dev want to be competitive and stay at the edge, whereas i will stuck to the goold old LAMP stack cause it is stable and reliable as fuck.

>>59609146

Web development looks so messy...
And I don't even know how it works. Do web devs use compilers? Or everything are interpreted scripts?

There also seem to be way too many frameworks, languages,...

>>59616611
It's KISS, it's fast, it's supported. What's not to love?
>muh generics
>muh static linking
>muh gc

>>59616735
those autistic generics user should leave their previous PL

gc is negligible with the latest version and improved a lot

>>59609146
This is exactly the reason I'm trying to avoid anything closely web related and why I hate the everliving fuck out of my job

File: 1473323718571.gif (3MB, 420x300px) Image search: [Google]

3MB, 420x300px

>It's another "every library sucks, every website should be served as plain HTML with minimal CSS" thread

/g/ is laughable.

>>59617329
The internet is laughable.

>>59617329
and there would be nothing wrong with
pretty graphics and muh animations are for sheep, UI design is nothing but psychology applied to software

>>59609146
>no redux
>no react router
kek

>>59612153
>>bootstrap
>>modern CSS
so what would you use instead of bootstrap then?

>>59614080
>take my advice

What advice? you offered no alternative, you just said stop...

Here's a fun JS exercise: create a way to have pseudo-fullscreen support on iPhones.
Found one library that was made for this, Brim, but it used loads of hardcoded sizes to detect if the addresbar was hidden, and didn't work on iOS Chrome.

Had to write that shit from scratch, took me 4 days to get something stable

>>59617721
For layout I'd just use flexbox ain't nobody need a grid system anymore
For styling the world is yours. Some material design based meme or just roll your own
Bootstrap just screams low effort these days desu

>>59617395
>UI design is nothing but psychology applied to software
No, it's a science of class theory and visual hierarchy

You don't keep your shoes in the fridge

>>59610898
>what is angular 1
>jquery
>ember
>vue
Etc

Anyway I can see the need for a transpiler with jsx, but people need to fuck off transpiling simple node apps. Node is only really missing es import/export which is pretty much irrelevant unless a package doesn't provide commonjs.

And anyone that can't grok a JavaScript setup probably shouldn't be touching code. Please kill yourselves

>>59618382
>>You don't keep your shoes in the fridge

Actually people do that. It stops them from smelling.
https://lifehacker.com/5562291/throw-smelly-shoes-in-the-freezer-to-freshen-them

>>59618382
>science
lmao

>>59618325
>For layout I'd just use flexbox
flexbox a shit on iOS tho

>>59613790
Damn nice, I only make 250k.

>>59616735
>less powerful than C
>less safe than C
>still manages to have a gc, compile times, and no modern features

>>59609146
>brianlet can't understand FP ui design

just go back to your php MVC kido.

>>59618758
1. Define "powerful"
2. Hahahaha
3. A FAST gc, VERY FAST compile times, one of the best concurrency model

>>59618790
Gotards everybody.

>>59609146

bootstrap & jquery is all you'll ever need.

>>59618790
>A FAST gc

a gc no matter how fast will make it irrelevant for systems programming, for web dev where you are IO bound you have better languages. whats the use case for golang again?

>one of the best concurrency model

its one step up from java primitives but its still garbage.

http://www.jtolds.com/writing/2016/03/go-channels-are-bad-and-you-should-feel-bad/

>>59612174

Single Page application.


A website design that updates parts of a single view rather than load and transition to static pages.

File: this-guy-fucks.png (452KB, 538x560px) Image search: [Google]

452KB, 538x560px

>>59613202

>>59618013
That doesn't sound fun at all. Why are you lying?

>>59609645
No it's fucked, but if can master it and use it within IT outside of dev departments you will ascend.

I used to be a.low paid Linux admin making $50k. I knew programming but rarely got to use it. I learned web development and a lot more programming, config management, continuous integration/delivery, etc. Now I make $130k in Midwest money as a Devops engineer.

>>59611465
>ruby
>empty project
Try rails project with a bunch of random shit installed including an entire administrator engine and running in development mode.

>>59609799
This is something I don't understand.

>you need to support old shit because of companies

This argument is stupid to me. It cost nothing to people to upgrade to a better web browser

>they are using windows xp
This is a shitty argument. In this days, using old machines and OSes that are probably slow as shit makes the people less productive. It would cost less to a company to have people working on decent machines, with things up to date than forcing them to use old and slow shit.

>>59618948
Before I had to write that shit, I hated apple stuff because its so expensive.
Now I hate apple stuff because the damn niggers aint even got support for the fullscreen API

>computers get stronger
>people suddenly start using more and more resources because "who cares lmao more resources we are in the future now"
what a terrible time to be alive

>>59616113
You mean "they shouldn't".
And no, in fact I would fucking expect a fucking compiled runtime, not the whole goddamn cock cunt fucking shit show.
But that's not happening, because edgy devs have to show of their pecs with server side ES6 full compliant cunt crap shit goddman code.

Whoever calls web design as well engineered is a faggot
t CompEngi student

>>59609146
it's honestly like node has turned developers fucking retarded.

File: 1489971525698.gif (4MB, 280x302px) Image search: [Google]

4MB, 280x302px

>>59609305
>So you need a bunch of polyfill libraries to support legacy browers

>Implying modern webdevs actually test in anything outside chrome and say "it works in chrome" when told it doesn't work outside of chrome.

Most of these dependencies are actually devDependencies and should not be included into dependencies. The tweet is from a retard

>>59611465
> programming language is even worse

Are you retarded?

>>59611287
There's nothing worse then a person that thinks they know what they're talking about. I could correct you but it's pointless, you've already demonstrated that you lack the capacity to understand.

>>59611465
Rails doesn't come with jquery-ui or active_admin by default. Are you shocked that you can install shitty third party packages?

Those files would also be getting concatenated and compressed if you were running in production mode.

>>59611525
The retard installed a bunch of third party packages and is running in development mode. In production mode they would be concatenated into a single file and compressed.

They aren't concatenated in development mode because it makes debugging far easier.

>>59613790
That must suck living in poverty. I'd kill myself if I made less than a million a year.

>>59616185
> take 20 years to figure out XML isn't the ideal way to code
> it will take another 20 to realize annotations aren't much better

I really hate the Java community. There is no community that is even half as bad.

>>59609146
never liked using other people's libraries
I usually make the stuff I need myself

>>59623364
But source maps don't care if it's a single file or 20, no?

>>59623505
>I usually make the stuff I need myself

I guarantee the stuff you make is far worse than all of the popular libraries.

>>59623553
YES THAT'S RIGHT NEVER ROLL YOUR OWN LIBRARIES NEVER REINVENT THE WHEEL OR MAKE A BETTER WHEEL
ALWAYS IMITATE NEVER CREATE

i fucking hate nu-coders.

>>59623536
Source maps aren't as precise and aren't always completely accurate. There's really no good reason to concat and rely on source maps during development.

>>59623572
Nice strawman idiot. People that roll better libraries have almost always used the popular libraries extensively. The reason they can create better ones is that they're intimately familiar with the flaws of the previous attempt.

The retard that just says "never liked using other people's libraries" is almost guaranteed to make complete fucking garbage.

>>59623572
Right, because the problem with JS today is too few people trying to reinvent the wheel.
Are you mentally retarded by any chance?

File: 1484047239409s.jpg (5KB, 250x178px) Image search: [Google]

5KB, 250x178px

Non web-dev can fuck off to the hole they crawl from

>>59624167
Frankly, COBOL is starting to look better than all of webshit.

>>59624274
Frankly you can fuck off there then

>>59624292
Did I hurt your webshit feelings?

>>59609799
supporting legacy browsers is fucking cancer, go fuck yourself

Js was a mistake. The DOM was a mistake.

We should go back to static HTML and CSS or redo everything in lua/scheme

>>59625457
We should redo everything in a purely data-oriented way. The browsers should decide how to render, what to render, and generally how to partition the entire application. The application should be pure data.

>>59623572
>YES THAT'S RIGHT NEVER ROLL YOUR OWN LIBRARIES NEVER REINVENT THE WHEEL OR MAKE A BETTER WHEEL
>ALWAYS IMITATE NEVER CREATE
ah yes, exactly what we need, more framemeworks!

>>59609305
how about:
>If it works in ie4.0 it's good enough

many such cases, so SAD

>>59623288

try to improve your shitposting next time

>>59609146
You scared little dinosaur. Fuck off and leave this industry to us.

>>59611535
>>59609928
You hate it because you feel threatened. You are pitiful.

ok, I get it, web development is a mess

how can I make a webpage with simple elegant code? tools? framework?

>>59626573
>simple elegant code?
You write it in plain HTML/CSS with no javascript.

>>59623101
IE users can fuck off desu.

>>59626732

it won't be interactive and how could you use it for example in a shopping cart webpage?

saving sessions, SPA?

>>59626832
You only asked for elegance.

I write html templates and markdown articles and then use make and sed to render the static website. I use Sass and TypeScript though so I may not qualify for the absolute minimum.

Next project is a backend crud framework in Go with vanillaJS Ajax calls to build SPA's.

Maybe in a year or 5 I'll understand what the things in this thread mean.

Life is comfy

>>59626832
> it won't be interactive and how could you use it for example in a shopping cart webpage?
We've done that with sessions for over a century.

>>59617721
>so what would you use instead of bootstrap then?

Notepad and write your own css? It takes 1 minute to get up a basic layout, and after that you just add stuff as you need them.

>>59626832
Remember when people wrote simple CGI scripts and maintained state through auth cookies, logins and databases?

Those were the days.

>>59626875
>>59626910
>>59626940

that means that people have been using bloated, unnecessary features for most tasks?

>>59627820
YESSSS

nobody ever needed SPAs

>>59626940
mm yeah cuz .net webforms were so great fagit
react is gonna own ur ass soon so get with it

>>59612246
What are you on about? Post is the only thing that maintains history properly. Good luck using the back button on your spa.

im a vanilla js programmer

feels bretty good. I dont even know what dependencies are lol

>>59628355
Post doesn't maintain history. Get does.

>>59612172
>That is a good reason to stay the more you can true to html bare standards than JScript magic.
And yet Microsoft ""Edge"" doesn't even support basic HTML <details>...

>>59609928
T H I S
H
I
S

>>59611789
either server or client statesave should still be referenced in URI changes

be caferul not to develop a facebook-timeline experience

>>59611891
>>59611939
hey, fellow pajeet fake-malefeminism blm brogrammer here, a computer scientist could do my job faster, costing the company less, or even pay them a bigger salary.. but thats no excuse for me to vomit garbage code, i do my best study a lot on my free time and stitch my libraries in the most sane and less bloated way as possible

>>59623101
>tfw I do this

>>59613870
>why you need physics to write software
not that anon, but i guess he was ferering to 'engineer' term, not 'software engineer' which is just a fancy term for 'programmer', seeing as its no engineering

>>59619251
>windows xp... In this days, using old machines and OSes that are probably slow as shit
baby born with an iphone on his hands thinks humans in the ancient year if 2001AD you had to wait a few seconds for each keypress to reflect a change on a black&green screen , computers couldn't possibly be useful with only 500MHz right??????????
midori works just fine,
in fact only now firefox is distancing itself from windows xp and vista

>>59612073
so its not a hotsite advert campaign page, yet you cant have a simple string search result page for mechanics that just want the parts and couldnt careless about your material design, with a noscript fallpage functionality, just because you dont want anyone looking at it with a UI that doesn't match your 'vision'

>>59623439
I know! Ever since I've been raking in at least 1 million a month, I'd be killing myself if I ever made less than this again

are websockets inherently trustworthy?

I have a section of my website that interacts with users through websockets after they have been verified and the session is managed through express-session. Could one user potentially hijack another user's websocket if they disconnect? Anyone have any strategies for websocket "verification"? Should I just ask for cookie information through the websocket to verify the correct user is in the websocket session?

>>59632931
Use a meaningful session token that has information about the user's machine for verification.

>>59609799
Sure, I'll just tell my customers to go fuck themselves
Brilliant!

>>59632290
Do you really think people in corporate world use Midori? I'll let you in on a secret now, they don't. They use whatever was packaged with Windows XP, and that's Internet Explorer. Not updated.

We had an old guy who was like 80, been working in that firm forever and he just wouldn't switch from Windows XP and Internet Explorer. Not to mention the clients, everything had to be tested on Internet Explorer.

>>59609146
web designers aren't programmers
web programmers aren't programmers

>>59633070
If we all universally agreed to tell them to go fuck themselves, then they would have no one to turn to and they would have to update.
Shame that's not possible.

>>59611771
would consider that anyone who has to put themselves into a boot camp to get external motivation to "get skill" isn't motivated at all

>>59633250
'boot' camp .. make it as unpleasant as possible . wow suddenly you are out of it, with a scrambled brain & 'total aversion to pain' .. you would hate programming for life after going through a boot camp

>>59633271
to prepare for the utter shit working in industry is?
nothing is worse than i.t. industry. It cannot be simulated. It's not about the technology, it's about the personalities

>>59633205
>corporate world use Midori?
uh okay then, seeing as you argued "companies wont upgrade" argument is silly, i infered you were talking about Home users and believing they were locked out of browsers released since...idk +- 2014, around the time of both html5 ascension and XP deprecation

thing is other than IE6 quirks mode and texts outside of their squares that could make shit unusable, most sites blocking ie6 were just doing it because it would look fucked up, but still functional.

if your site need to keep legacy support for a few users its silly to worry about a pretty UI, and in such cases it would also make sense to have noscript functionality as well.... so might as well serve the noscript one to ie6 if the enterprise JS aplication is too convoluted with canvas webrtc.

>>59610898
>vanilla JS library
>>59626888
>framework in Go with vanillaJS
>>59628453
>vanilla js programmer

at first i thought it meant "regular old plain JS", but then i thought "wait, what if they are talking about some hipster that actually called his new framework VANILLAJS"

and then http://vanilla-js.com/
>Vanilla JS is a fast, lightweight, cross-platform framework for building incredible, powerful JavaScript applications.

REEEEEEEEEEEEEEEEEEEE

>>59623288

This.
What an idiot.


>>59609146

>this is all considered serious and well engineered code design

Consider this:
You're entering a new field for the first time. Let's say chemistry.
All those weird glasses, why not just using one type of glass?
So many differnt lights and fluids, why making it so complicated?
All those huge ass formulas - ZOMG they must be kidding?

What I'm trying to say is, that it's pretty easy to critisize somethign if we don't care why things are done that way.
Web development has come a long way in only three decades, the "PHP 960 grid + psd" workflow is lightyears from responsive webdesign and single page apps.

Still there's a lot of work to do and yes, the internet carries the legacy of this evolution. But the frameworks get better every year, and the last years some really decent frameworks have been established. It's getting better all the time, but it is a long way too perfection..

>>59616556
It's k, she's immortal.

it's okay guys web assembly will fix everything

>>59609928
t. Rejeesh

>>59634571

>muh latest meme

It surely will, anon.

>>59634447
Do you only read the first sentence on any website you visit?
You got meme'd

>>59609200
bloated maintainable shit

>>59609418
no, just use typescript

>>59634571
>web assembly

That's been in development for what, 10 years now?

Literal vaporware.

>>59634715
The first version was literally just finalized this month you fucking retard

>>59612111
Then why rely on xperimental features?

>>59613809
>calls himself an engineer
>only knows how to operate a train
fucking bulshit

>>59609146
I tried to add React to a Node project that I am doing, but I couldn't figure out how to set it up. :(

>>59634630
i thought i would spare the spoiler and let others have a chuckle reading that page

i just wonder if they bothered to run that speed comparison or if its just made up

>>59634840
Nah, it's skewed though. The other libraries use the same function for the end result but pre- and postprocess the in- and output, and do other things on the side. Eg the Jquery selector:

 * Attribute not equal selector
 * Positional selectors (:first; :eq(n); :odd; etc.)
 * Type selectors (:input; :checkbox; :button; etc.)
 * State-based selectors (:animated; :visible; :hidden; etc.)
 * :has(selector)
 * :not(complex selector)
 * custom selectors via Sizzle extensions
 * Leading combinators (e.g., $collection.find("> *"))
 * Reliable functionality on XML fragments
 * Requiring all parts of a selector to match elements under context
 *   (e.g., $div.find("div > *") now matches children of $div)
 * Matching against non-elements
 * Reliable sorting of disconnected nodes

It's the age old flexibility vs. overhead tradeoff.

>>59634518
>why things are done that way.
>this convoluted buggy mess is necessary because we're forced to use html, css and javascript which were originally intended for documents, not applications
>in essence, we do it for backwards compatibility reasons
>we don't support old browsers tho, we need the latest flexbox & shit
the why the fuck not throw this garbage out and start fresh with a saner foundation?
>soon we have webassembly so we can add another layer of interpretation to what will basically amount to a return of the java applets and flash
y'all must be crazy

>>59616155
> then let them eat cake

>>59628461
That's why you always redirect client to GET after they POST something.

>>59616696
>Stable
>LAMP

How I knew you were retarded

>>59635738

The distinction of HTML, CSS and JS is clever. have you never heard of "seperation of concerns"?

HTML is WHAT graphical elements you see.
CSS is HOW you see them.
JavaScript is the logic behind it (at least the client side logic).

Of course it's a little bit confusing in the beginning, but basically there are only two approaches:
-server does the work and delivers the result, while client (JavaScipt) only does some calculations for stuff like moving elements
-server only gives endpoints with some bare data coming out while the client (often single page apps) do most of the work


A lot of newbies get intimidated by the sheer number of JS frameworks, but you have to understand that they are OPTIONS and POSSIBILITIES, not strict rules.

>hmm, I'd like to have virtual DOM manipulation, but React seems overkill..
>cool, I'll just use Vue!

With jQuery and Bootstrap you are already fine for 70% of all websites out there. And you can learn both in a few days.

So why are you complaining about options? web developers just like to try out new things. Making the same thing over and over again is boring. Trying out new things is fun. It's taht simple.

>>59636097
>HTML is WHAT graphical elements you see.
it's markup, you fucking dufus, aka. text formating
>CSS is HOW you see them
it's styling for text
>JavaScript is the logic
it's a clusterfuck
>OPTIONS and POSSIBILITIES, not strict rules
codewords for "we don't know what the fuck we're doing"
>web monkeys just like to try out new things
ftfy

>>59636165

>it's markup, you fucking dufus, aka. text formating

No.
Like I said, HTML is the markup of WHAT you see (and not only text, btw):
"<p>" means here starts a paragraph, "<a>" means here starts a hyperlink, "<video>" means now comes a video and so on.

It's still the job of the CSS to decide how much spacing, which colors and so on - the "HOW do I see it?"


>it's a clusterfuck

It's no secret that JS is not the most beautiful language out there, but the latest version got much better. Also nothing keeps you from using a different syntax, you just just have to include a single library and you can write C# (i.e. Script#) or OCaml (Ocamljs) or Haskell (ghcjs) or ..whatever.

You could also try out one of the cool front end languages like TypeScript or Elm.


So much toys to play with. Don't be a grayface.

>>59636320
>Like I said
yes, you made it clear you're retarded, no point in proving it further

>>59636336

Nice arguments you proivded there, buddy.

>>59609928
t.pajeet

>>59609146
>npm install meme-of-the-week

File: 1488926451258.png (30KB, 176x184px) Image search: [Google]

30KB, 176x184px

>tfw use plain text files and awk to generate my site

>>59639289
>not using make and m4 macros

>>59639364
make triggers me because Makefile has a capital letter.

>>59639388

alias make='make -f makefile'

>>59639400
>adding several bytes to my perfect .shrc

>>59639289
>use css file for look of the website
>use js file for functionality of the website
>basic index.html

I don't understand why people need all this plugin shit and megabytes of frameworks and libraries and what not

File: 1345999339976.png (35KB, 471x385px) Image search: [Google]

35KB, 471x385px

>>59639388
All my filenames have capital letters because I use Windows, not some shitty case sensitive OS.

>>59639502
it's not perfect without

[ -f .config/alias ] && . .config/alias

File: 1483225546294.gif (315KB, 263x271px) Image search: [Google]

315KB, 263x271px

>>59639510
If you have a big site, say a blog with a few hundred posts, it gets to be a pain. Still, you could conceivably do

cat * > index.html

instead of invoking the four horsemen of the apocalypse with several megabytes of frameworks. Even markdown.pl would be preferable to this shit.

>>59639557
>having to press shift to type any filename
Wasted keystroke.

File: 1350594293765.jpg (111KB, 500x500px) Image search: [Google]

111KB, 500x500px

I use PHP to generate a front-end with only jQuery as non-local dependency.

Is this considered good or bad?

I use web technologies for all of my guis

Am I cancer?

File: 1483224913330.jpg (114KB, 970x880px) Image search: [Google]

114KB, 970x880px

>>59639560
Nice bashisms nerd.

if [ -f "$HOME/.config/alias" ]
do
    source "$HOME/.config/alias"
fi

Now you are POSIX compliant.

>>59639657
Neither "." nor "&&" is a bashism

>>59639692

[ -f .config/alias ]

doesn't evaluate to anything but a syntax error in a correctly implemented shell. `.' is a matter of taste.

>>59639743
What?
"[" is a program not a shell command (in the standard shell - bash and others implement it as a shell command for performance reasons) and it will give you the appropriate return code.

>>59639782
>bashfags
The problem is the unquoted string.

>>59609146
I don't know any code language.

What is wron giwht your picture?

>>59639818
It's normal command-line program.
The "unquoted string" is a command line argument and does not need to be quoted if it doesn't contain a space or special characters.

Also, I'm not a bash user.

>>59618880
Green text questions are rhetorical.

>>59639861
>and does not need to be quoted if it doesn't contain a space or special characters.
That's wrong though, the string

.config/alias

is expanded by the shell to the full path.

>>59639899
What?
No it isn't.

File: 1483224708210.jpg (58KB, 778x720px) Image search: [Google]

58KB, 778x720px

>>59639921
The shell expands ".config/alias" to '/home/you/.config/alias', that's literally how it works. This is before we get to your assumption that the current working directory is necessarily the home directory.

>>59639995
It's a normal command line argument, without anything that has special meaning to the shell.
The shell doesn't do shit with it, if it would you'd get

% /bin/echo .config/alias
/home/you/.config/alias

And yes I forgot $HOME in the original but that got shit to do with bashisms.

>>59639824
imagine renting a garage full of tools and shit but you just have to inflate a tire

Is gulp/grunt now obsolete?

>>59635373
>The other libraries use the same function for the end result but pre- and postprocess the in- and output
well thats a given, php is c wrapper, ruby c++ wrapper, python idk, btw i skimmed thru some comparissons and it seems some jquery functions are only 3x slower, but in the end, with lots of self wrapping and library over library you get the mess that is the web today with simple ajax forms talking up surreal cpu time on modern laptops.

isnt there a way to write less but instead of ending up with a convoluted multi level wrapping of JS, you pre macro it all to vanilla js? why isnt anyone working on something like that? even google has their own JS framework instead of making some new kind of language that renders efficient vanilla js without making it more complex to design aor write like it was jquery or angularjs, althewhile making your webapp faster or allowing it to be more complex while still running on most users computers

fuck frontend looks like such a mess, idk if im either too dumb to understand the reasoning behind the tools creators, or too smart to adopt them, heheh

File: 1487362300114.jpg (92KB, 795x1204px) Image search: [Google]

92KB, 795x1204px

>>59609799
Alternative idea: Only support legacy browsers.

Do the layout in frames or don't do it at all motherfucker.

>>59612246

this >>59635773
its like bootcampers dont even know how shit was done before, and that would explain why every 2 months a new convoluted paradigm comes up and js stitchers jump to it.

>>59639642
>all of my guis
as long as theres an accompanying backend binary on lowlevel or at least something that is not JS that servers websocket or http, and the browser really just runs the GUI and not any deep logic, its not cancer, just some mild STD,

but only if your backend compiles to nearly all existing platforms that run browsers capable of some minimal JS (or newer websockets browsers), or else you might just have gone with some GUI framework that is only cross for your targeted kinds of computers

anyone have a torrent for algorithms 2 by princeton that used to be on coursera?

>>59641615
I think jQuery was developed mostly for dev convenience.

>isnt there a way to write less but instead of ending up with a convoluted multi level wrapping of JS, you pre macro it all to vanilla js?
TypeScript was designed with what you are describing in mind. I haven't really used it so I can't say if it's been really successful at it.

Mind you that frontend has been developing at a ridiculous pace which means that at some point shitty 'just werks' solutions with awful design were the best one could get.
The shitty part is that some developers have started defending those.

Mozilla is doing it's best to provide better alternatives btw, like the Fetch API for Ajax calls and Template literals