The JavaScript Juggernaut Rolls On

JThaddeus writes "An article in TechWorld Australia summarizes the latest opinions on JavaScript from ThoughtWorks: 'There is no end in sight to the rise of JavaScript... "I think JavaScript has been seen as a serious language for the last two or three years; I think now increasingly we're seeing JavaScript as a platform," said Sam Newman, ThoughtWorks' Global Innovation Lead.' The article touches on new additions to JavaScript tools, techniques, and languages built on JavaScript. As the fuller report (PDF) says, 'The ecosystem around JavaScript as a serious application platform continues to evolve. Many interesting new tools for testing, building, and managing dependencies in both server- and client-side JavaScript applications have emerged recently.'"

Comments

SubjectsInCommentsAreStupid

[lesincompetent]

Great... more single-core-hogging fun ahead...

Re:SubjectsInCommentsAreStupid

[Anonymous Coward]

Yeah I hate the fact that none of the Javascript implementations actually scale past one core, and this is a serious problem. Yes there is the ability to use threads, but it's so useless since it's equal to forking a new process that has no access to anything of the parent process.

Same with WebGL, it's not going anywhere, anything that can be done with WebGL will produce something barely any better than a DS/N64/Playstation 1 game. This is not a problem with the hardware, but the fact that there's way too many cruft layers:
Hardware
Driver
OS
Browser
Javascript
half dozen Jquery plugins or other framework crap
then the actual app/game.

If there was a way to write Javascript directly against the hardware, then it might be a viable use as a game platform, but that's generally not the issue. The issue is that the raw calls needed to manipulate the GPU are going to be just too slow to ever use. This is why the majority of web and flash games stick to making ripoffs of NES and SNES games just with higher resolutions. They can't make a Duke3D or a DOOM clone because the underlying cruft won't let them do clever things with hardware that 3D games require.

Basically the best you can hope for, game-wise is Late SNES or PS1 (Think FF8 or FF9.) Applications that do not invoke sound may be able to get away with doing more (eg a HTML5 spreadsheet app would be slower than a native app, but 90% of that slowness will be the network transfers and not the browser.)

With 4 and 8 core computers, I'm just baffled at why anyone would want to use Javascript to create a game beyond using it as a scripting engine. Build a game engine that already knows how to read 3D models and textures, and just expose the game engine's state through javascript, not try to render the game state itself.

Re:SubjectsInCommentsAreStupid

[beelsebob]

Yeah I hate the fact that none of the Javascript implementations actually scale past one core, and this is a serious problem. Yes there is the ability to use threads, but it's so useless since it's equal to forking a new process that has no access to anything of the parent process.

Good!

Anyone with even 5 minutes of experience writing multithreaded code will realise that no shared memory, and message passing is far superior as a method of thread synchronisation.

Re: SubjectsInCommentsAreStupid

[Anonymous Coward]

Experienced engineers wil indeed be able to properly use thread-shared state. Mailboxes are inefficient in quite a few scenarios.

Re:SubjectsInCommentsAreStupid

[MillionthMonkey]

Anyone with even 5 minutes of experience writing multithreaded code will realise that no shared memory, and message passing is far superior as a method of thread synchronisation.

Anyone with even 5 days of experience will come up against a task where it isn't an option.

Re: SubjectsInCommentsAreStupid

[Joce640k]

Experienced engineers wil indeed be able to properly use thread-shared state.

True, but we're talking about Javascript programmers.

Re: SubjectsInCommentsAreStupid

[Pino Grigio]

Nobody really talks about false sharing and fine grained control of cache lines that you really need to make multicore actually perform in the kind of scenarios a game makes use of, most of which are totally beyond the scope of a Javascript app anyway.

Re:SubjectsInCommentsAreStupid

[phaggood]

http://nodejs.org/docs/latest/...

Re:SubjectsInCommentsAreStupid

[Anonymous Coward]

They can't make a Duke3D or a DOOM clone because the underlying cruft won't let them do clever things with hardware that 3D games require.

You're aware that they used Emscripten with Asm.js to port the Unreal engine to the browser, right?

Granted, this will never be as tight as native code, but it is clearly possible to do real-time 3D games in a browser. It may be lower resolution and/or lower frame rates, but using nothing but JavaScript it’s still more capable than the original Doom. Take a look at BananaBread, the first browser-based FPS.

Peak "platform"

[PCM2]

Is the word "platform" officially over? My fucking toaster is a bread-browning platform.

Re:Peak "platform"

[wonkey_monkey]

Would you like some toast?

Re:Peak "platform"

[Lumpio-]

Not as bad a "solution". I've seen knives marketed as "cutting solutions".

Re:Peak "platform"

[hey!]

And once we've done in "platform", we can start the campaign against "system". By the time we're done engineers will only be able to talk about software using mime.

Re:Peak "platform"

[swb]

So which is the parent? Do we have solution platforms or platform solutions?

Or do we have nested combinations, like platform solution platforms?

Re:Peak "platform"

[Lumpio-]

We could also add "service". SPaaS - Solution Platform as a Service.

Re:Peak "platform"

[ArcadeMan]

Count yourself lucky. My toaster is a bread-burning platform.

Re:Peak "platform"

[swb]

...platform solution..

Re:Peak "platform"

[ArcadeMan]

I'm launching a new company that will offer an operating system, platform-based solution services hosted in the cloud. Our first server will have a capacity of 1000 Megabitcoins per Doge/Watts.

Re:Peak "platform"

[Bing Tsher E]

We have platform precipitates.

And precipitate platforms.

Re:Peak "platform"

[Anonymous Coward]

My fucking toaster is a bread-browning platform.

Is it RESTful? Are you naming your upgrade project Butter? Care to share the private API? Can I fork it? Your propriety breakfast support system is harming OSS, you insensitive clod!

Re: Peak "platform"

[cyber-vandal]

COBOL is a far better language than JS. JS is more likely to be the next VB6.

Re:Peak "platform"

[Anonymous Coward]

We used to have a toaster, but after he stopped standing up he was just a drunk.

Re:Peak "platform"

[Anonymous Coward]

By the time we're done engineers will only be able to talk about software using mime.

Some of the engineers I work with already use hand gestures when communicating with me about my projects. Are we calling that "mime" now? And would that be a "mime system" or a "mime platform"?

Re:Peak "platform"

[Anonymous Coward]

That's not a toaster, that's the stove top. Of course you're burning the bread, the stove top lacks a mechanism to flip the toast or eject it.

Re:Peak "platform"

[SunTzuWarmaster]

Words have meaning and purpose. What words would you use to express the following concepts?

  - a collection of tools which allow you to build a new component through their leverage, while not contributing significantly to the overall effectiveness of the tools (or won't particularly be used in operation)

  - a collection of functional components which you will use as part of the operation of a new component

Currently, the words are "platform" and "system". I'm happy to switch to other words if they express the concept better.

Re:Peak "platform"

[Anonymous Coward]

The browser's typically the platform people refer to when talking JavaScript. Just as the C# language is not referred to as the platform, but .NET is (i.e. the CLR) in that tool chain. I wouldn't consider the Java language the platform, but the Java VM is. If you're going to describe every tool in the system as a platform, the word loses a lot of meaning.

Re:Peak "platform"

[Anonymous Coward]

I think we're talking about MAsS, Miming As a Service

Re:Peak "platform"

[S.O.B.]

You guys don't know anything.

Clearly Javascript is a super-tangential component of the solution platform meta-space.

Re:Peak "platform"

[Anonymous Coward]

> Clearly Javascript is a super-tangential component of the solution platform meta-space.

Yes, but will it help me leverage my synergies in the cloud?

Re:Peak "platform"

[Anonymous Coward]

My colon is a FoodBrowner<T> where T : Food. (C# syntax, just because it's pretty damned concise for this type of thing.)

Re:Peak "platform"

[hicksw]

...My fucking toaster is a bread-browning platform....

I use my other toaster for that.
--
And, please STOP thinking about the children!

Re:Peak "platform"

[Anonymous Coward]

Sounds great, I'll switch over to "toolchain" for a collection of tools, and use platform as the "toolchain+otherStuff" version. Thanks.

Re:Peak "platform"

[Anonymous Coward]

Exactly - if it doesn't have built-in support for talking to the DOM, I would say it's not truly targeted to the browser platform.

Re:Peak "platform"

[cwsumner]

Different specialties have different dialects of "technish", If you want to say something that someone will understand, then you need to find out what dialect they talk. And translate.

We need Google Translate to work on Javascript to SQL to C++ to Clarion !
( to Salesman to Manager to Politian !!) 8-)

Bring on Blink

[Anonymous Coward]

Isn't Blink supposed to be doing some serious changes to how JavaScript is handled?

I dunno, either way, bring on Blink regardless because it sounds better than what Chrome(ium) has become right now, a massive heap of slow and awfulness.
Glad they ditched CSS Regions. What a terrible specification that was. More needs to be ditched and turned in to libraries that can be loaded dynamically instead of bloating up everything else to high hell.
Screw the initial expense of loading the libraries.

How about a generic scripting engine?

[Anonymous Coward]

Why don't browsers come with a generic scripting engine (in addition to, or replacing, the JavaScript virtual machine) that runs some sort of intermediate language a la .NET CIL, or LLVM IR, etc.?

Then we could use pretty much any scripting language we wanted with the content.

Re:How about a generic scripting engine?

[tepples]

The last time that was tried, it was called Java applets, and Oracle's incompetence at keeping its virtual machine secure led to browsers blocking Java applets by default.

Re:How about a generic scripting engine?

[qpqp]

Why don't browsers come with a generic scripting engine

They do, it's called JavaScript.

LLVM IR? Here you go
Your other request inverted, compile CIL to JS: jsil.org

Re:How about a generic scripting engine?

[Anonymous Coward]

because speed. If you want fast javascript, you need a javascript engine.

Re:How about a generic scripting engine?

[Anonymous Coward]

Silverlight does that.

Re:How about a generic scripting engine?

[Anonymous Coward]

It really says something about JavaScript that the "hottest" technologies are ingenious ways to avoid actually CODING in JavaScript!

Re: How about a generic scripting engine?

[Anonymous Coward]

Dart Go

Re:How about a generic scripting engine?

[ebno-10db]

But did that really have anything to do with it being a byte code interpreter, as opposed to a lousy implementation of one? Or perhaps applets having access to library functions that were too difficult to make secure?

In other words, is a byte code interpreter necessarily less secure than say a JavaScript interpreter? I honestly don't know, and would be interested if anyone can explain this.

Re:How about a generic scripting engine?

[jbolden]

Why don't browsers come with a generic scripting engine (in addition to, or replacing, the JavaScript virtual machine) that runs some sort of intermediate language a la .NET CIL, or LLVM IR, etc.?

You mean like: Java applets, Active-X, flash or quicktime?

Re:How about a generic scripting engine?

[jbolden]

It isn't the byte code interpreter that's the problem. It is running in a low level state. Let's take a simple example division by zero. If you send a division by zero to an x86 processor something has to catch that exception or the entire system crashes. So either:

a) You have to have a wrapping layer which prevents virtual instructions from causing exceptions in the kernel or OS (Javascript) which is slow
b) You have to have a complex system where things move up and down abstractions layers freely (C++ / Active-X)
c) You have to have a high speed specifically tuned VM that offers services of all types and that is big and complex (Java)
d) You have to have a family of high speed specifically tuned VMs that each offer only some services (flash, movie players, reader...)

etc...

Re:How about a generic scripting engine?

[Anonymous Coward]

No, I mean something like asm.js, but instead of using the JavaScript engine use something that pretty much any scripting language could compile down to.

Re:How about a generic scripting engine?

[Anonymous Coward]

Yep, it says millions of dollars have been invested in making javascript run fast.

Re:How about a generic scripting engine?

[Anonymous Coward]

ALU exceptions like that are easy. Just wrap the whole bloody thing in a try... catch block. The harder problem is securing underlying functions such as file access. You might need your script to access information about the browser or the OS; but you damned well better not allow it to change the configuration of the browser or OS. That's where everybody screws up.

Re:How about a generic scripting engine?

[backspaces]

Any language compiler that will emit LLVM rather than machine code can run in the browser via Emscripten to asm.js. So C/C++ is already ready to run in the browser. Other languages use different strategies such as in-browser translation to JS. Python has several stunts for running in the browser now.

Asm.js was the game changer by getting within 1.5x the C/C++ perf numbers.

Re:How about a generic scripting engine?

[Anonymous Coward]

Yes, Emscripten/Asm.js are cool, but I was kind of hoping for an answer to "what can we use instead of JavaScript" other than "translate language x to JavaScript."

Re:How about a generic scripting engine?

[Anonymous Coward]

It isn't the byte code interpreter that's the problem. It is running in a low level state. Let's take a simple example division by zero. If you send a division by zero to an x86 processor something has to catch that exception or the entire system crashes.

This is Slashdot, when did technobabble become legitimate instead of actual knowledge? What the fuck does that have to do with anything?

You do know that every browser has a built-in assembler, right? Every browser has a JIT which converts JavaScript to machine code then runs the actual machine code it just generated directly on the hardware CPU in order to execute the JS program.

a) You have to have a wrapping layer which prevents virtual instructions from causing exceptions in the kernel or OS (Javascript) which is slow
c) You have to have a high speed specifically tuned VM that offers services of all types and that is big and complex (Java)

A and C are identical, A is just a less efficient version of C. In C you have simple, lightweight bytecode, in A you have a giant heavy-weight bytecode that looks like English words and requires a parser pre-pass to interpret.

d) You have to have a family of high speed specifically tuned VMs that each offer only some services (flash, movie players, reader...)

The fact that you think Flash has less features than Java or JavaScript is amusing. Flash has a comprehensive bytecode engine and a large standard library. Flash "movies" are written in ActionScript 3 which is an extended derivative of ECMAScript, ECMAScript being the standardized version of JavaScript where JavaScript is "ECMAScript for web browsers". Anything you can write in JavaScript can be compiled to bytecode for the Flash VM since the Flash compiler is a JavaScript compiler.

Perhaps you should actually try writing a Virtual Machine before you try making uninformed statements on how they work
PUSH 30 [bytes: 0 30]
PUSH 3 [bytes: 0 3]
DIV [bytes: 1]
Engine:
var stack = [];
var program = [0, 30, 0, 3, 1];
var program_ctr = 0;
function has_next() { return program_ctr < program.length; }
function next() { if (!has_next()) { throw new Error('Program End'); } return program[program_ctr++]; }
var actions = [
      function() { stack.push(next()); },
      function() { var denom = stack.pop(); if (denom === 0) { throw new Error('Tried to divide by zero'); } stack.push(stack.pop() / denom); }
];
var action;
while (has_next()) {
        action = actions[next()];
        if (typeof(action) !== 'function') { throw new Error('Invalid bytecode, not an instruction'); }
        action();
}
alert("Result = " + stack.toString());

Jesus Christ, this is first year CompSci, it's nowhere near rocket science.

The hard part has nothing to do with the bytecode itself, it has to do with the interaction between the bytecode program and the rest of the system. Everything that keeps going wrong with Java, and JavaScript "sandbox escape" attack bugs, are because the safe bytecode can call functions in the environment, and those functions are not always well behaved. If the bytecode has access to an OpenFile function and that function fails to determine that a file is privileged then you have an attack vector. Of course, the smart thing to do is to not expose generic APIs like file access to a sandboxed program but that requires competent designers who actually think about what they're doing instead of just exposing the entire C standard library with a poorly conceived blacklisting function intercepting each call.

Re:How about a generic scripting engine?

[BZ]

For what it's worth, "millions of dollars" is a pretty low bar. A million dollars in the US will get you _maybe_ 5 person-years worth of work from anyone at all competent (using the normal rule of thumb that an employees cost to the employer is about 2x salary once you take into account benefits, employer taxes, equipment, office space, etc).

So 10 million dollars will get you a 10 years worth of work from 5 developers. As an example, the PyPy project is 10 years old....

For JS, between the various browser vendors, the right number is probably closer to 300-500 person-years (see https://news.ycombinator.com/i... for an attempted breakdown). Figure $100 million as a low estimate. Chances are, the people involved are being paid more than $100k a year, so adjust the estimate up accordingly...

Re:How about a generic scripting engine?

[Anonymous Coward]

How about a scripting engine that isn't shit? Do they have that?

Re:How about a generic scripting engine?

[Anonymous Coward]

Google is trying to do something like that with Portable Native Client...

Re:How about a generic scripting engine?

[Anonymous Coward]

Not exactly, although their LLVM bitcode interpreter is equivalent to the kind of target I had in mind. From Google's PNaCl: Portable Native Client Executables : "Developers compile their source to bitcode and ship the bitcode as a portable executable."

I'm not thinking of downloadable native executables, I'm thinking of keeping the scripting in the browser without having to have a complete front-end/back-end interpreter for each and every language supported. Think something like:

<script type=”text/perlscript”>
<script type=”text/pyscript”>
<script type=”text/phpscript”>
<script type=”text/tclscript”>
<script type=”text/myfavoritescript”>

etc.

Re:How about a generic scripting engine?

[shutdown -p now]

JavaScript has very wonky semantics that make it poorly suited for an intermediate representation for another scripting language. Sure, you can pull it off, but it's not trivial and perf is pretty bad.

Re:How about a generic scripting engine?

[shutdown -p now]

asm.js is a neat hack, but doesn't it require you to preallocate the entire heap? It really needs to have some more flexible memory managing policies.

Re:How about a generic scripting engine?

[shutdown -p now]

Google keeps pushing for this with Portable Native Client (where LLVM bitcode is the portable low-level language).

Re:How about a generic scripting engine?

[Anonymous Coward]

See above; PNaCl is targeted to pre-compiled native binaries rather than including scripts on the page.

Re:How about a generic scripting engine?

[Anonymous Coward]

Nope, you only need to be good at language design... look at the "juice" plug-in for NetScape... very fast and safe pe (portable executable format) profiling-compiler (some how, like jit style compiling from compressed platform independent intermediate representation) for the oberon language; made by some doctoral student of Michael Franz

Juice was verifiable, fast, efficient, capable, extensible... the result of many years of research in language design.

Oberon is: a tiny language, and a small but very powerful os, that could run on top of other os, or in the juice case inside other programs like plug-in. Later Niklaus Wirt evolved Oberon to the smaller and more capable oberon2, and finally to a design pattern based framework in OberonF ( one GPL comercial implementation is the BlackBox Component Pascal )

Re:How about a generic scripting engine?

[shutdown -p now]

PNaCl is not precompiled - that's the whole point; it's compiled to the target architecture only once it is delivered to the browser.

And there's no reason why it couldn't be used for scripts on the page.

Re:How about a generic scripting engine?

[Anonymous Coward]

That's not correct; according to Google's own docs, "The intermediate product, an LLVM bitcode file, is distributed."

Re:How about a generic scripting engine?

[Anonymous Coward]

Actually, replying to my own post here, I think you are right-- the binary is pre-compiled bitcode, but it is not native. I misinterpreted what I was reading.

Nevertheless, this is not the kind of solution I'm talking about. Running native binaries or translating portable bitcode is interesting, but not equivalent to writing a Web application that targets the browser, in the way that JavaScript is used on most sites.

It's as if I said, "I wish I could write Qt applications in language x," and you said, "here's a tool that lets you compile your language x app and run it in a QT window." Neat, but not the same target platform.

Re:How about a generic scripting engine?

[shutdown -p now]

PNaCl does run in the browser The only step missing there right now is exposing the entire DOM API to it, but that is a minor thing and relatively easy to do should some browser want to.

The big deal would be making it an accepted standard...

Re:How about a generic scripting engine?

[Anonymous Coward]

I didn't say it didn't run in the browser. Having a low-level virtual machine to target, whatever the instruction set, is a necessary but not sufficient condition to running other languages in the browser. The point is to avoid having to write a complete interpreter - front and back-end - for every language you might want to use on the page, not in a separate sub-process.

Re:How about a generic scripting engine?

[shutdown -p now]

The point is to avoid having to write a complete interpreter - front and back-end - for every language you might want to use on the page, not in a separate sub-process.

"On the page" and "in a separate subprocess" are not contradictory - in fact, I would argue that "in a separate subprocess" is good sandboxing practice in any case. And it's certainly quite possible to set up IPC such that the subprocess can manipulate the page in the same way JS does today.

I don't see anything wrong with "having to write a complete interpreter" - after all, that's the same requirement that we already have for desktop and server-side development, why should browser be any different? And we already have many languages targeting LLVM IR, so, really, it's not such a big deal.

Now the fact that if it is an interpreter, then it has to be delivered to the client alongside with the actual code, is somewhat inconvenient. The entire Lua interpreter is just under 100 Kb gzipped - for comparison, minified & gzipped jQuery is around 30 Kb. So I would expect this system to favor AOT compilers, which would only have to ship, at most, a minimal VM alongside the actual code. If the host would furthermore provide some common services, such as compacting GC, the VMs could be made even smaller if not completely avoided in many cases.

Re:How about a generic scripting engine?

[Anonymous Coward]

Well, I guess if you've got everything you need... have fun playing with PNaCl!

Re:How about a generic scripting engine?

[Anonymous Coward]

Remember the "HTML5" versus "HTML 5" debate ?

Replusive

[gweihir]

Why do the worst technologies that are just barely able to solve the problem always make it? Is the developer community collectively really this stupid? I fear it is...

Re:Replusive

[Nkwe]

Why do the worst technologies that are just barely able to solve the problem always make it? Is the developer community collectively really this stupid? I fear it is...

Because technologies that just barely solve a problem allow people who can just barely do the job to barely solve the problem. People that can barely do the job are less expensive than people who do the job right. Unfortunate enough people are willing to live with "just barely".

Re:Replusive

[sandertje]

Why are some developers obsessed with performance? The user probably isn't. They don't care whether something loads 2 seconds quicker, or whether something has beautifully concise code.

Re:Replusive

[Anonymous Coward]

Having been a developer for nearly 20 years, I can confidently say that yes, the vast majority of developers really are this stupid. It doesn't help that because of retarded corporate policies the mean age of developers is something like 23 years of age.

Re:Replusive

[hibiki_r]

They don't always make it: Many do not make it at all. Survivor bias and all that.

JavaScript thrived because the alternatives were arguably far worse. Java applets were terrible. ActiveX a platform specific disaster. Flash is heavy. JavaScript allowed you to do the very minor things most web developers wanted at the time without having to turn your website into a plugin that disregarded base web technologies.

Yes, the language design is pretty silly. The function declaration syntax is silly. It tries to look like a member of the Algol tree, but its internals behave more like Lisp. The automatic type conversion system is the source of many jokes. But it still beat it competition at the time, because it was built into the browser, it talked directly to the page's DOM, and the competition did not. Today we'd have little trouble designing a better language than JavaScript for what we currently do with it, but our best bet to get something like that working is to build a language that compiles to JavaScript and then hoping browsers start building VMs for that language directly, skipping the JavaScript step. Still, not bloody likely.

Re:Replusive

[Anonymous Coward]

If users are happy with bloated and buggy scripts that never report and handle failures properly, then let them have it!

Compilers are great, because they reduce errors even before execution. Javascript just runs and you never know the correctness of the state.
Errorhandling and exceptions are great, because they don't ignore errors, but force you to deal with them. Javascript just runs and you never know the correctness of the state.

If someone knowledgeable complains about something, it's usually because there is a great point in it less knowledgeable people don't recognize.

Re:Replusive

[Skinny Rav]

Good enough is perfect.

You can endlessly polish your elegant solutions for decades (see Hurd) while the rest of the world happily uses "technologies that are just barely able to solve the problem".

Re:Replusive

[ebno-10db]

a/k/a the Race to the Bottom

Re:Replusive

[Kremmy]

You sound like someone who's never dealt with users.

Re:Replusive

[leptons]

If that is how you think about Javascript, then you have just illustrated your own stupidity. Or maybe it's fear that you are displaying?

Javascript is far from "the worst technologies", and your quip about being "barely able to solve the problem" is pretty ignorant.

I'll get off your lawn now.

Re:Replusive

[hey!]

Because it's easy to mistake one's personal biases for sound judgment about what is "good".

I won't sit here and defend every design decision in javascript, but it's a lot more sophisticated than meets the eye. I think of it more like Lisp than Java; it encourages (among skilled programmers) a functional programming style, which turns out to be both under-used as a programming paradigm and very nicely fitted to the kind of event-driven tasks people use javascript for.

If you aren't writing higher order functions in javascript (functions that take functions as values or return them as values) you aren't fluent in javascript and aren't qualified to pass judgment on it.

Re: Replusive

[Anonymous Coward]

Keep throwing that Dart and eventually it will stick.

Re:Replusive

[Anonymous Coward]

Why do the worst technologies that are just barely able to solve the problem always make it? Is the developer community collectively really this stupid? I fear it is...

Please don't blame this on us developers :-(

We'd be more than happy to code in something else if the browser consortium gave us another way to screw up everyone's web experience. Honest we would.

Re:Replusive

[BurfCurse]

Why is it that everytime I see a truly insightful post, its always the day after my mod points expire?

Re:Replusive

[jbolden]

Because "worst language" is generally a compromise between competing ideas and interests of various stakeholders. Compromise is rarely pretty.

Re:Replusive

[Crudely_Indecent]

Write once - deploy everywhere.

Name one other language that is as well supported and as broadly implemented as JavaScript?

With technologies like Node.js - it can run on both the client and the server.

Sure, there are inconsistencies in implementation - but the bulk of functionality is supported in every implementation.

Re:Replusive

[jbolden]

Actually users care a great deal. Especially as we start moving to touch. There is nothing in nature that causes your visual image of your finger to lag your finger. Latency over 1ms are distressing to human subconscious under those conditions. They experience negative emotions. 0% of the current systems are that fast even using low level languages.

Re:Replusive

[westlake]

Why do the worst technologies that are just barely able to solve the problem always make it?

They are the first to reach critical mass and they get the job done.

Re:Replusive

[ebno-10db]

Latency over 1ms are distressing to human subconscious under those conditions.

Do you have more info on that? Seriously, no snark, but I'll admit that I'm skeptical. I've never heard of human perception time being less than 30ms. I hate slow responses in UI's with a passion, but 1ms?

Re:Replusive

[Anonymous Coward]

I would have to say that I haven't seen this. I work at a company with over 1,000 programmers and our average age is about 40. We have about 5 guys who have more 23 years of experience on our platform alone. We do have (what I would consider) a high washout rate for new guys (10% a year in the first five years) but after that people tend to stick around. Interestingly age of acquisition seems to not be a good predictor of success. In fact the four biggest flops I have seen were all 45-55.

Re:Replusive

[TheGratefulNet]

I am not a java guy (I have the basics, but I still don't 'get' java like some fans do). I asked a guy who works at a hardware company (that does some java accel stuff) what the big draw was. his reply - pertaining to server-java code in business, at least - was that it could tolerate bad programming and still 'run ok enough' for business use. lots of cheap *very bad* programmers in the world and java works well for them. malloc and no free - wow! you can be as stupid as you want and your app will still run 'fine'.

of course, java is also big on phones and embedded devices. but on servers, I never quite understood what the draw was. I'm a C guy (from quite a long time ago) and still prefer C, but I realize that the new breed of programmers mostly don't even know or care about C anymore and its always java this and java that for any new project that starts.

Re:Replusive

[Jane Q. Public]

"Why are some developers obsessed with performance? The user probably isn't. They don't care whether something loads 2 seconds quicker, or whether something has beautifully concise code."

Although I am a programmer, I am also a "user". And from purely a user perspective: I sure as hell DO care.

Slow programs are slow. Computers are supposed to wait on people, not the other way around.

Re:Replusive

[StripedCow]

You are modded funny, but if this really is a joke, then there certainly is some melancholic undertone to it.

The web is being developed for "the average" developer that just wants to get things done. This is a pity. Because if the web were developed for the advanced developer (read: computer scientist), those developers could lift the web to a higher level, and there would be much more traction.

We just have to wait until Javascript gets sufficiently powerful to function as an intermediate language.
From that moment on, we can have other languages targeted to the web. E.g., Python, Haskell, you name it.

Right now, it is possible to use Javascript as an intermediate language using the ASM.js subset (what a name, what a concept), and there are actually projects that do stuff along these lines, but still, performance suffers a lot in this area. Perhaps a better approach is Google's NaCl, where intermediate code is translated more directly to native code, while putting security guarantees in place. But here, of course, cross-platform support is an issue.

Perhaps then a hybrid between NaCl and ASM.js would be the best approach. That is, a compiler back-end that can target both NaCl *and* ASM.js.

Time will tell...

Re:Replusive

[TheloniousToady]

Good enough is perfect.

That's a concise and interesting twist on the old adage Perfect is the enemy of good. Well done!

Re:Replusive

[TheloniousToady]

You can endlessly polish your elegant solutions for decades (see Hurd) while the rest of the world happily uses "technologies that are just barely able to solve the problem".

Which got me thinking...since the Mythbusters recently proved that you can polish a "Hurd", I guess the GNU folks can retire their experiment now.

Re:Replusive

[Areyoukiddingme]

Do you have more info on that? Seriously, no snark, but I'll admit that I'm skeptical. I've never heard of human perception time being less than 30ms. I hate slow responses in UI's with a passion, but 1ms?

See, for instance, John Carmack's analysis of head mounted display latency. Under 20 ms is acceptable for most people. 30 ms is too high, and leads to motion sickness.

For ~1 ms latencies, experiment with a mouse on an old CRT running at 180Hz. Try out a hardware rendered mouse, then a software rendered mouse. You can indeed see the difference.

Also for time perception that accurate, see a competent analysis of a fighting game. Many fighting games feature reaction windows no more than 5 ms wide, and some have windows ~1 ms wide. And players can hit them.

Human visual perception and motor reaction is a very complex, multipath system with a lot of variance between individuals. Some of those paths are very slow. Some of them are suprisingly fast. Touch screen systems are some of the worst high latency offenders, in a visual perceptive/motor reactive realm that is least tolerant of it. When you touch something in the real world, it responds instantly. Push a marble and it rolls, no delay. Touch screens can't do that. Not even close. It's really really jarring, and it makes the choice of scripting languages on mobile devices particularly unfortunate, since they're not capable of executing fast enough to solve the problem. Your entire reaction and perception system is definitely slower than it feels, but it is also equipped with compensation for that slowness. Your brain conceals the lag from you when interacting with the real world. That process of concealment has a lot to do with expectations. Touch screens especially have a tendency to violate those expectations.

There's a lot of work yet to be done in human-computer interaction, and getting touch screen responsiveness down to the equivalent of the venerable mouse is seriously important, at this point. Unfortunately, it doesn't seem to be much of a priority. See complaints about hipster programmers elsewhere in this thread.

Re:Replusive

[bearinboots]

What does Java have to do with JavaScript -- besides the stupid naming decision, that is?

Re:Replusive

[Anonymous Coward]

I am a C programmer as well. While it is awesome for low level/systems stuff, you want something OO to solve complex problems, and you want something script-like to solve easy problems. Also, the job market around where I am has few to none C openings as of the last decade. So the logical extension of C for OO capability is C++, right? Right? You would not beleive it, but modern C++ techniques hide memory management to where sun never shines: Singletons; because manually managed memory on a complex system is like blood of a lamb being torn apart by a lion and a crocodile.

Re:Replusive

[quetwo]

Actually, the human eye can really only perceive images with 16ms at the very fastest. Most people's cognitive recognition is at about 50ms, and auditory recognition is about 70ms.

Re:Replusive

[Billly Gates]

Worse?!

Javascript is bad ass rockstart and webscale language. It invents this new thing called threading that no one has ever thought of before and introduces non-blocking async i/o. Wow this is sooo new and cool and writting it in javascript is a pleasure of all the languages.

Re:Replusive

[SnowZero]

Perhaps a better approach is Google's NaCl, where intermediate code is translated more directly to native code, while putting security guarantees in place. But here, of course, cross-platform support is an issue.

You might find Portable Native Client interesting, which is built out of LLVM+NaCl.
http://www.chromium.org/native...

I'm hoping that eventually browsers, mobile phones, and cloud hosting will just become sandboxed LLVM targets. Then people could use whatever language they like, wherever they want.

Re:Replusive

[matthewv789]

if the web were developed for the advanced developer (read: computer scientist), the "web" as we know it would be confined to a few geek niche markets and the rest of us would be using some Microsoft- or Adobe-pushed proprietary technology instead.

FTFY.

Re:Replusive

[StripedCow]

The problem with PNaCl and NaCl is that it is not really portable to platforms like Firefox and IE.

Sadly, I don't believe even that Mozilla will ever adopt NaCl, judging from these responses (wikipedia):

Mozilla's vice president of products, Jay Sullivan, said that Mozilla has no intention of running native code inside the browser, as "These native apps are just little black boxes in a webpage. [...] We really believe in HTML, and this is where we want to focus."[31]
Mozilla's Christopher Blizzard criticized NaCl, claiming that native code cannot evolve in the same way that the source code-driven web can. He also compared NaCl to Microsoft's ActiveX technology, plagued with DLL hell.[4] Mozilla currently employs the inventor of JavaScript, Brendan Eich as chief technological officer.

The best approach, therefore, is to build a compiler back-end that targets *both* ASM.js *and* (P)NaCl.

The performance benefits of NaCl will, in time, become clear, and hence other platforms will be more or less forced to either improve their Javascript implementation, or adopt NaCl.

Re:Replusive

[Wolvey]

Why are some developers obsessed with performance? The user probably isn't. They don't care whether something loads 2 seconds quicker, or whether something has beautifully concise code.

You sir, are part of the problem. Thinking like this leads to layers upon layers of inefficiencies and the user experience often suffers dramatically. Why don't people just code efficiently? Is it too hard or are you just too lazy?

Re:Replusive

[jbolden]

Microsoft labs is the best source for this. Here is a quick video where they introduce the concept: https://www.youtube.com/watch?...

Re:Replusive

[Anonymous Coward]

Mostly Java was the alternative pushed by Sun and IBM as a replacement for C++, because C++ was Microsoft's language of choice (and C++ was also horrible).

Re:Replusive

[jbolden]

Where are you getting those numbers from? For example in lip-sinking 2.9ms errors are routinely detected even with distracting visuals.

Re:Replusive

[Hewligan]

Why do the worst technologies that are just barely able to solve the problem always make it? Is the developer community collectively really this stupid? I fear it is...

Because the people who spend their time worrying about which is the theoretically superior solution rarely come up with a practically useful solution.

Re:Replusive

[smellotron]

Many fighting games feature reaction windows no more than 5 ms wide, and some have windows ~1 ms wide. And players can hit them.

I've heard of fighting games (e.g. Street Fighter) with 1 frame windows, but the games were 60 fps ~ 16.67ms. That is more in line with your figures from John Carmack's experimentation. Your 1ms figure does not sound credible, can you provide a reference?

Re:Replusive

[smellotron]

Name one other language that is as well supported and as broadly implemented as JavaScript?

C is ubiquitous.

Re:Replusive

[Anonymous Coward]

because it was built into the browser, it talked directly to the page's DOM, and the competition did not

This is why it made it so far. Compared to Java or C# or any other myriad of languages, JS did not start with (and still doesnt have) any *major* advantage. Rather it used to have a lot of major disadvantages (and still has a handful, though much less). However, JS, and the tools, and the browsers that provided the runtime, matured. But it was positioned in the runtime that people were already using: the browser.

It is a much more robust nowadays, especially with HTML5 and CSS, but that wasnt always the case.

Re:Replusive

[PCM2]

The best approach, therefore, is to build a compiler back-end that targets *both* ASM.js *and* (P)NaCl.

Or just compile your code twice, using two different back ends? I don't see much wisdom in building a compiler that tries to do two things at once.

Re:Replusive

[StripedCow]

Well that's what I meant actually.

Also note that some aspects of the target intermediate languages may limit the actual programming language in certain ways.

For example, the language from which you are compiling may support some primitive types that the back-end languages do not support. Then, to stay efficient, it may be necessary to find a common denominator of primitive types in ASM.js and NaCl, and introduce these as new primitive types in the original language.

Re:Replusive

[Darinbob]

Because it's not how good the salesman is, it's how much of his foot he can get in the door (and how much pain he can withstand when they try to slam the door shut).

Basically, Javascript got the foot in the door very early on with browsers; not for applets (which really never worked that well with Java), but as just some simplistic things to do like pop up windows (with ads!) or move windows around (with ads!) and so forth. So when people wanted a bit more complexity with the scripting they could either evolve what was already present, or try to get a different language into every browser.

Re:Replusive

[gweihir]

I am currently working on a problem where the user is "obsessed with performance", because the JavaScript monster created for them by some other people is so slow, it markedly affects user performance and satisfaction. Unfortunately, this being JavaScript, nothing can be done to help them.

Re:Replusive

[gweihir]

Makes sense. What I have seen in code-reviews is atrociously bad, despite being touted as "quality development".

Re:Replusive

[Darinbob]

Originally Java was supposed to be platform independent (there's that nasty "platform" word again). That is it didn't matter what type of CPU you had, it would run on your computer, PC, Mac, Sun, minicomputer, mainframe, etc. The it was quickly diverted to run on the web so it cobbled together some security features ala safe sandboxing, although it didn't really work out there.

These are great goals by the way. Except for me at least, Java was a moving target, it never stood still long enough to learn it well, with frameworks becoming popular and then discarded a few years later. As a language it's pretty nice (or was in version 1.0 anyway), bringing together tried and true features that really hadn't gotten into the mainstream. However the early implementations were not very good, Microsoft sabotaged it pretty well keeping it off the PC. Everything that is .NET today could have been Java instead and be portable.

Re:Replusive

[gweihir]

"Barely good enough" is not "good".

Re:Replusive

[gweihir]

And it happens to weed out all those cretins that cannot program to save their lives pretty fast. Mind, it is nice to embed it in, e.g. Python, for simple glue-code.

Re:Replusive

[gweihir]

This is far too simplistic. Sure, your first version may be there earlier, but I have seen large projects where the people cannot move forward anymore and cannot even fix essential issues because JavaScript does not allow them to. They cannot even rip it out anymore.

Re:Replusive

[tepples]

C is ubiquitous.

Except on platforms that ban third-party native applications. For example, third-party applications for Chrome OS and Firefox OS are written in JavaScript, as are iOS applications that contain prohibited behaviors. Applications for Windows Phone 7 and the Xbox Live Indie Games are written in .NET Compact Framework bytecode, which in practice means C#.

Re:Replusive

[tepples]

In the Groove (arcade) polls the controllers more than once per frame and can resolve step timing to within 10 ms. The PS2 version has slightly wider timing windows because it polls the controllers only once per frame (16.7 ms).

Re:Replusive

[Crudely_Indecent]

and you must compile for every platform.

Re:Replusive

[smellotron]

C is ubiquitous.

Except on platforms that ban third-party native applications.

Even if the walled gardens don't permit software implemented in C, it is extremely likely that some part of the walled garden's own software stack is still written in C—because C is ubiquitous.

Re:Replusive

[Anonymous Coward]

Yet the web works, and you haven't proposed even step one of a "replacement"

Re:Replusive

[Anonymous Coward]

You try and run C in a browser.

Re:Replusive

[smellotron]

and you must compile for every platform.

Yes, deployment requires compilation once per target. How many platforms exist which have a javascript interpreter but no C compiler? I believe my point stands.

Since nobody has offered any other languages to your challenge, I'll suggest Lua as a trivial alternative: the core language is implemented in ANSI C, so it is likely that Lua is portable to any platform that supports C. However, the core functionality of Lua is ridiculously hamstrung (ANSI C doesn't support sockets, for example) so the language is not well-suited for general-purpose programming.

Re:Replusive

[Anonymous Coward]

that was a really long way of saying "no"

Re:Replusive

[Anonymous Coward]

So, you're completely intellectually dishonest and you know you're misrepresenting the truth, or you're so stupid you have no idea what you're talking about you prattle on anyway, repeatedly.

Which is it?

Re:Replusive

[rycamor]

Exactly. I'm sick of all this talk of how bad Javascript is. What were the alternatives in 1997? Brendan Eich had to solve a problem quickly and actually get something into production. Same with Tim Berners-Lee and HTML. Not perfect or elegant technologies, BUT they actually solved a problem within a finite time without expending ridiculous amounts of brainpower in design-by-committee cycles. The real world directly benefited from these solutions.

Really, Javascript only has two major problems: 1. The security model (freezing of libraries so they can't be modified by other code loaded in the browser) and 2. A few ugly aspects of how the language handles this or that type of expression... most of which have been solved in latest versions of Javascript (Harmony). Most of the other problems are just questions of implementation and add-on technologies (WebGL, Websockets, etc.. expanding the scope of a scripting language into areas that may... not... just... be the wisest)

Yes, there are all sorts of esoteric reasons why Javascript is a Terrible Lisp, and a Downright Horrible Scheme, and an Ugly Hybrid of Object-Oriented and Functional, blah blah blah... So what? It's meant to solve small finite problems in the front end. Let's keep it that way, and just clean it up. Fortunately, it is a fairly simple language.

To all those who want a type-checked, compiled language running in the browser... Uh no. I don't want to allow any site to run code I can't inspect.

Re:Replusive

[Crudely_Indecent]

Yes, I understand that C is virtually everywhere, but requires much more to deploy an application everywhere.

Javascript: write once - deploy everywhere
C: Write once, compile for every target, deploy everywhere.

See the difference?

And how often is it necessary to alter a compiled application for nuances on specific targets? Javascript is often written wifh fallbacks for targets that don't support a specific feature.

I'm not saying Javascript is superior or inferior to C. They have different strengths and weaknesses. Why must there be a "best language" or arguments over "my language is better than yours"? It's all a personal preference when more than one language can perform a given task. What about tasks that are possible in one language and not in another? Apples and oranges comparing C and Javascript.

Thursday I was writing Objective-C, Friday I was writing Java, today I was replacing the fuel-injectors in my truck, sometime next week I expect to be writing some PHP. Personally, I enjoyed replacing the fuel-injectors more than the work I did Thursday or Friday - and now my truck is running better. Because fuel-system work makes my truck run better, does that make auto-repair better than Objective-C or Java - because neither of them make my truck run better? Objective-C and Java pay the bills while working on my truck does not - does that make programming better than auto-repair? Apples and oranges.

I think it's safe to say that you prefer C to the exclusion of everything else.

Re:Replusive

[smellotron]

I'm not saying Javascript is superior or inferior to C.

I'm not trying to argue that any language is better, either. I originally responded to this:

Name one other language that is as well supported and as broadly implemented as JavaScript?

and your subsequent reply moved the goalposts by arguing about deployment effort. Yes, you're right, C software requires more work to deploy to N platforms, but that is irrelevant to the presence of broad multi-platform support. I'm really not trying to say that C is better than Javascript. But surely it has a larger footprint across platforms!

I think it's safe to say that you prefer C to the exclusion of everything else.

Actually, no. I can read C fine, but I don't code fluently in it. I usually switch between C++, Python, and bash depending on the task.

Re:Replusive

[Crudely_Indecent]

OK. Unlike the Dos Equis "Most Interesting Man" - it occasionally is "my bad"

Frequently these types of threads turn into bashing contests between different language enthusiasts. Usually things go downhill for my contributions to the thread when I mention that I enjoy my (infrequent) opportunities to code in the first language I learned - COBOL.

(cue the COBOL haters in 3, 2, 1)

Javascript is one of those languages that I love to hate and hate to love.

Re:Replusive

[Anonymous Coward]

Whoosh

Re:Replusive

[Anonymous Coward]

>Write once - deploy everywhere.
Only with 1,000 framework/cross platform libraries.

>With technologies like Node.js - it can run on both the client and the server.
Yay! We can have the same shit performance on the server! Woo! Enabling shitty, bottom-tier "developers" to write server code is the exact thing COBOL was designed for. And now look at it.

>Name one other language that is as well supported and as broadly implemented as JavaScript?
Java ME. Far more devices support this.
C - The thing every language interoperates with because everything important is written in C or C++.
C++ - Every device javascript runs on, can run c++ code. And then some.

Re:Replusive

[Anonymous Coward]

I think the reason why JavaScript is so popular is fairly simple. To write your first bit of JavaScript you just need to be able to write a text file with some script tags and save it as file.html.

Compare that with the amount of preparation you need before using EJBs for example.

Re:Replusive

[Anonymous Coward]

http://www.youtube.com/watch?v=RnqAXuLZlaE

Re:Replusive

[trenobus]

JavaScript thrived because the alternatives were arguably far worse. Java applets were terrible. ActiveX a platform specific disaster. Flash is heavy. JavaScript allowed you to do the very minor things most web developers wanted at the time without having to turn your website into a plugin that disregarded base web technologies.

No, JavaScript survived because it wasn't seen to threaten corporate interests until it was too late to stop it. Java applets failed because Microsoft did everything in their power to kill them. Yes, there were problems with the technology, but early versions of JavaScript offered far less functionality. The only reason people still talk about how bad applets are/were is because they actually got used to solve problems which JavaScript could not. Flash was also used heavily and is still hanging on, though probably not for much longer.

JavaScript is the result of the major players of that time, Microsoft, Sun, Netscape, and Adobe, failing to realize that an open standard would be in their best interests. It is unfortunately the case that corporations tend not to endorse open standards until no alternative is left. This happened with TCP/IP - by the time corporations like IBM and DEC got around to collaborating on ISO/OSI as the future replacement of their proprietary network protocols, it was too late. Network companies like Cisco and many others were bringing products to market based on TCP/IP. The ISO/OSI protocol got to the party just a little too late.

So we have TCP/IP with its NAT and its ever-so-drawn-out migration to IPv6, when we could have had something better. And we have JavaScript, which was a cute solution to making live widgets on a web page when it was invented, but like TCP/IP has been pushed far, far beyond what it was designed to do. The inevitable fate of such technologies is to become a victim of their own success. So it will be with JavaScript. I don't know whether that will be because it is replaced by some language that compiles into JavaScript, or whether some new browser runtime such as Dart will do it. Or perhaps the mobile device market will make browsers themselves obsolete, in favor of some mobile O/S.

What will not change is that corporations will continue to try to own the next standard. That will inevitably fail, and the software industry will be the poorer for it.

Re:Replusive

[PCM2]

My understanding is that the most common use case for both NaCl and ASM.js is to write your code in, of all things, C.

Re:Replusive

[Anonymous Coward]

s/H/T/urd

Re:Replusive

[spongman]

Java applets failed because Microsoft did everything in their power to kill them.

Really? For a long time the MS JVM was the fastest and most compatible (according to Sun's own verification suite) VM available for any browser on any OS.

Re:Replusive

[Anonymous Coward]

"Whoosh" does not apply in this instance.

Re:Replusive

[trenobus]

Really? For a long time the MS JVM was the fastest and most compatible (according to Sun's own verification suite) VM available for any browser on any OS.

MEMORANDUM OF THE UNITED STATES IN SUPPORT OF MOTION FOR PRELIMINARY INJUNCTION

The whole thing makes interesting reading, but just search for Java and JVM to see what I'm talking about. Microsoft was following their "embrace, extend, and extinguish" strategy, which had worked well for them many times before. They had a Java-like language, J++ that was not compatible with Sun Java.

Sun's has to share some of the blame for Java failing to become a browser standard, as they did everything they could to maintain control of Java, rather than turn it over to an open standards body. But in fairness they did invent it, and Microsoft had the resources to dominate any open standards effort through sheer numbers of representatives. In contrast, Netscape did turn JavaScript over to Ecma (a very wily choice, I think), and I don't think we'd still be discussing it if they had not.

Re:Replusive

[spongman]

yeah, hardly an unbiased source, that.

the truth is that MS's JVM was more compatible and faster than any other VM (although the Semantec VM eventually caught up) at running 100% java-compatible applet code. sure, it was also capable of running non-compatible code but you could say the same about the browser, the OS and the machine in general.

Re:Replusive

[shutdown -p now]

In what sense does it work "more like Lisp"? Just because a language has first-class functions and closures doesn't make it "more like Lisp". Every single mainstream language these days also has closures, with the sole exception of Java which is getting them in the next release.

Re:Replusive

[tigersha]

Agree. I always thought Brendan Eich's decision to base his scripting language on Self and make it a functional language instead of the Algol-like mediocre dreariness of PHP and Java and Python was a brilliant idea. Javascript has been held back for many years because of the DOM and browser wars. It also has a somewhat limited (and silly at places) standard library, but the base language is not bad at all.
The fact that compiling technology for Imperative languages was 20 years ahead of anything functional and dynamic did not help either.

Higher-order functions? Quit advanced object system? In 1996? The language was waaaaay ahead of its time. And Brendan Eich is a genius. I wish I will have the opportunity to meet him one day.

Re:Replusive

[cwsumner]

Latency over 1ms are distressing to human subconscious under those conditions.

Do you have more info on that? Seriously, no snark, but I'll admit that I'm skeptical. I've never heard of human perception time being less than 30ms. I hate slow responses in UI's with a passion, but 1ms?

I can read a line on a screen that is only there for 1/10,000 of a second.
And so can you, with a little practice and assuming the background light is not too bright.
Of course, it's not a computer screen, they can't do that.
Human senses are not very good, but they are a lot better than people think.

Re:Replusive

[cwsumner]

Actually, the human eye can really only perceive images with 16ms at the very fastest. Most people's cognitive recognition is at about 50ms, and auditory recognition is about 70ms.

Those numbers are from the wrong kind of measurement. They were a test of how long the image would remain in the visual cells if a black background was presented. Like for avoiding flicker in a video.
They have nothing to do with how fast an image can be percieved, which is several orders of magnitude faster depending on light level and contrast.

jscript

[Horshu]

I just started on a web project that is heavily javascript (even though it's an ASP.Net app). It feels like I stepped back in time 10 years. I'd have code not execute. Why? Some error somewhere in the code that was ignored. I had a typo on a property set somewhere else. No error. Why? Because javascript created a new property with the typo and set *that*. And every change requires me to run the app to see if it works, because I don't have a compiler to check the basics out ahead of time. And then, because I'm debugging an app in one window, when I'd open a new one to read the news, I'd hit other, non-ignored errors on those pages (advertisements) because 90% of commercial pages these days have error-laden javascript because people rarely check for error conditions. Horrible language.

Re:jscript

[Anonymous Coward]

JavaScript is a horrible language because you haven't heard of things like jslint and your team doesn't develop with TDD using one of the numerous testing frameworks that exist for JavaScript?

Re:jscript

[Anonymous Coward]

Yes, javascript is a horrible language. The fact that there exists tools and practices that make it almost, but not quite, tolerable does not change this fact.

Re:jscript

[Anonymous Coward]

Yes, but why is it a horrible language? Do you really think it is horrible just because (1) it isn't compiled, (2) you don't understand all the concepts, and (3) people can write code that has errors in it? This describes a lot of languages. Personally I think it's a case of PEBCAK.

Re:jscript

[Anonymous Coward]

because 90% of commercial pages these days have error-laden javascript because people rarely check for error conditions. Horrible language.

Bad programmers programming in any language will result in bad code.

Hell, my monkey programming in C is going to leave at least one dangling pointer. Does that make C a horrible language? No, it doesn't. But it does make my monkey cry.

Re:jscript

[timmyd]

You might want to look at TypeScript if you're already using Visual Studio. It infers types, type checks your code, is open source, and supports writing plain JavaScript. When using Visual Studio, you can do the things you're used to doing like "go to definition" and "find all references". If you decide to annotate your definitions with types, it can do type checking and catch errors which is really useful when you need to refactor a lot of code. The video at the bottom of http://www.typescriptlang.org/ is a really good tutorial. It compiles to JavaScript and accepts plain JavaScript so you can use it without having to rewrite all your code.

Re:jscript

[Laxori666]

Try getting a good IDE, like PyCharm (which does javascript as well). It gives you pretty good static checks on things like var name and property name typos. Not 100% of course since that's undecidable.

Re:jscript

[leptons]

Oh you're such a big, brave programmer because you aren't a "Script Kiddie", and hiding behind "Anonymous Coward". Give me a break. slashdot seems to be filled with people with narrow vision such as yourself.

Re:jscript

[Anonymous Coward]

Forget its awkward object model, Javascript has two flaws that really annoys me:
1) Javascript is bad at the only thing is should be good at: the DOM. Javascript should adopt jQuery's way of accessing and modifying the DOM. Is it really necessary to name the single most important function of the language "getElementById"? Why did it take so long before "getElementByClassName" appeared? Why can't we combine them and access them by their CSS names?
2) Although not the language's fault, the implementations of the language differs between browers, and some browsers (guess who) does an incredibly poor job of implementing it. And we have to support and work around that. I've spent so many hours making sure my scripts run in IE it's not even funny.

I, for one, beg that Dart will catch on!

Re:jscript

[Bite The Pillow]

Sounds like you are not using the correct tools and environment, which makes debugging harder. And using that same setup to read news with a coding environment is just stupid. I don't read news with my development database or web server, so why would I risk infecting my development environment from known malware vectors of even legit advertising?
You conclude with "horrible language", but your comments support "horrible programmers" all the way around.
I'm making no defense here, only pointing out that your opinion is not based on objective consideration. Were it otherwise, you would support your conclusion better, naturally.

Re:jscript

[ebno-10db]

my monkey programming in C is going to leave at least one dangling pointer. Does that make C a horrible language?

The difference is that there are good reasons for that design tradeoff in C. The point of C is that it's fast as hell and gives you almost complete control. You need that to implement a kernel, which is what C was originally designed for. The downside is that you have to watch out for things like dangling pointers.

The GP's complaints are a result of what tradeoffs in JS?

Re:jscript

[anchovy_chekov]

Just compile your Dart to JS (https://www.dartlang.org/docs/dart-up-and-running/contents/ch04-tools-dart2js.html). Then you can debug the generated code on ALL platforms :)

Re:jscript

[laffer1]

From a developer perspective:
JavaScript is a pain in the ass to debug.

From a user perspective:
JavaScript can be slow. -- this one is getting better and may not matter in time.

Full stack JavaScript is insane. There are faster languages for server side and everyone pushing for it didn't live through the first try with classic ASP which was VBScript or JScript. There was even an implementation that ran on top of Java for netscape servers called chillisoft or something like that. It was even worse than running on windows with IIS. I don't want to go to the bad old days.

NodeJS is an attempt to get cheaper backend programmers because everyone has extra front end developers lying around for projects. The problem is that most front end developers I know don't know shit about big data or working on real problems. Their biggest fear is if a button is pretty and the popup works in IE8 and the latest jQuery UI.

It's stupid.

Re:jscript

[Anonymous Coward]

I'm not really concerned with JavaScript's warts, they are manageable. I'm concerned about what I can DO with a language, and you can do a LOT with JS.

However, one criterion you can use to judge a language is how easy/hard it makes certain common tasks for the programmer to accomplish quickly, correctly, and maintainably. In that respect, JavaScript is great for some tasks and horrible for others. You do have to know what you're doing.

Basically, you can run through a minefield at high speed if you know where the mines are!

Re:jscript

[Anonymous Coward]

Modern browsers support querySelectorAll() which is effectively the same as the jQuery model.

To answer your question, it's because W3C bozos got involved and came up with the getElementById() finger-twisting. Both Netscape and Microsoft had much simpler APIs where you could type document["foo"] or document.all.foo to get an element ... these still work but nobody likes to talk about them because they're non-standard.

Re:jscript

[Anonymous Coward]

But node.JS is soooo much faster that it invents async i/O?!!!

Re:jscript

[Anonymous Coward]

>Some error somewhere in the code that was ignored. I had a typo on a property set somewhere else. No error. Why? Because javascript created a new property with the typo and set *that*.

You can seal objects to prevent objects from being extensible.

> And then, because I'm debugging an app in one window, when I'd open a new one to read the news

Yeah I hate when I open up Newsweek in Eclipse. (Treat your browser instance as part of your development platform).

> because I don't have a compiler to check the basics out

GruntJS can watch for changes to your JS, HTML, CSS, whatever, and autorun your tests for you, as well as JSHint/JSLint, etc. You... are writing tests for your JavaScript, aren't you? Search for GruntJS, grunt-contrib-jasmine, grunt-contrib-watch, etc.

>Horrible language.

It's actually a really good language when it clicks :) First-class functions are absolutely powerful.

Re:jscript

[matthewv789]

There are faster languages for server side

Such as what? Well sure, you can write your web apps in C using G-WAN if sheer performance is your primary concern, as opposed to PHP, Perl, Python, Ruby, Java, C#, etc. But there's a reason people moved from C to other languages: quite simply, developer time is waaaaay more expensive than server time for all but the heaviest uses. (On the other hand, Node.js is very fast for high concurrency transactions and can cut server costs as a result, so you can have both.)

NodeJS is an attempt to get cheaper backend programmers because everyone has extra front end developers lying around for projects.

From this statement it seems you don't know much about NodeJS or why it exists. One thing that's different is that Node.js IS the server, and its pervasively asynchronous, non-blocking, event-driven IO is not something other web servers or language ecosystems offer. JavaScript is a good match for that kind of programming, there are many developers with experience in asynchronous JavaScript programming from the web front end (not all of whom are completely stupid), and Google's V8 engine (which compiles to machine code) is not exactly slow. People are using Node because it is faster than other web backends for transactional responses with high concurrency - AND it happens to be easy to develop for too, compared to, say, Java (at least according to PayPal and others working with it).

It is certainly not anything like ASP.

Re:jscript

[Nivag064]

How about putting the following line at the start?
'using strict';

Re:jscript

[Tablizer]

There are times and places for scriptish languages and times and places for anal-retentive languages (AKA strong typing). The problem is that JavaScript being the browser's only built-in language tends to result in trying to make one-language-fit-all. It practically removes choice of tools to fit a specific need.

Re:jscript

[Horshu]

As should be obvious from what I wrote, I'm a .Net developer so my environment is VS.Net debugging against IE. The javascript project is not my primary target, so I don't have, nor do I wish to have, special tools specific to JS. It isn't "stupid" to debug against IE while reading news with IE, as it's a damn web browser. As for your last statement, this is a fucking comment section, not a dissertation, hence the lack of arguments up to your exacting standards.

Re:jscript

[ebno-10db]

you can run through a minefield at high speed if you know where the mines are!

That doesn't sound like a very enticing analogy for a programming language. Then again I use a lot of C++, so who am I to talk?

Re:jscript

[PCM2]

The point of C is that it's fast as hell and gives you almost complete control.

No, assembly language is fast as hell and gives you complete control. The point of C is that it gives you almost as much control but makes it easier to build and maintain large systems without you being some kind of semirobotic idiot savant. In other words, C IS a "dumbed down" language, just like some of the other ones people are complaining about.

Re:jscript

[PCM2]

To clarify your last statement a bit, TypeScript is designed as a superset of JavaScript. All valid JavaScript is valid TypeScript. All you have to convert your project is say "my project is in TypeScript now," and you can then start adding TypeScript features to your code -- or not -- as you see fit. That's as opposed to Dart, which really is a different language with a different syntax and you have to start over.

Re:jscript

[Jeremi]

The point of C is that it gives you almost as much control but makes it easier to build and maintain large systems without you being some kind of semirobotic idiot savant.

I think the other real point of C is that it allows you to write portable code. With assembly, if you wanted to port from one CPU architecture to another, the only option was to throw all the code away and rewrite from scratch.

Re:jscript

[Anonymous Coward]

I had a typo on a property set somewhere else. No error. Why? Because javascript created a new property with the typo and set *that*.

FFS... This is normal problem for the vast majority of scripting languages. JavaScript is a scripting language.

There's lots of tools available to help you, but my favourite is Google Closure because it can minify my code when its done statically analysing it for errors.

I can't defend JavaScript as being the perfect tool for how it is used now, but it does the job if you work with it the right way.

Re:jscript

[Anonymous Coward]

Do you feel like getting angry somehow makes your point more meaningful? Those of us who actually know something already dismissed you in the first place. This reply just lost half of the rest.

Re:jscript

[Anonymous Coward]

It's actually a really good language when it clicks :) First-class functions are absolutely powerful.

Basically every language I use has first-class functions. Did you mean something else specific to Javascript?

Re:jscript

[XcepticZP]

The javascript project is not my primary target, so I don't have, nor do I wish to have, special tools specific to JS.

That's just silly. "This is my only project in language A. Therefore I won't get any tools that help me program in language A". If you're too lazy to download a free IDE and install it for the single project you are working on, then you deserve to suffer trying to use that programming language. Right tools for the right job, and if you can't bother to work smarter then you're just screwing yourself over.

Re:jscript

[joss]

> NodeJS is an attempt to get cheaper backend programmers because everyone has extra front end developers lying around for projects. The problem is that most front end developers I know don't know shit about big data or working on real problems. Their biggest fear is if a button is pretty and the popup works in IE8 and the latest jQuery UI.

No, it really isn't and nobody with any experience of it believes that. It's for when performance requirements are such that you want to control your own threading rather than letting the operating system do it for you. It's more like programming in assembly than in a high level language and as such requires a much higher level of knowledge on behalf of the server side programmers than something relatively straightforward like java let alone python or php.

Just for the record, I was a c++ programmer for 15 years an EJB (curse its fucking name) programmer for 6 years, and have been doing a lot of front end javascript stuff for 6 years plus some node.js for the last 2 years. I just delivered the server end of a project which deals with more data than youtube ( http://www.manything.com/ ). Get off your fucking high horse about front end web development. Big data is a fucking walk in the park compared to delivering a decent user experience on web platforms.

I'm not here to praise node.js: many environments have pitfalls, but node.js greases the sides, put stakes at the bottom and threw down a bunch of venomous snakes just to be on the safe side. Plus, you're working in a language where
[1, 2] + [3,4] === '1,23,4'

However, having said all that, it has its place, its just not where you think it is.

Re:jscript

[PCM2]

Well, that's true, too.

Re:jscript

[Anonymous Coward]

JSLint, JSHint and a good JS IDE would help identify most of the "basic" issues ahead of time, and even some of the not-so-basic ones. There are very good tools out there in JS land - you should use them.

Re:jscript

[shutdown -p now]

Because it is poorly designed. There are already lengthy essays written on the subject, so it's not worth the bother rehashing all the arguments - they are too numerous, anyway.

Re:jscript

[shutdown -p now]

To clarify this even further, TypeScript is, for the most part, ES6 with (optional, compile-time-only) type annotations and modules tackled on top. Everything else in it is directly from the ES6 draft spec, like the arrow functions and even classes. If you take a TS program, strip out all the type annotations, and replace import statements with require() or your local equivalent, you'll get an ES6 program.

Re:jscript

[jlowery]

[1,2].join([3,4]) === '13,42';

FTFY

I has a sad

[Snotnose]

In my 35 years of professional programming, getting good at dozens of languages, I've only run across 2 I've actively disliked. Javascript is one of them (tcl was the other). JS is a crap language that IMHO can't be fixed. If they ever add an honest garbage collector to the base language then most programs will delete themselves upon execution.

Re:I has a sad

[anchovy_chekov]

In my 35 years of professional programming, getting good at dozens of languages, I've only run across 2 I've actively disliked. Javascript is one of them (tcl was the other). JS is a crap language that IMHO can't be fixed. If they ever add an honest garbage collector to the base language then most programs will delete themselves upon execution.

Lol. And the garbage collector would then send out an email to every web designer who says "I know how to code in Java" when they mean Javascript and clean up that confusion once and for all.

I've been forcing myself to get good at writing JS lately (if only because Node looks like it'll make all my other skills irrelevant in the web development market). It.. just.. feels.. wrong. Nothing in the language lends itself to building architectured solutions. Maybe the testing tools have caught up with other languages now, but you're still testing ugly code.

Javascript is the smallpox of coding languages. Maybe once it's finally eradicated Brendan Eich will only be remembered for this, the equally damaging Rust language and attempting to remove the marriage rights of same-sex couples in California. Hey, did I just politicise Javascript there? Flame on.

Re:I has a sad

[narcc]

So... What's wrong with the language? Or were you just repeating the meme?

Re:I has a sad

[anchovy_chekov]

The fact that there is a need for a book like this: http://shop.oreilly.com/produc...

Maybe? :) Still haven't found a "C - Just the Good Parts" book, but still looking.

Re:I has a sad

[leptons]

Dynamic typing scares him. It scares a lot of people that don't understand how it works or how to use it.

Many people have no problems with Javascript.

Only programmers who cling to their compiler to tell them they did something wrong fear Javascript. I never, ever, have any errors due to an unexpected type, or whatever boogeyman people like @narcc seem to be afraid of.

Or, maybe they are just afraid of Javascript's market share and the possibility that they will have to learn to use it to earn a living someday.

Re:I has a sad

[Anonymous Coward]

narcc loves JavaScript. He hates jQuery.

Re:I has a sad

[PolygamousRanchKid]

I've only run across 2 I've actively disliked.

. . . oh, then you need to meet IBM mainframe JCL . . . it's a hoot and a half . . . then you can ++ your count up to three . . .

Re:I has a sad

[dr_blurb]

The fact that there is a need for a book like this: http://shop.oreilly.com/product/9780596517748.do

Needed, perhaps, but that is a good book, recommended for anyone who wants to delve a little deeper into Javascript.

Re:I has a sad

[wiredlogic]

TCL has a nice simplicity to it in that is has essentially no grammar. Compared to Lisp, Forth, or Postscript it is a nice small language. It's not my go to for everyday scripting but I still appreciate its flexibility. You can redefine everything. You want infix assignment expressions? A simple handler for unknown commands will do that for you. Want a new control structure? Make a command for it and roll.

Re:I has a sad

[Snotnose]

It's been 20 years, but in '94 I spent 6 months writing something in TCL. I don't remember why I ended up hating it so much, but I seem to recall I was never sure if I wanted to use *foo, foo, or &foo in any given situation.

Re:I has a sad

[Tablizer]

Why is that book only 3 pages?

Re:I has a sad

[Anonymous Coward]

That's because C is the answer to 'C++ - Just the Good Parts'.

Re:I has a sad

[Anonymous Coward]

So... What's wrong with the language? Or were you just repeating the meme?

Let's see:
The syntax is fundamentally broken because of Automatic Semicolon Insertion which creates lots of weird edge cases.

The type system is bullshit. There are 2 types of null that have to be checked for separately; JS has primitive types despite being fully object oriented which creates lots of weird things due to boxing and unboxing (Unlike Java, where boxing is explicit so not a problem, JS has no type declarations so boxing is random and has unpredictable consequences), boxed integers let you attach properties like every other object but will then suddenly unbox themselves and drop all the properties.

The structural support syntax is non-existent, classes are based on prototyping but there is no built-in syntax for that which results in massive repetitions of "FunctionName.prototype.property = function(...) {...}" over and over.

The typeof operator is basically retarded, typeof(null) === 'object' but typeof(undefined) === 'undefined'. That's great, null is a valid object so obviously trying to access a property on the null object will work fine, right? If you consider a "variable is null" exception to be "working fine" then sure.

Following from bullshit types, we have bullshit automatic type conversion. What does "1" == true do? Why, it's true, of course! Because "1" converts to integral '1' and true converts to '1' so it's equal. Also, ' \r\t\n ' == 0 is also true because converting a string of white-space to a number gives 0. How useful. Especially since JS is dynamically typed so you frequently end up comparing variables with unknown types and getting wonderfully unpredictable consequences for it.

Multiple fucking equality comparators. We have the roulette wheel comparator I mentioned above "==", and then we have reference comparator "===". Anyone who knows anything about JS knows that you never use double-equals, ever (Except for comparing to null since you either write "x == null" or "x === null || x === undefined", yay 2 null types, just what we always wanted).

The standard library is shit. [5, 10, 1].sort().toString() === '1,10,5'. String comparison of integers. What happened to dynamic typing? Who the fuck knows.

I could go on but just thinking about JS makes me angry and I've had enough.

Re:I has a sad

[Anonymous Coward]

Dynamic typing scares him. It scares a lot of people that don't understand how it works or how to use it.

I understand dynamic typing just fine, I wrote JavaScript software for 9 months. The longer I did it, the more I hated it.

Only programmers who cling to their compiler to tell them they did something wrong fear Javascript. I never, ever, have any errors due to an unexpected type, or whatever boogeyman people like @narcc seem to be afraid of.

You're a liar. "I never mistyped a property on an object ever". Yeah, and I'm the King of the USA. If that's actually true it'll be because you never wrote anything with more than 100 lines. I've written shit in JS with more than 10000 lines and it happens enough to be fucking annoying to have things 'work' but produce incorrect results so that I have to trace through the entire execution to find the problem.

Or, maybe they are just afraid of Javascript's market share and the possibility that they will have to learn to use it to earn a living someday.

What a fucking hipster. The most popular language is plain C. Maybe you should learn C and use that if popularity is so indicative of quality, smartass.

Re:I has a sad

[Anonymous Coward]

Actually, that book (Javascript: The Good Parts) is very analogous to the orignal K&R "C Programming Language" ... very, very useful.
That said, the chapters that talk about the 'bad parts' are in fact sad ...

Re:I has a sad

[Anonymous Coward]

CLEARLY you have never laid eyes on IBM's Job Control Language (JCL). It is the most unholy collection of abuses taken to the extreme! There are masses of arcane features, all of which have a single use, and none of which work with any of the others! Javascript has structure. It might be a security disaster, but at least it has structure.

Re:I has a sad

[Anonymous Coward]

"Nothing in the language lends itself to building architectured solutions."

JavaScript supports object-oriented programming, as well as functional programming. Node.js even has a require function for modular design.

Have you ever actually written JavaScript?

Re:I has a sad

[ndykman]

Some would argue the book you seeking is this: The C Programming Language (2nd Edition)

Re:I has a sad

[Anonymous Coward]

You can make your own by modifying a copy of K&R: Just go to page 247 and use a big black marker to make the section that talks about gets() illegible. There you go, the worst part is gone.

Re:I has a sad

[Anonymous Coward]

I'm pretty sure it's heavily frowned upon to use gotos in C.

Re:I has a sad

[Pherdnut]

That book is written by a guy with a love/hate relationship with JavaScript. I do much appreciate Doug's contributions in regards to Monkey Island and JSON but beyond that I wish he'd just butt out of the vast majority of JS conversations given some of the truly inane stuff he's said about it.

Re:I has a sad

[wirelessduck]

It's called K&R?

Re:I has a sad

[Anonymous Coward]

I share your dislike of TCL, but it's not the worst language I have used. That dubious honour belongs to SNOBOL.

SNOBOL was designed over half a century ago, before people knew any better. It makes heavy use of gotos, global variables, and even assembly-language like fields.

If you ever have to maintain a SNOBOL code base, I suggest you just port it to a real language like AWK or Python. You will likely pick up speed on the pattern matching and your sanity will be much improved.

Re:I has a sad

[Anonymous Coward]

Only if you are a retard.

gotos in C are local to the function, meaning you can goto out of scope.

They are damn useful for cleaning up after an error. It keeps your critical path free of clutter that will cause you to blow up your cache, which will grind your program to a halt.

Just goes to show...

[Foresto]

...a programming language (still) doesn't have to be good in order to see widespread use.

Re:Just goes to show...

[jones_supa]

What would be good?

Re:Just goes to show...

[Anonymous Coward]

What would be good?

What would be good? If we didn't have to regurgitate these language discussions every week. Different programmers use different tools for different jobs - often the tools that they are familiar with or that they prefer. Get over it.

Re:Just goes to show...

[anchovy_chekov]

Oh, you're no fun.. why can't we get all jihaddy from time to time. "The sports team from my general geographical location is better than yours" works for the general public. What's wrong with a bit of rough and tumble over each other's favourite programming language?

Re:Just goes to show...

[Tony Isaac]

What makes a language good?

French is a language that is kept pure by the tight control exercised by the French Academy. It has a unique, pleasant sound. It does have weaknesses, such as the lack of phonetic spelling. But is it good? By many measures, yes. Is it successful? Well, it's not dying, but it's not exactly taking over the world, either.

English is nearly the opposite of French. English is promiscuous, allowing in words from any language that is convenient at the time, and even new made-up words. New words and syntaxes become standard just by being used by enough people. Is it good? Well, it does the job. Is it successful? Yes, certainly, it is a language considered essential in many countries. It may be that the very lack of purity of English is what has made it so successful.

JavaScript is more like English than French. It's not pretty, but it works, and it's everywhere. Many people are trying different things with it, different ways to construct useful software. The very lack of enforcement of structure and looseness of the language may be what accounts for its success.

So is JavaScript good? There is already a massive amount of JavaScript software out there. Much of it is shoddy, but there is a steadily growing amount of high-quality software as well. Measured by that standard of how useful it is, I'd say it certainly is good.

Re:Just goes to show...

[XcepticZP]

I thought that was long ago proven by PHP? If you even want to call it a programming language, that is.

Web Workers

[tepples]

single-core-hogging

I don't see how that needs to be the case. Everything except IE pre-10 supports Web Workers. Firefox is moving toward the process-per-tab model popularized by Chrome, and the refactoring project is codenamed Electrolysis.

Re:Web Workers

[jones_supa]

Maybe he meant that it's ridiculous to begin with that a simple JS application can hog an entire core.

Re:Web Workers

[fast turtle]

yet, any god damn thread that hogs a CPU still locks the entire god damn UI of Firefox because it's not multi-thread capable. Fix the fucking UI and other bugs that sill haven't been fixed insted of trying to keep up with Googles insane numbering.

Re:Web Workers

[Anonymous Coward]

If you have javascript running in 20 to 200 tabs your computer is probably frozen solid already. If the scripts are disabled then you should not be affected.

Re:Web Workers

[Anonymous Brave Guy]

Why should having many tabs open in separate processes cause any problem? No modern operating system will actually require the entire browser codebase to be loaded 20 or 200 times in that situation, and any overheads for context switches should be imperceptible with this kind of application.

In general, the code itself should automatically be shared between those processes. The main things kept separate would be the data in each tab/process and the system resources and permissions.

Those are things you really want to be kept separate, for security and robustness reasons, for example if a process crashes or if the OS kills it for security reasons. Firefox is now many years behind the curve on this issue.

Re:Web Workers

[robsku]

Oh phlueez, I've had similiar tab counts for several (>=5) years now and it hasn't been a problem, though nowdays it does help to be selective about domains you allow scripts from.

Re:Web Workers

[hairyfeet]

I'll get hate from the programmers but fuck it, truth is truth. Wanna know what is wrong with JavaScript? The entire damned concept, that is what!

I mean think about it, if I told you "I have an idea! Just to get some information you are gonna have to run strange code from over a dozen places, you don't know these places, if they have malware, you have to trust ALL these places or you can't have the information" you'd say that is seriously fucking STUPID, wouldn't you? That is how modern ad driven web works, calling shit from God knows where with zero control or fine grained permissions, its all run by default....STUPID.

What we need is either a new language built from the ground up with the realization that there are seriously nasty people out there trying to fuck your systems up, or JavaScript seriously needs a rewrite with the modern situation in mind. As it is now everybody keeps trying to fix the bullet wounds of failed design with bandaids like sandboxing, low rights mode, but at the end of the day the very concept of clicking on a link and suddenly getting code from a dozen sites you don't know and executing that code? its just a dumb idea.

Re:Web Workers

[Tough Love]

Why should having many tabs open in separate processes cause any problem?

A process has a far bigger resource footprint than a thread, with its own page tables, heap, fd table, etc, etc. It's a pretty big multiplier. Against that, we have improved robustness, it better be worth the bloat.

Re:Web Workers

[Anonymous Brave Guy]

A process has a somewhat bigger footprint than a thread, but it's really not a big multiplier at all in this kind of context, where we're talking about RAM measured in GB and running any modern operating system. Right now, Firefox typically grabs hundreds of megabytes of RAM just to start up with a handful of popular plug-ins and load a few simple pages, so I don't think a bit of process overhead is where they need to be worried.

Re:Web Workers

[Darinbob]

I would prefer that every single tab except the visible one by 100% idle, no background loading, no updating of message counts in web mail, etc. It's my computer, I don't want some web developer taking control of it in order to serve me ads.

Re:Web Workers

[Darinbob]

They're pushing ads at you, so to them it is working. It's futile to resist, because HTML5 standardized on that ridiculous underpinning, turning the browser into an application platform (sorry, "app" platform, I didn't mean t to insult real applications with the association).

Re:Web Workers

[Tough Love]

Could it be possible that you are not clear on the concept of "multiplier"? In this case, it means the per-task overhead for a process is a multiple of that for a thread. Of course, the code and data of the application itself may be bloated, but that is a another story.

Re:Web Workers

[Mitchell314]

"Multiplier" is meaningless without some kind of context (no pun intended).

Re:Web Workers

[Anonymous Coward]

You will definitely get hate from programmers.. because you talk without actually knowing what you are talking. THAT is stupid.

Re:Web Workers

[WaywardGeek]

Heck, you're right and didn't even have to go into how f-ed up JavaScript is as a language... "dynamic scoping"... really??? That idea sucked in Lisp and I thought we got past it in the 60's after that cock-up. And now we're supposed to take this bastard child serisouly?

Re:Web Workers

[FuzzNugget]

To be fair, that isn't really the fault of JavaScript in and of itself, it's more of a failing of browser security policies.

If a same-domain-only policy was enforced similarly to as with frames, that would remove a *lot* of the issues you're taking about. Of course, that would introduce quite a number of problems, but I think it would solve more (like, for example, sites don't take ten fucking minutes to load a single page because it's reliant upon a thousand different ad servers, tracking systems... seriously, there are times when my 1080p screen can't contain the NoScript blocked list menu, it's absolutely absurd)

Yeah, you could still do some server trickery to load external scripts, but it would be a pretty big roadblock for incompetent developers who include a million external scripts to deal with a simple task that could handled by... well, not being incompetent.

Re:Web Workers

[Anonymous Brave Guy]

In this case, it means the per-task overhead for a process is a multiple of that for a thread.

If the overheads in question are fixed and non-zero, that is self-evidently true.

However, it isn't very interesting, because what actually matters is the resources required to run the overall Firefox application. It appears that those resources are typically a few orders of magnitude larger than the scale of the thread vs. process overheads, so the important multiplier will be very close to 1 here.

Re:Web Workers

[nitehawk214]

You are right, it shouldn't be a problem. However chrome hogs an unbearable amount of memory on my linux system.

The good news is that when (not if, when) individual tabs crash, either due to html5 sound/video support being shit on linux or my own abuse of the chrome javascript debugger (it still crashes less than firebug)... you can almost always end the single tab's process and continue on without crashing the entire browser. (almost... it still occasionally leaves dead visual artifacts on the screen or hogs up enough memory/cpu to bring a rather powerful multicore machine to its knees)

I can only imagine how horrifyingly bad developing javascript would be on windows. Perhaps chrome's multithreading works fine there, but I don't expect to ever have to find out by doing development on windows.

Re:Web Workers

[Anonymous Coward]

Amen.

Re:Web Workers

[Anonymous Brave Guy]

FWIW, my main development machine is a Windows one, and I find it's refreshingly straightforward to use the built-in developer tools in all the major browsers these days. Mercifully, the age of relying on bloated and bug-ridden add-ons like Firebug is over. :-)

Re: Web Workers

[bloggerhater]

Chrome might have "process per tab," but Firefox creates container processes for plugins. All isolated from the page they are rendering to so if they go down they don't take out Firefox. Might not be as robust, but it works. Also, time to update your knowledge, kittlings. Electrolysis has been in development for almost a year and is available in Firefox nightly builds right now. Still buggy but FF is well on its way.

https://wiki.mozilla.org/Elect...

Re: Web Workers

[bloggerhater]

Chrome might have "process per tab," but Firefox creates container processes for plugins. All isolated from the page they are rendering to so if they go down they don't take out Firefox. Might not be as robust, but it works. Also, time to update your knowledge, kittlings. Electrolysis has been in development for almost a year and is available in Firefox nightly builds right now. Still buggy but FF is well on its way.

https://wiki.mozilla.org/Elect...

Re:Web Workers

[Anonymous Coward]

Java = write once, run anywhere
JavaScript = write once, run everywhere

Re: Web Workers

[CockMonster]

It doesn't work, Flash can still take down Firefox pretty regularly

Re:Web Workers

[ultranova]

In this case, it means the per-task overhead for a process is a multiple of that for a thread. Of course, the code and data of the application itself may be bloated, but that is a another story.

No, it isn't. As is, Firefox with 200 tabs open will consume about 2GB+ (and is unusably slow and on the verge of crashing, presumably due to address space exhaustion). A multi-processed Firefox with 1 MB of overhead per process and one process per tab will consume 2.2GB (and likely still crash in a chain reaction if any of the tabs goes, but that's life with perpetual beta). That's 10% increase in memory use for an absurd amount of assumed overhead. So yes, code and data bloat absolutely matter by making everything else shrink into insignificance by comparison.

Re: Web Workers

[Anonymous Coward]

I am not a Web developer. Why can't you run chrome and Firefox and load the same page and see which one take more resources?

Re: Web Workers

[Anonymous Coward]

Right. Heaven forbid you want pandora to keep playing while you switch away and youtube to keep buffering a stream while you peruse another slashdot article...

Re:Web Workers

[Anonymous Coward]

Lua 5.2 has an interesting system. It uses lexical scoping (as it has for previous versions), but added a special variable called _ENV. Any variable reference not lexically bound uses this _ENV table, which can be manipulated programmatically. _ENV is also lexically bound (you can define as many as you want), but Lua has a concept of "metatables", which means you can hook any _ENV definition and catch key lookups or insertions, allowing you to implement dynamic binding precisely where and when you need it, such as a sandboxed CGI script.

Lua 5.1 used something called function environments, which allowed you to hook non-lexically bound variable lookups and assignments on a per-function basis, but was too fine grained and also too prone to "WTF?" moments as the feature wasn't part of the syntax and not easy to spot when being used. Lexically bound _ENV is a rather elegant idea.

Re:Web Workers

[shutdown -p now]

To be fair, they did realize that dynamic scoping of "this" was a hilariously bad idea. It'll have to stay in the language for method calls to work as they do, but ES6 introduces a new syntax for lambdas that makes it lexically scoped as god intended.

Re: Web Workers

[nitehawk214]

It doesn't work, Flash can still take down Firefox pretty regularly

My Linux machine does not suffer from this... flash does not work at all. And nothing of value was lost.

Re:Web Workers

[Tough Love]

Wow, 10% more memory usage is significant in my books. Maybe that's just me. No wonder nobody cares about light and tight any more.

Re:Web Workers

[Anonymous Coward]

Javascript doesn't have dynamic scope.

Re:Web Workers

[Anonymous Coward]

Developers typically don't make core functionality dependent on malware...

You don't see stuff like

define(['malware'], function () {
});

That's just not a thing. So I'm just not certain your complaint isn't a red herring. JavaScript developers cannot enforce that Ad-block not be installed. And if they're doing dependency management, then making functionality dependent on malware would be a *design choice*.

In which case, JavaScript is the least of your problems. Hello, RSA!

Re:Web Workers

[Anonymous Coward]

And just to add...

It would be great if the bitcoin block chain functioned as an anonymous Web cache.

Hence, rolling apps with calls to third-party CDNs. Again: that is a *design choice*, and it's application design — Web Application Design.

JavaScript is just the language that make the design you described possible. Is it the fault of a language, as a programming language, the fact that it makes certain designs possible?

First off, that's a utilitarian argument, and it isn't even about JavaScript, or formal languages, or analysis, or closures, prototypal inheritance, how it's becoming Pythonic, etc. It's an opinion about how people are using the language, it's purpose in building the Web.

The best response to this persistent wailing is "If it wasn't JavaScript, it'd be Smalltalk." (notice there is not direct assertion that J$ and Smalltalk are even qualitatively similar)

Re:Web Workers

[Anonymous Coward]

I increase memory usage by 10% or slightly more and get 50%+ performance increase.

Less memory usually means slower.

Hopefully this is like Wall Street

[Anonymous Coward]

When the media is awash in tips on making killing in stocks or real estate, that's when you know that the market has reached its height.

If only Guido hadn't blown it with Python.

[Animats]

Arguably, it should have been Python, which is a better language. But Python has a problem. Python's little tin god, Guido von Rossum, is in love with his own implementation, CPython. CPython is a naive interpreter. (A "naive interpreter" is one which does the specified operations in the specified order, with little optimization across operations.) In CPython, everything is a dictionary and a lot of time is spent doing lookups. This allows everything to be dynamic. In Python, one thread can patch objects in another thread while both are running. Objects can gain or lose elements on the fly. Even running code can be "monkey-patched".

The price paid for that excessive dynamism is that a Python compiler is hard to write, and an optimizing Python compiler can't optiimize much. Google tried in-house to make Python faster, and their "Unladed Swallow" failed humililatingly. (A different group at Google then developed Go, aimed at the same problem of producing something good for server-side processing.) The PyPy crowd has tried, hard, to make an optimizing Python compiler, and with an incredible amount of complexity under the hood, has made considerable progress, but not enough that PyPy is used much in production.

Pascal went down for a similar reason. Wirth was in love with his elegant recursive-descent compiler. But it didn't optimize, couldn't handle separate compilation, and had no way to handle errors other than aborting. Python seems to be headed for similar irrelevance. It hasn't even been able to replace Perl, which ought to be as marginal as "awk" by now.

Re:If only Guido hadn't blown it with Python.

[Anonymous Coward]

I think the virtue of python is not in its dynamism but how fast and easy development happens. Some large clusters such as the one CCP uses for EVE-Online runs python without fuss.

Python should be used when developer time costs more than CPU time. As cruel as it is, that means python is a wonderful replacement for client-side code such as web based scripts. That is where python should be going. Bill for the same projects, code faster, and make the client browsers worry about an optimized interpreter.

Re:If only Guido hadn't blown it with Python.

[sourcerror]

Javascript has the same dynamism as Python, why is Javascript still much faster?

Re:If only Guido hadn't blown it with Python.

[StripedCow]

IMHO, Python is a complicated mess compared to Javascript.
And a complicated mess is only warranted if at least it adds type checking.

Re:If only Guido hadn't blown it with Python.

[dkf]

Javascript has the same dynamism as Python, why is Javascript still much faster?

I suspect it is because it is easier to identify a consistent low-level type interpretation — including clearly delineated points where you need to throw the code away and recompile — with Javascript than with Python. That's what you need to do a decent compiler.

Re:If only Guido hadn't blown it with Python.

[jbolden]

Interesting points. I agree that Python is missing the window regarding Perl. What are your thoughts on Ruby's approach?

Re:If only Guido hadn't blown it with Python.

[Anonymous Coward]

Making Python 3 code incompatible with Python 2 code didn't help Python adoption either...what the fuck was he thinking, come on man.

Re:If only Guido hadn't blown it with Python.

[sourcerror]

I don't see how that is true. In Javascript you can add new methods to objects during runtime, so that's the same problem that prevents Python to be properly compiled, isn't it?

Re:If only Guido hadn't blown it with Python.

[Anonymous Coward]

What do you mean by 'separate compilation'? Do you mean like libraries and modules? I'll admit to only ever meaningfully writing Pascal on CP-6, so it may be that LADC included some features not available in TurboPascal.

Re:If only Guido hadn't blown it with Python.

[Anonymous Coward]

Complicated in what way ?

  I might be biased because I learn Python first (then C\C++) but I have not once, in the 3 weeks i'm dedicated to javascript, found it to be "simple".

  Regarding the mess part i'm just going to assume you are trolling.. if python's design is messy than what is JS's ?

Re:If only Guido hadn't blown it with Python.

[Animats]

In Javascript you can add new methods to objects during runtime...

In Python you can do that from another thread. Not that you'd ever want to. But the code has to be prepared for that. In Javascript, you can tell at compile time if the code is doing that to itself.

The optimization killer in Python is that the compiler can never look at a block of code and decide "X can't happen between these two statements". Because in Python, it can. That kills most loop-related optimizations.

Re:If only Guido hadn't blown it with Python.

[Nivag064]

From memory: Pascal was a language meant for training, not for production use!

Re:If only Guido hadn't blown it with Python.

[smellotron]

It hasn't even been able to replace Perl

Python has definitely replaced Perl where I work. It's also very quickly catching up to R (statistical/quantitative analysis), but from what I hear there is still a network effect keeping R on top for bleeding-edge research.

Re:If only Guido hadn't blown it with Python.

[Pherdnut]

You think type checking is more complicated than dynamic?

Re:If only Guido hadn't blown it with Python.

[Animats]

From memory: Pascal was a language meant for training, not for production use!

Most of the original Macintosh applications were written in a slightly modified version of Pascal. Turbo Pascal had wide usage in the DOS era. But the UNIX compilers for Pascal sucked. The Berkeley one was a hack on the C compiler, and made a subroutine call for every subscript check. So it was too slow to use.

Re:If only Guido hadn't blown it with Python.

[shutdown -p now]

Adding methods is a minor ability. In Python, you can do things like get a dictionary of the local variables of the current frame. Or even walk all the frames on the stack and patch them up. Or mutate the bytecode of a running method.

Re:If only Guido hadn't blown it with Python.

[shutdown -p now]

Python is certainly more complicated than JS. But it's also much more consistent. And, ironically, usually less verbose, as well.

Re:If only Guido hadn't blown it with Python.

[Anonymous Coward]

You can't open built in classes...

Re:If only Guido hadn't blown it with Python.

[vilanye]

That isn't where he screwed up.

The problem is that there are two current, incompatible versions.

Ruby EOL'ed 1.8 and Python should have done the same for the 2.x branch.

Worse is better

[tepples]

Read the essay "Worse is better" and you might understand more about what causes this

Re:Worse is better

[gweihir]

This is not "worse is better", what I see. It is "utter crap gets into production and causes serious issues".

Re:Worse is better

[Hatta]

Funny, "worse is better" describes the phenomenon in your signature as well.

Not sure if story is advert for Node.js or Google

[Anonymous Coward]

regardless, is there a co-relation between many independent developers and bloat?
or are you saying this will create more dependent developers and it won't bloat?

Load fast, or user will click Back

[tepples]

Why are some developers obsessed with performance? The user probably isn't. They don't care whether something loads 2 seconds quicker

I thought loading faster was the difference between the user staying on a document and the user hitting the Back button to return to someone else's document. Web search engines have recognized this and have started to penalize slow-loading documents.

Re:Load fast, or user will click Back

[TubeSteak]

If I hit a slow loading site, I immediately try for the google cache because it is almost always responsive.

Re:Load fast, or user will click Back

[Anonymous Coward]

Too bad the asshats at eBay will never see this.
Their site now is a performance abortion.

Re:Load fast, or user will click Back

[reikae]

The back button rarely works on modern Javascript-heavy sites, so that's not a problem.

Re:Load fast, or user will click Back

[Billly Gates]

yeah like 99% of the rest of the internet users, regardless of having a boss who only cares what the SEO ranking by week is when doing performance evaluations.

Re:Load fast, or user will click Back

[Anonymous Coward]

As both a video game nut and a developer: Nothing. pisses me off faster than crap thrughput.

Kinda Suprised...but I guess I shouldn't be...

[neorush]

I'm kind of surprised that there is not one good comment about the benefits of javascript up above this yet. I mean you can off load sooo much data to the client cpu. With the latest in webstorage and the sqlite port to JS I can actually create a friggen database server running on the client. WebRTC and WebSockets are seriously about the change everything in the next 1-2 years....I'm curious how many of the above posts are done by folks who actually do web development? It is pretty much indispensable these days, and really pretty awesome, so get used to it.

Re:Kinda Suprised...but I guess I shouldn't be...

[jeffb (2.718)]

Yes, and thank you for dropping my laptop's battery life from 5 hours to 2 hours.

Tell me, do you work at Gawker?

Re:Kinda Suprised...but I guess I shouldn't be...

[rubycodez]

people like you need to die in a fire. we don't appreciate your rubbish bogging down our machines and opening a multitude of channels for marketing scum's wet dreams

Re:Kinda Suprised...but I guess I shouldn't be...

[laffer1]

Spoken like a true front end developer. You guys don't belong in the backend. Real applications store millions of rows and run on clusters. Grandma's iMac won't cut it.

I don't want to run my whole stack on javascript and worse yet not in the BROWSER. I need speed. I'm processing large amounts of data. I'm trying to build real applications.

Do you really think Facebook or twitter could run in your browser? Seriously?

Re:Kinda Suprised...but I guess I shouldn't be...

[Jartan]

Any language can do that. You're just impressed that javascript has come so far.

The only reason to "get used to it" is that web browsers still aren't supporting other languages.

Re:Kinda Suprised...but I guess I shouldn't be...

[Anonymous Coward]

Do you really think Facebook or twitter could run in your browser? Seriously?

Be careful about giving them ideas. I wouldn't think Facebook could run on PHP either, but here we are.

Re:Kinda Suprised...but I guess I shouldn't be...

[iONiUM]

All the features in JS that the parent is talking about are very important. But they're important and good for web applications, not web sites. Everyone here seems to assume that all people do with HTML/CSS/JS is write web sites, but that's no longer the case. Application which were traditionally written in WinForms/WPF/C++/Mobile apps are starting to get offloaded to web, which means we need those features in JS. I mean, why do you think the canvas tag and others were introduced?

The problem is advertisers/annoying websites are also using them.

Re:Kinda Suprised...but I guess I shouldn't be...

[Admiral_Bob2000]

If you want to offload more bloated bulky Javascript onto my own CPU, you can chip in to my electricity bill. Deal?

Re:Kinda Suprised...but I guess I shouldn't be...

[Anonymous Coward]

Lol your shitty little backend "processing" "millions of rows" is what passes for a "real application" around here these day?

This site has really gone downhill. People have always shit on things they didn't like around here, but at least they used to drop the turds from some kind of height!

Millions of rows. Fuck. You can process millions of rows IN A BROWSER now. Frankly if you can't make that blazing fast on an iMac you're a wastrel the likes of which would make my grandma cringe.

Re:Kinda Suprised...but I guess I shouldn't be...

[neorush]

...and poorly written python sucks up server resources. You can't say a language is bad because there are people out there who don't know how to efficiently leverage it. The reality is it uses less power / network to say calculate an md5 hash in-browser than to send a json or a full http request to the server. Despite that "bulky" inclusion of Crypto.js

Re:Kinda Suprised...but I guess I shouldn't be...

[Anonymous Coward]

If you want to offload more bloated bulky Javascript onto my own CPU, you can chip in to my electricity bill. Deal?

Why? Did he force you to open your browser and type the URL to his web application?

Re:Kinda Suprised...but I guess I shouldn't be...

[Anonymous Coward]

I'm kind of surprised that there is not one good comment about the benefits of javascript up above this yet.

Some of use just know when we are wasting our time arguing against the latest meme. When I started reading slashdot, the memes were Natalie Portman, hot grits, and Cowboy Neal. They were mere annoyances, rather than trolls. Now the memes are things like suggesting how every little thing Google does is somehow evil (it apparently doesn't take much to be considered evil these days...even something like "that service you provided for free and never made any promises about is now going away" is somehow evil). Hating on Javascript is just another of those memes. And they aren't mere annoyances like a First Post. They are childish and malicious. Disregarding the historical perspective, how and why things are the way they were, all of the good aspects that come along with the bad, etc.

Somehow it's just popular to spew hate. Meme's seem to have gotten that way these days. Reminds me very much of the Nickelback meme you used to see all over the place. Was Nickelback the best band? No. Was their music great? No. It's just mediocre pop rock that really should have slipped under most people's radar. But somehow they got crucified as the worst band that ever existed, and idiots were even paying the band for concert tickets just so they could go to their show and boo them.

Javascript has grown a lot over the years and can do some pretty great things. There are decent tools for them, and decent libraries covering almost anything you might want to do. A lot of people are accomplishing a lot of great thing instead of just harping on all of the imperfections. Most are probably too busy to waste their time arguing with trolls.

PS. No, I'm not just dismissing everyone who doesn't like it as a troll. But anyone who simply says "It sucks because of X, Y, and Z" without acknowledging that X, Y, and Z have certain tradeoffs (which make them attractive to some, unattractive to others) is either a troll, or remaining purposefully ignorant (which is close enough to one for me)

Re:Kinda Suprised...but I guess I shouldn't be...

[jeffb (2.718)]

If you're writing and executing sloppy code on your server, it's your problem.

If you're writing sloppy code and executing it on my machine, along with hundreds of thousands of others, you're wasting my resources and theirs.

No, calculating one md5 hash isn't going to make any difference to anyone, no matter where you do it. Doing whatever it is that Gawker's code-bonobos are doing cuts my battery life in half. That's kind of a big deal to me -- enough to make me avoid their sites when I'm running unplugged. And it turns out to be easier just to avoid the sites entirely.

Re:Kinda Suprised...but I guess I shouldn't be...

[Anonymous Coward]

Red herring. You are not really concerned about the electricity your CPU is using.

Your CPU spends most of its cycles waiting for you to do something. If you really want to save on electricity bills, turn off your fucking computer.

javaScript is okay.

[Kremmy]

Platform is the wrong word. Something like node.js might be considered a platform, but not JavaScript itself. JavaScript is flexible, C-like, has first class dictionaries and JSON makes them super simple to serialize. It's one of those languages whose flexibility can actually be a hindrance because you have end up having to get pretty deep to find the structure ... maybe it is a platform.

Re:javaScript is okay.

[Anonymous Coward]

(As someone forced to use Lua in a production environment...)
Javascript is better than Lua - at least JS gets array indexing right, and doesn't require filling your code with Pascal-style begin...end spam. And who the fuck looks for Lua on a resume? I would swap out Lua for JS in a heartbeat.

Re:javaScript is okay.

[TheRealMindChild]

Node.js isn't anything close to a platform. It just serializes/deserialized objects.

Re:javaScript is okay.

[Kremmy]

Yup, and JavaScript is just a programming language.

RE:javaScript is okay.

[Pherdnut]

How, beyond syntax is JS C-like?

Lisp M-expressions

[tepples]

[ECMAScript] tries to look like a member of the Algol tree, but its internals behave more like Lisp.

In other words, it implements the original plan for "M-expressions" in Lisp.

use strict

[tepples]

Errorhandling and exceptions are great, because they don't ignore errors, but force you to deal with them. Javascript just runs and you never know the correctness of the state.

Have you tried to use strict? That should turn more silent failures into the "error handling and exceptions" that you prefer.

Re:use strict

[ebno-10db]

From your link:

Browsers don't reliably implement strict mode yet, so don't blindly depend on it. Strict mode changes semantics. Relying on those changes will cause mistakes and errors in browsers which don't implement strict mode.

It sounds like adding a source of errors more than removing one.

Re:use strict

[Billly Gates]

Doesn't work in IE 6 so it is a no go.

Re:use strict

[shutdown -p now]

Even Microsoft websites no longer support IE6.

Re:use strict

[cwsumner]

Even Microsoft websites no longer support IE6.

And you know what we think of Micro$oft !

JS is a C..P Language - see here why

[Anonymous Coward]

My beef with JS is that it is a regression from the time of PASCAL or Ada (or should I say ALGOL ?) where good programs are nicely structured by means of proper data structures.
This means:
1.) Reliability is unecessarily low. You never know how many feet an object in variable "monkey" might have. Usually two and sometimes just one, zero or 312.
2.) Security: A lack of type-safety regularly results in insecure programs. Plus, it is quite difficult to reason about program safety in absence of strong typing.
3.) Hard to optimize: Strongly typed languages result in programs which can be quite easily optimized for execution efficiency. Compare that to JS, where the optimizer has to "infer" "typical runtime type" and then optimize for that "dynamic" type constellation. Of course, exception code must handle the "differently typed corner cases", too.
That implies: Bloated optimizer/JIT compiler plus all the nice zero-day exploits which result from this. Donning my tinfoil hat I would say JS is a godsend for the TLAs.

In summary: JS is a regression from the state of the art of 1970s. "IT" is actually forgetting the great achievements in Software Engineering that resulted in Algol. With JavaScript, computing has gone even further down the path of sloppiness and anti-reliability. Let's face it: 99% of people are in this profession for the love of money, not for the love of constructing reliable and correct systems.

Dipl.-Ing.(BA) Informationstechnik Frank Gerlach
Gäufelden
Germany

As a counter-concept, look at this invention of mine:http://sourceforge.net/p/sappeurcompiler/code-0/HEAD/tree/trunk/doc/manual.pdf?format=raw

Re:JS is a C..P Language - see here why

[jbolden]

If you are going to argue for a serious counter proposal you may want to get an account.

Regardless the battle between typed and untyped languages on the web was lost during the CGI days when Perl replaced C. Too much of the data coming in is untyped. http://happstack.com/docs/cras...

As for Saupper reading the manual you don't seem to be considering the problem domain at all just creating an alternative strongly typed language with some different features than C++ or Java.

Re:JS is a C..P Language - see here why

[Anonymous Coward]

Sappeur aims to be as type-safe as Java/C# while retaining all the efficient features of C++ (e.g. stack allocation, arrays of objects instead of arrays of references, synchronous destructors,smartpointers/ref counting. ...)

Regarding a SD handle - I had one and used it until the PC brigade modded the handle into oblivion.

Re:JS is a C..P Language - see here why

[Anonymous Coward]

Frank, I fully kicked sappeur's arse when you spammed it all over theregister, don't bring your shite here until you've fixed it (which will be never because you have no formal language theory nor any significant compiler experience. In fact, you don't even understand C++ very well).

Re:JS is a C..P Language - see here why

[Nivag064]

ALGOL, COBOL, FORTRAN, Java, C, C++?

Nah! Too hard, their compilers throw too many errors...

Now JavaScript, I don't need no stinking discipline! Good lets go for it...

I'm having to master JavaScript/AngularJS as part of a JEE7 App. JavaScript is great in some ways, but the lack of rigorous static type and error reporting is a massive handicap for serious development. But then, I understand the value of coding discipline from experience - as I have programmed in all the languages of the first line - mostly COBOL (in the distant past) & Java (since 1997).

Re:JS is a C..P Language - see here why

[Anonymous Coward]

The problem I see with JS is that it did not get to where is at intentionally. In the same way that unintentional coupling is a pain. JS growth into places that it should not be can be extremely painful. It seems we have made the tradeoff of usability against cpu usage and size with the frameworks.

Javascript is paradigm-free

[mark-t]

One thing I personally like about Javascript is that it covers all three of the currently most popular programming paradigms.

You want an imperative style of development? Javascript can do that, check.

You want an object-oriented style of development? Javascript can do that, check.

You want a functional style of development? Javascript can do that too, check.

Some would argue that by covering so many different paradigms, it ends up covering none of them as well as languages that are designed for a specific paradigm from the ground up, and I wouldn't really refute this point... but it easily does all three of them well enough to still be profoundly productive when developing in any of them, and this means that a programmer is relatively free to pick the paradigm that best models the original problem when designing a solution. This, in my experience, results in shorter development cycles, and frequently much less buggy code.

Re:Javascript is paradigm-free

[anchovy_chekov]

Why pick one when you can pick all three in the same application? :)

No, you're absolutely right - being able to choose a mode of programming is neat and Java does lend itself to doing neat things. But it still feels like a language that someone quickly hacked together. And the freedom to pick a paradigm means your fellow coders get to pick whatever happens to be in their clue bucket for the day. At least with a language that focuses on imperative or functional coding you can be reasonably sure that the guy sitting next to you has a similar view of reality as you do. "Multi-paradigm" is a bit like saying "post modern", with all the positive and negative connotations. I prefer my languages neo-classical :)

Re:Javascript is paradigm-free

[anchovy_chekov]

Did I say Java? Sorry.. Javascript. Ah.. so easy to get them confused. Damn you Netscape Marketing Division - you got me again.

Re:Javascript is paradigm-free

[Anonymous Coward]

For some values of "can."

Re:Javascript is paradigm-free

[stox]

Does everything, none of them well.

Re:Javascript is paradigm-free

[mark-t]

It does them well enough to still be useful... which was my point.

Re:Javascript is paradigm-free

[Tablizer]

When you want imperative, you get object-oriented. When you want object-oriented, you get functional. When you want functional, you get.....bugs.

Re:Javascript is paradigm-free

[ndykman]

This is an excellent point. It doesn't pick any particular paradigm to be structured around.

Languages like C# and Java have added functional features, but they integrate with their primary object oriented focus. Compare this to Scheme/LISP, whose object oriented features were integrated with its primary functional focus.

JavaScript is showing the signs of it's hasty development. Things like programming language semantics, type theory and more exist for a reason. Programming language benefit from careful design and analysis of tradeoffs.

Re:JavaScript is paradigm-free

[mark-t]

What does C# have to do with Javascript?

Also, writing strictly imperative code in Javascript is as straightforward as writing it in plain C.

Just avoid the JavaScript trap

[Anonymous Coward]

Read all about it at https://www.gnu.org/philosophy...

The hipsters need to go. Now.

[Anonymous Coward]

I've been in the industry a long time. Decades, in fact. I've worked with many different technologies over that time. Some were good, some were bad. I've worked with many different people over that time. Some were good, some were bad.

But when I look back, the worst of the technologies and the worst of the people came bundled together. They're commonly referred to today as "hipsters", and there's been a helluva lot of them involved with software development since around 2006.

Anyone in Silicon Valley, and in the industry in particular, knows exactly who I'm talking about. It's the guy who refuses to take off his fedora hat, even when meeting important clients and customers. It's the girl who wears huge glasses with no lenses and constantly complains about "sexism" without any justification (even the women in HR, who are typically hyper-sensitive to anything that could potentially be perceived as "sexism" think she's full of shit). It's the androgynous "designer" who constantly talks about "obscure" beer or "ironic" music, and who never manages to do anything productive. It's about the failed athlete who also failed out of business school, then flipped through a book on Ruby on Rails once, and now thinks he's a web developer.

These people have systematically destroyed decades of real growth and real productivity. Generally lacking any formal education in computer science, mathematics, physics, or engineering, they don't realize how wrong they're doing software development. They seriously think that JavaScript, Ruby on Rails, and NoSQL are "good technologies". Of course, anyone with any knowledge and experience sees right through this awful charade these hipsters put on.

The software these people produce is total, complete, utter shit. I can't emphasize this enough. I've seen a lot of really bad code in the past, but none of it compares to what these hipsters crank out. It's unmaintainable, it performs horribly, and it's usually written in a shitty dynamic language so it's rife with errors that any compiler would have caught. Just look at the projects that these people have worked on or been heavily involved with. GNOME 3, for instance. Or Windows 8. Or Web 2.0 sites. All of these are totally crap.

Don't misunderstand me, it's not a problem with young software developers in general. I've had the pleasure of working with many who were very talented, and they have a bright future ahead of them. The problem is squarely with those who fall into the "hipster" camp, regardless of age. They are truly a disease upon our industry, the worst I think it has ever experienced. They need to take their ignorance, stupidity, and extremely crappy technologies with them. We would truly be better off without JavaScript, without NoSQL, without Ruby on Rails, and without the hipsters.

Re:The hipsters need to go. Now.

[jbolden]

As someone who actually has 2 decades in the industry you don't sound like someone with that kind of time at all. Us middle aged folks look at NoSQL and think "Network database" or CODASYL databases are coming back. We weren't attached to Java-EJBs and expected a lighter weight version, Ruby seems like a scripting language...

Stop trying to pretend to be older. Older people saw shit when you were young that isn't used anymore and that's hard to fake.

Re:The hipsters need to go. Now.

[Anonymous Coward]

+5 Insightful if I had mod points...

Re:The hipsters need to go. Now.

[Anonymous Coward]

Isn't Ruby that lousy free variant of VW Smalltalk ?

Re:The hipsters need to go. Now.

[zieroh]

I have decades of industry experience in Silicon Valley as well. And really, your post comes off more as a bunch of whining than an actual critique. There are always going to be new technologies, and the people who are heavily invested in the previous generation of technologies will always groan about these dag-gone kids with all their newfangled ways of doing things. Heck, I've been guilty of this kind of thinking myself.

But you know what? When we entered the industry as young whippersnappers, the previous generation of programmers said the exact same kinds of things about us.

And somehow, despite all that, progress marches on.

Re:The hipsters need to go. Now.

[Jane Q. Public]

"We would truly be better off without JavaScript, without NoSQL, without Ruby on Rails, and without the hipsters."

It's all very well and good to have opinions, but you make some strong specific complaints here without actually giving reasons for those opinions. Many people disagree with you.

What is your problem with JavaScript? JavaScript DOES, in fact, have some very serious problems. For a classic example that was linked to here just yesterday, try

[5, 10, 1].sort();

I happen to share your dislike for JavaScript, but I can give actual reasons for having that opinion.

NoSQL has its problems as well. While for some narrow uses it does offer great performance, if you step outside those narrow bounds even a little bit, you are forced to code many of the functions normally handled by a relational database, manually. I have never had to implement a huge distributed high-performance database, so in my case the NoSQL use-case is limited to very simple local non-relational data stores. But again, I can articulate a reason for not caring for it: the theoretical basis for it is weak, what actually constitute proper use-cases for it are rare, and it creates a lot of manual coding that relational databases handle on their own.

But now we come to Ruby and Rails (which are 2 different things). Yes, Ruby is a dynamic language, with all the (well-known) shortcomings of dynamic languages. But other than that, what do you have against them? Please be specific. We already know you don't like them. The question is: why?

Re:The hipsters need to go. Now.

[Anonymous Coward]

The "hipsters" are generally an improvement over the previous generation of self-taught dipshits who churned out tons of Visual Basic spaghetti crap and worshiped all things Microsoft. Or everyone who wrote a hackjob PHP4 website, such as the mogoloids behind abortions like Drupal. Most of that is due to better tooling and frameworks however.

The programming industry sucks in cheap retards and then spit them out in a few years when their skillsets are obsolete, it is how it has always worked.

Re:The hipsters need to go. Now.

[Joce640k]

It's not just the hipsters, it's also people like Joel Spolsky - the god of hipster programmers.

eg. I've lost count of the number of times I've seen this article referenced on the web and every time I see it I want to bash his head against something solid.

"Making Wrong Code Look Wrong"? Really? How about making wrong code *be* wrong? How about not having it compile at all? How about an SQL library that doesn't accept unsafe strings as parameters?

If you really want to, you could get the compiler to convert unsafe strings to safe strings on the fly ("implicit type conversion") to avoid having to educate the junior programmers.

All SQL injection problems could be eliminated at a stroke, now and forever. How is possible that people aren't doing it?

They're not though, because that would require a strongly typed language...and that's not fashionable.

Joel also rants endlessly against languages with exceptions (there's some of in that article), fails to grok very basic C++ concepts every time they're mentioned (eg. stack unwinding), but the hipsters worship him as some sort of a guru. /rant over

Re:The hipsters need to go. Now.

[tranquilidad]

I started back in 1978 and I remember someone coming in to pitch a database technology in 2006 for which they had patents pending and it would replace relational databases. They kept describing it using catch phrases and turning rows into columns and I just couldn't grasp WTF they were talking about. I finally asked them to draw a picture and they mapped it out on the white board.

I then asked if they had ever read about IMS and hierarchical databases. They had not. I wished them good luck on their patents and sent them packing.

Re:The hipsters need to go. Now.

[Anonymous Coward]

Because they're not /really/ web-scale.

Re:The hipsters need to go. Now.

[Oligonicella]

Heh, the database version of the recent computer sciences grad that applied for a position and was sure he could write a better compiler for us. Probate Court, K.C., IBM mainframe. He took a walk as well.

Re:The hipsters need to go. Now.

[Billly Gates]

But Node.js invented threading and is a webscale and a rockstar language!! non blocking async/IO was invented by javascript. I can get code and can piece it together in parts and run it in parallel. Amazing this was never invented before.

Re:The hipsters need to go. Now.

[Anonymous Coward]

This idea comes back around every few years under a new guise. First it was Object-Oriented databases, then it was XML databases, now it's JSON databases. And every time they come crawling back to SQL.

Re:The hipsters need to go. Now.

[Anonymous Coward]

Did you even bother to read the comment you just replied to? I think you missed each and every point that it made, and replaced them with junk that you just pulled out of your rectum.

The GP makes no mention of 'new' technologies versus 'old' ones. In fact, the ones he'd listed aren't exactly new. JavaScript and Ruby are both nearly 20 years old now. The web is even older. Non-relational databases go back to the 1950s and 1960s. The common thread is that they're bad technologies. Their age doesn't matter.

And the GP explicitly states that this has nothing to do with the ages of the people involved, either. Here, I'll quote it for you, since you apparently can't be bothered to read through a few sentences on your own, without assistance:

Don't misunderstand me, it's not a problem with young software developers in general. I've had the pleasure of working with many who were very talented, and they have a bright future ahead of them. The problem is squarely with those who fall into the "hipster" camp, regardless of age.

Read the comment again, and try to comprehend what it's actually saying.

- Ronald

Re:The hipsters need to go. Now.

[StripedCow]

But Node.js allows only threading without a shared address space.
This means that in order to communicate between threads, you must serialize all your data (!)
Think about the overhead...

Re:The hipsters need to go. Now.

[jbolden]

Ruby Is Smalltalk Minus Minus

Re:The hipsters need to go. Now.

[jbolden]

That's a good story. I've had similar experiences. For example someone who was creating extensions to .pdf to essentially reinvent PostScript.

Re:The hipsters need to go. Now.

[Billly Gates]

But it is FAST AS HELL?!

Re:The hipsters need to go. Now.

[Anonymous Coward]

SQL is slow because it uses joins. No SQL is 1,000% faster! I have seen it with my own eyes and there is no way I will go back to crappy SQL.

Re:The hipsters need to go. Now.

[Billly Gates]

NoSQL has a great future. The article posted with true ACID will be a blessing.

Here is the problem with a crappy ultra expensive solution from Oracle or Microsoft. It can't scale and prices go up on an expontential basis if you try to make it do so. Zdnet a few years ago put in a price tag for running youtube.com on Oracle's database instead of Google's NoSQL solution. The price tag was almost $8,000,000,000!!

NoSQL does not mean no sql. It means not only sql. For quick webscale performance you need low latency in the milliseconds. SQL is not designed for this. Joins slow things incredible and so do threads with blocking I/O which is the appeal of node.js and ngix that use event driven technologies.

It is nice when you have small data and an employer who already has a license but outside of this for a startup traditional relational databases can't cut it for big data on a shoe string budget.

Re:The hipsters need to go. Now.

[colinrichardday]

What if one needs joins?

Re:The hipsters need to go. Now.

[tomhath]

NoSQL is just a cache, except if it gets messed up you're screwed, unlike a cache which can be reinitialized. But if you want to use it, skip the overhead and use BDAM

Re:The hipsters need to go. Now.

[Anonymous Coward]

NoSQL has it's place, but 99% of "hipsters" would be better off denormalizing their data to remove the joins and using a time-tested relational system versus crappy crap like Mongo for their stupid social recipe sharing app. Especially since rdbmses are adding support for json documents which gives you the best of both worlds.

Re:The hipsters need to go. Now.

[Joce640k]

Do they know what denormalizing is?

Re:The hipsters need to go. Now.

[Joce640k]

Please be specific. We already know you don't like them. The question is: why?

Because they trade away everything good computer science strives for in exchange for being easy to type and so that uneducated people can call themselves programmers without having to educate themselves. It's Visual Basic and Excel Macros all over again.

Re:The hipsters need to go. Now.

[Jane Q. Public]

"Here is the problem with a crappy ultra expensive solution from Oracle or Microsoft. It can't scale and prices go up on an expontential basis if you try to make it do so. Zdnet a few years ago put in a price tag for running youtube.com on Oracle's database instead of Google's NoSQL solution. The price tag was almost $8,000,000,000!!"

You don't have to use Oracle's paid version. Almost all you're getting for your money is tech support. (Well, maybe a BIT more, but not much.) MariaDB is plug-and-play compatible (plus some enhanced features) and is free and open-source.

"NoSQL does not mean no sql. It means not only sql. For quick webscale performance you need low latency in the milliseconds."

You mean "for quick, massive-scale performance". Lots of modern, large web businesses use MySQL or PostgreSQL and have no complaints.

And yes, "NoSQL" does mean "no SQL". It is not appropriate to say "SQL is not designed for this", because SQL has nothing to do with that. You were referring to the database engine. SQL is just a query language and has about zero to do with the database engine. But in general, by popular usage, when most people say "NoSQL" what they actually mean is "non-relational", like for example key-value data stores. But while most "NoSQL" databases are non-relational, using "NoSQL" to mean non-relational is just plain technically wrong. I admit it has been a popular thing to do, but so what? Astrology is popular too.

"It is nice when you have small data and an employer who already has a license but outside of this for a startup traditional relational databases can't cut it for big data on a shoe string budget."

Anybody who is trying to design for genuinely "big data" on a "shoestring budget" is an idiot. Amazon is big data. Ebay is big data. Joe's Surplus Warehouse is not big data. If the company becomes wildly successful, it might get to BE big data someday... but for the vast majority of startups, a relational database works just fine.

All this was part of my point: while non-relational databases might be all the rage, their actual justified use-cases are relatively rare. I can easily find you lots of articles by folks telling everybody how their decision to go with a non-relational database turned out to be a very bad decision. But then you can find them yourself easily enough.

Fads are fads, and recognizing them when they come around is a learned skill. Nobody is perfect at it, but getting caught up in one can be costly.

Re:The hipsters need to go. Now.

[Jane Q. Public]

"Because they trade away everything good computer science strives for in exchange for being easy to type and so that uneducated people can call themselves programmers without having to educate themselves."

That's not really an answer. It's as vague as the GP to whom I was replying. What "things"?

Using Ruby as an example (because it was already used here): please explain what "things" it trades away? I am not aware of any. Not counting raw performance. That is one of the mentioned "known issues" with dynamic languages.

But what else? I am interested to know what "things" these are.

Re:The hipsters need to go. Now.

[Anonymous Coward]

SQL defines joins. Joins and processing data slows things down and creates unacceptable latency for web apps. That cost with YouTube was also things like infiband and other things required due to support joins and figure out relationships rather than just serving data.

Re:The hipsters need to go. Now.

[Darinbob]

Actually, Ruby's not a bad language. Though I'd avoid Ruby on Rails like a plague ridden rat carcass.

Re:The hipsters need to go. Now.

[Darinbob]

I once interviewed for a job at a company that made a compiler. They claimed to use techniques borrowed from holography to make their compilation better and faster. They even had a laser lab in the building, and I am not making this up. But I could never get any actual information about what this actually meant without an NDA... Half of me thought it sounded interesting, and the other half thought it was snake oil, and I was never really sure.

Re: The hipsters need to go. Now.

[Anonymous Coward]

PostgresXC says Hi.

Re:The hipsters need to go. Now.

[Anonymous Coward]

Ruby IS a scripting language. Duh

Re:The hipsters need to go. Now.

[Anonymous Coward]

If you think joins slow things down, then you're 1) doing joins wrong, and 2) an idiot.

SQL is Structured Query Language. Here's the structure:

SELECT [things that add up to the resultset row size]
FROM [some table] a
JOIN [another table] b ON [list of conditions to relate "a" and "b", that reduce the number of rows selected from "b", and this works just like a WHERE clause]
repeat above line as needed, with different tables or copies of the same table joined by different conditions...
WHERE [list of conditions that reduce the number of rows selected from any of the above tables, including "a"]

Notice how that "ON [blah]" part works. It reduces the amount of data used to match "b" back to "a".. If you write efficient join conditions, they're not slow at all, and can drastically increase query execution speed compared to matching everything up in the application layer. Doing database operations in the application logic is a rookie mistake.

As for YouTube, I'm sure they've got their crap figured out just fine. But to be perfectly honest, with proper clustering, replication, and tablespace allocations, they could have used a SQL database without any problems, as long as the database engine supports those things. (Read: not MySQL/MariaDB, and probably not PostgreSQL either, unfortunately.)

(And for those that didn't notice, my heavy usage of square brackets does indeed flag me as a SQL Server user.)

TL;DR: Joins are an optimization, not a bottleneck. Unless you're doing it wrong, of course.

Re:The hipsters need to go. Now.

[CrashNBrn]

Because "Web Developers" can't code, since they don't use the Holy C.

Of course one might also consider that "Web Developers" might actually be able to code better than those that "strive for all that is good about coding" since they have to deal with the bullshit that is HTML+CSS and get work done :-)
My favorite usage of C was LPC via DGD. A technically interpreted "Object" C language. It was very elegant compared to many languages that were "invented" to run crap with html. I always felt it was a downright shame that DGD/LPC didn't find that niche. I don't think there would of been a need for JavaScript, PHP, and possibly many other languages that were created (primarily) to deal with HTML.

Re:The hipsters need to go. Now.

[Anonymous Coward]

What is your problem with JavaScript? JavaScript DOES, in fact, have some very serious problems. For a classic example that was linked to here just yesterday, try

[5, 10, 1].sort();

http://www.w3schools.com/jsref/jsref_sort.asp

Default sort order is alphabetic and ascending.

https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Array/sort

The default sort order is lexicographic (not numeric).

granted though the ECMA-Script specification doesn't actually indicate "what" the default behaviour is to be
but a quick "google" will tell you generally what you need to know (browser vs browser vs etc)

i think this is where the old gaurd of programmers is having the most trouble with these newer languages.
javascript is not a "programming" language. it is a "scripting" language designed to work with TEXT inputs.
over the years it's been adapted to handle cases where the input most likely can be interpretted as some binary / numeric value.
but underneath the majority of everything it handles is still TEXT

Re: The hipsters need to go. Now.

[bloggerhater]

"[5, 10, 1].sort();"

You are going to throw down a basic sorting issue as a reason to hate JavaScript??? JS sorts alphabetically. Big deal. Frankly I think one of the biggest problems with JS is teaching people about prototypes and prototype chains. It can be a serious mind fuck for everyone from newbies to experienced programmers who have spent their lives coding in say, C++. I've had better luck teaching binary to six year olds.

Also...why is everyone hating on Rails so bad? It's been fantastic in the few places I've deployed it.

Re:The hipsters need to go. Now.

[Anonymous Coward]

Most compiled languages are faster.

Re:The hipsters need to go. Now.

[Zontar The Mindless]

perl -e 'print join(", ", sort(1,5,10)) . "\n"'

You were saying...? :)

Re:The hipsters need to go. Now.

[ultranova]

But in general, by popular usage, when most people say "NoSQL" what they actually mean is "non-relational", like for example key-value data stores. But while most "NoSQL" databases are non-relational, using "NoSQL" to mean non-relational is just plain technically wrong. I admit it has been a popular thing to do, but so what? Astrology is popular too.

The meaning of words are defined by their popular usage, while the correctness of astrology is determined by whether its predictions are correct to a statistifically significant degree, therefore your analogy is false.

Re:The hipsters need to go. Now.

[Anonymous Coward]

IMHO, the problem with JavaScript's behavior in this example is not the default alphabetical sorting, it's the implicit conversion from numerical to string types.

The whole point of polymorphism is supposed to be that you can use the same method name with behavior appropriate to the data type. If the semantics of a "sort" method on an array of integers implies sorting them as integers, yet they are not, that is broken.

Re:The hipsters need to go. Now.

[Jane Q. Public]

The meaning of words are defined by their popular usage, while the correctness of astrology is determined by whether its predictions are correct to a statistifically significant degree, therefore your analogy is false.

The meaning of words is eventually determined by popular usage. But it doesn't vary as much as you seem to think.

It is quite possible -- even common -- for popular usage to be different from the technical meaning of a word.

Even though current popular slang may say otherwise, NoSQL means "No SQL". Non-relational and "key-value store" mean non-relational. They are NOT the same things.

Further, my last sentence was not intended as "an analogy". It was simply an observation that popularity does not indicate correctness.

Re:The hipsters need to go. Now.

[Flere Imsaho]

Sounds like you're saying you were a coder before being a coder was cool. Typical hipster attitude...

Re:The hipsters need to go. Now.

[Anonymous Coward]

Check out NuoDB. Modern RDBMS that is object-oriented and scalable. Invented by the man who invented Interbase.

Re: The hipsters need to go. Now.

[shutdown -p now]

JS sorts alphabetically. Big deal.

It is kind of a big deal when the common sense dictates that it should work otherwise, and every single other language gets that right.

Re:The hipsters need to go. Now.

[Tablizer]

Some suggest the idea of a database is against the very philosophy of OOP. Databases tend to be attribute-centric, where-as OOP tends to want custom behavioral wrappers around each object: you don't "share verbs" in OO unless an object wants to.

If each object has its own little interface, then you have no standard database operations that apply to all objects, at least not without some sticky compromises. Databases are essentially about standardizing operations on things/entities/objects. OO doesn't like that: each object is "special".

Re: The hipsters need to go. Now.

[bloggerhater]

Whose common sense? Someone that is so thoroughly entrenched in their ways that they are unable to embrace change? Please.

Re: The hipsters need to go. Now.

[shutdown -p now]

Common. You know, that's why it's called "common sense".

There's nothing wrong with change in principle, but only when the change achieves some positive objective, not change for the sake of change. In this particular case, sorting numbers as if they were strings is 1) a change, that 2) serves no useful purpose, and 3) is actively harmful. Why would anyone want to embrace that?

Re: The hipsters need to go. Now.

[Anonymous Coward]

> It is kind of a big deal when the common sense dictates that it should work otherwise, and every single other language gets that right.

Well, to be fair, it's not strictly speaking part of the language. ECMA's documentation says the behavior of sort() is implementation-defined.

Re: The hipsters need to go. Now.

[cwsumner]

There are a finite number of ways to improve things. There is an infinate number of ways to make them worse. People who pick something at random and expect the change to make things better, do not have a good "track record" !

Re:The hipsters need to go. Now.

[Anonymous Coward]

The question is: why?

I'm not the original AC but I can tell you easily - things like Ruby are change for change's sake. Ruby is not significantly superior to the literally hundreds of existing scripting languages. All Ruby and it's ilk are doing is fragmenting the industry into even more competing camps, all with their own libraries, idiosyncrasies, developers and resources. What a huge waste. Scripting/interpreted languages were being written more than 50 years ago. To be rewriting one now is the height of hubris and while it's not completely unreasonable you'd better make damn sure the that the massive costs of moving to yet another scripting language is outweighed by dramatic improvements in productivity. Ruby isn't even close to being a dramatic improvement. The same applies to OOP (versus procedural) and dynamic+TDD (versus statically typed languages) incidentally - look up the research work.

Most of the people involved with these "new and improved" tools seem to be quite out of touch with large body of software knowledge we've painfully accumulated over the decades and these amateurs seem to be hellbent on forcing all of us to endure that pain all over again. Not surprisingly many people object. As just one example I for one now try to ignore anything to do with Ruby because all the Ruby talks I've seen and papers I've read have proven to me that most Ruby'sts are ignorant, don't even realize it and yet try to impose their ignorant beliefs on others. Apart from anything they endlessly obsess over syntax while ignoring far more fundamental issues like productivity, reliability, correctness and reuse.

Re:The hipsters need to go. Now.

[Anonymous Coward]

Smalltalk likes to pretend the underlying OS doesn't exist, so fuck that.

Smalltalk also has the worst syntax this side of Haskell.

Ruby has cleaner syntax and is about the best language to interact with the underlying OS, which is why it is overtaking Perl in sys admin.

Re:The hipsters need to go. Now.

[Anonymous Coward]

Ruby is NOT a scripting language.

It can be used as a scripting language but it is much more than that.

Javascript and Lua are scripting languages, Ruby, Python and Perl are not.

Re:The hipsters need to go. Now.

[Anonymous Coward]

Ruby development started about 21 years ago dipshit.

Re:The hipsters need to go. Now.

[Anonymous Coward]

In other words, you don't know the first thing about Ruby and just prattle on about it without giving any specifics.

Dollar

[Anonymous Coward]

Money explains a lot.

Run to the hills!

[KraxxxZ01]

Try this in your favorite console: [ 10 , 1, 5].sort()
You cringe you lose.

Re:Run to the hills!

[leptons]

The default sort coerces the array items into strings before sorting. If you want to sort integers, use the sort function like this:

[ 10 , 1, 5].sort(function(a,b){
if(a>b){return 1};
if(a<b){return -1};
return 0
})

The sort method in javascript gets it done, but you have to know how it works. Maybe you just don't really know javascript if you cry about this.

Re:Run to the hills!

[zippthorne]

String sort is the proper default, because for the prime uses of javascript, "numbers" should really be thought of and handled as strings, especially when sorting.

If you need it to sort numerically, javascript's array sort method takes an argument for a sort function.

Re:Run to the hills!

[JDG1980]

This is an incredibly non-intuitive way to do things. Sorting an array of numbers shouldn't require jumping through special hoops. This is one of the many reasons why non-typed languages suck.

Re:Run to the hills!

[Anonymous Coward]

One of the biggest issues with javascript is the standard stagnated for so many years and it's missing a lot of really obvious utility functions such as "sortNumber()". The string object in particular sucks compared to every other high-level language.

Re:Run to the hills!

[leptons]

Look, javascript has sort method. In its default use, it sorts strings because that's how it works. If you need to sort other things like complex objects in an array, that can be done too using the same function with logic you supply.

Just saying the whole language "sucks" because it doesn't have x y and z features is a bit disingenuous. Any language you name probably is also missing some things when compared to javascript, so you can't make everyone happy all the time.

Only noobs cry bloody murder when they get the wrong sort order back. It's that way for a reason, and it isn't impossible to sort an array of numbers. Everyone just needs to calm down and maybe learn how something works before declaring "it sucks".

Re:Run to the hills!

[Krishnoid]

The default sort coerces the array items into strings before sorting.

Which then leads to this point. Admittedly, this is a cross-language issue, but considering Javascript's origins are in the browser, it's hard to argue that this was a decision and not an omission.

Re:Run to the hills!

[Anonymous Coward]

Whats wrong with that?
The default is a lexicographic sort.
Try...
[10, 1, 5].sort(function(a, b) {
        return a - b;
});

Re:Run to the hills!

[smellotron]

If you need to sort other things like complex objects in an array

Oh god help us all, complex objects like integers?

Re:Run to the hills!

[kauaidiver]

If you don't like that you won't like:

> [] + []
""

Hmm, I wonder what [] - [] is?
> [] - []
0

How about adding an element to array?
> [1,2,3] + 4
"1,2,34"
And subtracting?
> [1,2,3] - 1
NaN

Come on, jeeze

Re:Run to the hills!

[leptons]

Your ignorance is on display.

Re:Run to the hills!

[Anonymous Coward]

That's bog-standard functional programming. The failure here is the default behavior: coercing to strings first. Any decent language/library would define some kind of Comparable interface with a operator or compare function and leave the comparison semantics up to the objects.

It's even worse than that though: it's actually impossible to sort an array of integers in JavaScript, because JavaScript has no integers! All numbers are floats!

RE:Run to the hills!

[Pherdnut]

That's the default behavior in a language that can mix numbers and strings and then convert them to match dynamically for purposes of comparison in the same array, yes. Not a very hard problem to solve:

[10,1,5].sort(sortNumeric);

function sortNumeric(a,b){ return a-b; }

So what you in ignorance thought was a flaw is actually a demonstration of one of JS's most powerful features. We've been writing functional code long before F# showed up.

Re:Run to the hills!

[shutdown -p now]

All other dynamically typed languages get this right. If you sort a list of strings, they are compared as strings. If you sort a list of numbers, they are compared as numbers. If you have a mix of both, the behavior varies, but e.g. Python will basically always compare two different types in the same way, with no meaningful semantics but guaranteed stability - so you'll get sorted strings followed by sorted numbers, or vice versa, but never interleaving.

The only reason why JS does not do this right because the way its sort method is designed is, well, retarded.

Re:Run to the hills!

[shutdown -p now]

it sorts strings because that's how it works.

"because that's how it works" is not a valid rationale or even excuse. Clearly, the way it works is not the way any sane person would expect it to work. It's also not the way it works in any other language out there.

And it's not just the sort order. It's all the various useless but bug-inducing automatic type conversions that let you do things with obvious semantics but give non-obvious results, like trying to apply + to two arrays - most languages would reject it, some would concatenate them, but only in JS the arrays are stringified and the resulting strings are concatenated. It's just useless, no-one ever wants that or expects that, why is it in the language?

Ditto everything else. Why can you declare a variable anywhere inside the function with "var" (including e.g. inside a loop), but its scope is always the entire function? That's plainly retarded - if you only want function-scoped locals, then only allow "var" at top-level scope. If you allow "var" inside a loop, then it should be scoped to the body of the loop.

Or the dynamic scoping of "this", which makes using closures an exercise in frustration.

Or ...

It's just too many things to list. It's a bad language, period.

Re:Run to the hills!

[Anonymous Coward]

The string object in particular sucks compared to every other high-level language.

Good news, AC! The ES6 standard adds startsWith, endsWith, and contains to String.prototype.

Re:Run to the hills!

[Anonymous Coward]

For a language that was originally intended to be run in a browser along with HTML(text) sorting by string makes perfect sense.

Just because you can understand simple concepts doesn't make it wrong.

As noted, you can easily override that behavior to sort however you like

Re:Run to the hills!

[shutdown -p now]

For a language that was originally intended to be run in a browser along with HTML(text) sorting by string makes perfect sense.

No, it doesn't. For one thing, even when you're rendering text to the user, you want to sort numbers as numbers, not text. Some file managers actually jump through hoops to get this right with file names like "1.txt, 5.txt, 10.txt", so that they're sorted by numeric order.

In any case, this is completely irrelevant here, because JS the language has the notion of data types, including two separate ones for text and numbers. The whole point of having types is to clarify this kind of ambiguous semantics. If I have a bunch of values that I want to be sorted as text, I'll make them be of string type. If what I have are numbers, then I expect them to be treated as numbers in all respects. And they are - if I say 1+2 I get 3, not 12, and if I use regular comparison operators they compare them as numbers, not strings. It's only sort that insists on doing its own way. It's not only inconsistent with sanity; it's inconsistent with the rest of the language.

Why isn't it PEBKA**M**

[zippthorne]

PEBCAK itself is an example of PEBCAK

It implies that the user is a filter that takes input from the chair to input it into the computer, and does not use the monitor feedback system at all. Go ahead and draw the control flow diagram some time.

Re:Why isn't it PEBKA**M**

[Fwipp]

It's talking about physical location. Between the keyboard and the chair, the user is sitting.

Besides, between Keyboard & Monitor, from a control flow diagram, is the computer. :)

Re:Why isn't it PEBKA**M**

[zippthorne]

Both, really. User sees the results of his actions, makes input to change the results to match what he wants. In a sense, the user transforms the screen output into keyboard input with additional input from "imagination." And the computer also transforms the keyboard into into monitor output.

Then there's a wider feedback loop from the results of the user's input affecting the real-world in some way and that information or effect reaching the user who transforms that into a desire for the computer to perform some task, and re-running the computer > monitor > person > keyboard > computer loop.

But putting the chair in the control flow mix implies an open-circuit input from the chair itself. I'm not sure that a chair is wise enough to make good computing decisions.

The lowest common denominator

[Anonymous Coward]

It's the Basic of the web generation. It is scorned, and rightly so.

Re:The lowest common denominator

[leptons]

You mean, scorned by anonymous cowards who haven't learned a bit of the language? Yeah, sounds about right.

Re:Can the Slashdot mobile site get any worse?

[phantomfive]

Yes, it can; check out the beta.

Re:I wrote the front and backend

[ebno-10db]

I think you should stop using computers for a year and meanwhile try to get laid by an actual homo sapiens woman.

Why are you in favor of inter-species mating?

Re:Can the Slashdot mobile site get any worse?

[anchovy_chekov]

Yeah.. wassup with that? The mobile version of slash seems to have gone away the last few days. Apologies for remaining off topic.

Re:Can the Slashdot mobile site get any worse?

[jones_supa]

Sorry for posting this off-topic but there is no other place to post this

feedback@slashdot.org

Javascript haters need to let it go.

[Anonymous Coward]

Most of the commenters must be stuck in 2000 as far is web development is concerned. Between jslint, phantom js, and jasmine you can have syntax validation and can use test driven development for every bit of your front end javascript code. Combine that with require js for managing dependencies along with jquery and knockout and ill write web apps that are more efficient, easier to manage, have better abstraction, are more reliable, and do it faster than pure server side web apps could ever accomplish. If your stuck in the mindset of server side mvc or god help us some proprietary web form garbage your pigeon holing yourself and will be irrelevant soon enough. A master programmer knows a languages strengths and weaknesses and uses the right tool for the job.

Re:Javascript haters need to let it go.

[anchovy_chekov]

You are wise and correct. A master programmer understands his place within the Tao.

http://www.canonical.org/~krag...

"Though a program be but three lines long, someday it will have to be maintained.''

why don't browsers block the popups anymore?

[Anonymous Coward]

Remember when checking the "block popups" box actually worked? there is these annoying new kind of javascript popup that doesn't technically open a new window but just takes focus of the browser and puts itself over the content, actually worse than a old style popup! can't browser block this crap? and even the "old school popup" is making a comeback...why are browsers no longer able to stop sites from opening new windows? i'm sure there are plugins that can do it but why should i need a plugin when most browsers claim to "block popups"? what's next blinking text making a comeback? also, how is it that those circular javascrip dialogs "are you sure you want to leave the site?" that go in a circle and make it impossible to close the window should be easy enough to block but for some reason i get stuck with those regularly enough which requires me to open a console and kill the process. The continued rise of javascipt wouldn't suck so bad browser would actually stay on top of blocking the garbage.

Re:why don't browsers block the popups anymore?

[wiredlogic]

The in-browser popups are rendered using DHTML floating regions. Since there are legitimate uses for the technology it isn't easy for browsers to identify the offending cases. The only guaranteed way to block them is with noscript.

"Progress"

[Anonymous Coward]

Knowing the internals of the JS optimizers leaves me unable to think of JS being "progress" as compared to stronly typed, imperative languages.
Rather the word "clusterfuck" comes to mind.

Frank Gerlach (see my other posts here)

Re:"Progress"

[zieroh]

And that very well may be. The main fallacy of your argument, though, is that "progress" can only be limited to a single metric.

JavaScript is not a first class language

[techfilz]

End of story.

Re:JavaScript is not a first class language

[Anonymous Coward]

butter always gives me moist farts also ur a nigger

Re:JavaScript is not a first class language

[PolygamousRanchKid]

Abraham Lincoln said of someone he detested: "He's a first class, second class, man."

You could rephrase that to say of JavaScript: "It's a first class, second class, language."

My gripe with JavaScript is that it enables people who shouldn't be programming to produce code. And this leads to negative productivity. It takes one of your good programmers a long time to debug something written by one of your bad programmers.

With JavaScript, you can create pretty mock-ups and prototypes very fast. Then the sales executive screams, "Ship it!" And then it takes you a long time to convince the executive that this app would break, fall over and die in production. And that a robust version of the app will take some more serious work.

Re:Can the Slashdot mobile site get any worse?

[Anonymous Coward]

shit bro ever try using slashdot in a text browser? equally ass experience, the stupid links on the left side that no one clicked on ever take about 30 seconds to tab past. for a supposedly "open source" friendly site deliberately shafting text browser users like that is a dick move, yeah i know you don't get ad revenue from text browsers but come on if this is truly a "nerds" site then you will not go out of your way to shaft the poor sap stuck in the server room with nothing but a text terminal.

Niklaus Wirth Wrote it

[Anonymous Coward]

It is called "programming in C using the PASCAL compiler".

Browser bytecode engine instead of a language

[kervin]

It's sad we're still using a single client side language instead of having the option of running bytecode in the browser. Obscurated JS is just as difficult to read as bytecode, and the browser can also have an automatic bytecode to text "viewer".

Browers should get behind pNaCL or something similar.

Re:Browser bytecode engine instead of a language

[Anonymous Coward]

Just think of JS as a form of bytecode. It won't parse as quickly and it has features that make its execution speed slower than it could be, but fundamentally it will serve the purpose and it's already available everywhere.

With Java, there's a lot of checking...

[Anonymous Coward]

that the compiler does, so at least you have to get your parameter
types right in the function call before it will compile. Plus you can
use interfaces and abstract classes to pretty much define an overall
structure and force everyone to do it roughly the same way. We've
been maintaining a quarter-of-a-million line codebase for 5+ years
now, with up to 6 active developers, and it holds up pretty well.

Re:With Java, there's a lot of checking...

[rubycodez]

that has nothing to do with the security exploits, privacy invading and cpu wasting marketing technology, very low bar for coder proficiency that the rubbish known as javascript enables.

Only game in town

[JoshuaRose]

Javascript is certainly leagues behind many other languages in terms of ellegance but it has one overwhelming advantage. It's the only option on the client that is generally supported. The javascript juggernaut rolls on because we can't write modern web applications any other way. Since most developers are getting forced into writing their client application almost entirely in javascript, it isn't too much of a stretch to throw in node and mongo to take it full stack, the performance and ease becomes pretty compelling at that point.

Re:Only game in town

[rubycodez]

people are learning to turn that rubbish off

I'm not the poster, but...

[Anonymous Coward]

...I guess, the situation is similar.
I wonder if Julia can fix some of the mess.

Re:I'm not the poster, but...

[syockit]

I hope not. I consider myself a rabid Julia fan, and for the past month I've been doing mostly Julia daily. But I don't want to see the language get dumbed down in order to accomodate general purpose programming, as it's proved to be stellar in what it's made for: numerical/technical computing.

not correct

[Anonymous Coward]

you can easily have sql injections in your c or c++ code if you are lazy. and prepared statements don't cut it in every case. think of dynamically adding query conditions in your c program.

Re:not correct

[Joce640k]

you can easily have sql injections in your c or c++ code if you are lazy.

Not if your SQL API only accepts parameters of type "sql_safe_string" (or whatever you want to call them)

sql_safe_string would be like std::string but all assignment to it will run the string through a sanitizer.

Passing any other type of string to the API would give an error when you compile the program. The compiler doesn't have a bad hair day or need you to be constantly alert in case a std::string slips though.

(PS: None of this applies to C, obviously, and your continued grouping of C with C++ is revealing...)

Re:not correct

[beelsebob]

No idea why you say none of this applies to C, this can trivially be done there too.

Re:not correct

[Joce640k]

OK, *some* of it applies to C.

C can't do it as anywhere near as neatly or safely as C++ though because it has no concept of privacy or encapsulation.

You also couldn't do things like implicit conversions to sql_safe_string to make it more friendly (assuming you want that).

(And Joel is still wrong and Javascript still rubbish, which is more important than starting a C vs. C++ thread...)

Re:not correct

[Anonymous Coward]

Realistically you use what your PHB decides and that is usually ado.net or some Oracle api. My info might be outdated by this sql_safe_string stuff is not part of SQL Server or Oracle and therefore doesn't exist in the eyes of the business world.

Re:not correct

[Joce640k]

Realistically you use what your PHB decides and that is usually ado.net or some Oracle api. My info might be outdated by this sql_safe_string stuff is not part of SQL Server or Oracle and therefore doesn't exist in the eyes of the business world.

My bad. The "SQL" example required you to follow a link and read something.

It's not about SQL, it's about how weak languages lead to problems like (for example) SQL injection vulnerabilities. Read the page and weep over the method being proposed to "solve" this program.

Re:not correct

[beelsebob]

OK, *some* of it applies to C.

C can't do it as anywhere near as neatly or safely as C++ though because it has no concept of privacy or encapsulation.

Sure there's a concept of privacy –simply don't let the compiler see the private stuff (i.e. don't put it in your header), and of course there's a concept of encapsulation – it's called writing functions that accept a data structure as their first argument, and only exposing the public ones via headers. C actually does a much *better* job of privacy and encapsulation than C++, as C++ requires you to write about private implementation details in public places (the header).

You also couldn't do things like implicit conversions to sql_safe_string to make it more friendly (assuming you want that).

You certainly can't do an implicit conversion, no. You can of course do an explicit one, but this is true of C++ too, and of course the programmer is a moron if they do this.

(And Joel is still wrong and Javascript still rubbish, which is more important than starting a C vs. C++ thread...)

Agreed, this is literally the example I use to demonstrate to people why they want a more strongly typed language than the one they're using. People seem not to realise that the type system can prove a lot more things than simply "you're passing a boolean to that if statement", and when exploited correctly can be made to prove some seriously important safety properties of your program. Another example that I use (which is very similar) is having a file transmitted over the network, and requiring verification that the file is actually valid, and is not giving all kinds of crazy values we don't expect, and may crash us. We can then trivially use the type system to prove that that verification actually happens before doing any real work with the file.

Re:not correct

[Anonymous Coward]

Ruby is more strongly typed than Java, C and C++...

Because of the JIT?

[Anonymous Coward]

Are you nailed to Lua for speed?

Pot - Kettle

[Anonymous Coward]

I have decades of industry experience in Silicon Valley as well. And really, your post comes off more as a bunch of whining than an actual critique.

wait....someone from Silicon Valley complaining about whining?

That's rich.

Re:Can the Slashdot mobile site get any worse?

[Anonymous Coward]

Is mobile.slashdot.org mandatory now? I was able to switch back to classic view on my galaxy nexus...but that was awhile ago.

JavaScript is FUCKING AWFUL

[Anonymous Coward]

JavaScript is FUCKING AWFUL, to use it in SMALL doses, fine, to use it for big projects, OH FUCK OFF:

It is the worst langauge I have ever had the displeasure to use. God knows why it is a standard given that there are BETTER languages out there.

I am a big believer in DSL's (Domain Specific Lanugages) and General Purpose Languages/Computing is really a major trade off and handicap.

If you want to do a job right, DO IT RIGHT! (Or hire a woman).

Re:JavaScript is FUCKING AWFUL

[leptons]

Says the anonymous coward.

Re:JavaScript is FUCKING AWFUL

[Anonymous Coward]

You want AWFUL, try one of the frameworks built on top of Javascript...DOJO comes to mind.

Re:JavaScript is FUCKING AWFUL

[Guspaz]

So don't use Javascript. Treat it like assembly and write it in something else like C++ that gets compiled down to Javascript.

SD handle?

[Anonymous Coward]

What's an SD handle? Please enlighten me, thanks.

Re:SD handle?

[Anonymous Coward]

> What's an SD handle? Please enlighten me, thanks.

SlashDot account name.

Re:SD handle?

[Anonymous Coward]

Slashdot Handle.

Re:SD handle?

[Anonymous Coward]

Thanks.

However, I fail to see how Sappeur is a solution to the devastated webstack. It's not like it's only JS.

Re:SD handle?

[colinrichardday]

SlashDot?

Java will soon die a horrible death!

[Anonymous Coward]

Nobody here seems to be talking about the hacking and vulnerabilities that JavaScript brings to the table. Even slashdot had an article about what some security researchers are seeing - and they point out that JavaScript is part of the security problem. As more hacks like Target, Neiman Marcus, Yahoo and many others come to light, then the role of JavaScript in the hacks will also come to light. The general public will respond with a push towards more security conscious behaviour and JavaScript will be out the door.

Re:Java will soon die a horrible death!

[Anonymous Coward]

Ooops - subject should say "JavaScript" - oh, well, as an AC, I can't edit that :-)

Yes. And there are other tools

[Anonymous Coward]

For example, the Scots created this wonderful drink to help with bitching wives.

And it helps whenever JS causes Firefox to hang - like it just did while posting this comment.

Oop! Need more ice....

Septic tank promotional design pattern

[Anonymous Coward]

Septic tank promotional design pattern where the a largest shits / turds rise to the top. Same as working at Microsoft.

Without hipsters, we would be using Mainframes

[Marrow]

And hating every minute of it. I have as much time in as anyone, and I am thankful for people who think outside the box. Heck, free software and linux are "hip" and the only thing that kept MS from gobbling the world.
The non-hipster people wrote OS/2, and maybe it was fine, but it was stagnant and smelled that way. So thanks for your opinion, but I will still give props where they are due to the girl with the glasses and the guy with the fedora hat.
Thanks.

popularity brings haters

[Anonymous Coward]

Name any language that gets popular, and the haters came out to flame it. Perl, JavaScript, whatever. The only worthwhile languages are the ones no one uses, like Erlang.

Where is the solution?

[MichaelHunter]

If in a room full of developers, I am an "average" developer, but I also tend to approach things logically. I do my research and choose the appropriate tool to get the job done efficiently, with code that is generally accepted by the greatest user base. My clients would not want me to hand them over a product that could not be easily understood and modified by other average developers, "Anyone know a Python developer?", "No, but I know someone that works with Ruby"... In the real world this is incredibly ridiculous. The term "average developer" is also looked upon poorly on this board, and unfortunately, this being 2014, people should not require a 4 year degree to code, code should have evolved enough by now that average people should be able to put words and thoughts into simple code and have it do what is intended. It is the hardcore developers that overcomplicate what should be incredibly simple. A business term, "Where there is pain, there is opportunity", and I hear all of the time, so-called hard-core programmers with their noses up complaining about JS. I come from a world where you do not complain about something without presenting a solution, and of all these posts there is hardly anything worthy of a "solution". Apparently there is a lot of pain towards JS, so maybe one of you smart people should chase the opportunity to build a solution; just understand that no matter how perfect you think your alternative is, if no one uses it, it is useless. I use JS as a carpenter would use his hammer, I don't complain about the hammer, but if one of you smart folk were to build a better hammer that saved me time and money, I would surely use it.

Re:Where is the solution?

[Anonymous Coward]

Maybe you educate yourself about Ada, Delphi or FreePascal. And I mean REAL education - reading lots of papers by EXPERTS as opposed to business whores. Read Wirth, for example. Learn about Cobol and how it failed to deliever "simplicity".

Learn why FORTRAN is one of the first and still one of the best languages !

Computing has not become easier, but harder as users expect lots of functionality, yesterday and with "total quality".

For starters, read all the comments here and follow the links. Try to appreciate what people are saying.

But be warned, to succeed in this you need to become an engineer and currently you sound like one of these MBA cynics who have only one love: MONEY.

Re:Where is the solution?

[Todd Knarr]

The complaint, though, is that people aren't using Javascript as a carpenter uses a hammer. They're using it like a carpenter would use a hammer to mill and tool a high-performance aluminum engine block. Of course your first thought would be "DUH! A carpenter's hammer's the wrong tool for that job, you need a CNC mill for that.". Which is the complaint's point. And we already have plenty of the right tools for that job, the problem is that Javscript developers go "But those tools are too complicated!". No, it's not the tools that're complicated, it's the job that's complicated. That's not what they want to hear, though.

Bear in mind that if you can approach a problem and work through it logically, you are not an average developer, not by a long shot. The ability to do that puts you in the top 25% of developers right off. Witness the question over on Ars Technica where a developer seriously asked, in essence, why we divide applications into modules with defined APIs and hide the implementation details of a module from the components that use it. This... should have been covered way back in introductory programming classes, and anyone who's been developing software for any length of time should be able to rattle off the reasoning behind it. Yet here we have Jason Swett writing Web applications without enough understanding of the concept to know why we do it that way. He's typical of at least half the developers out there today: they do cargo-cult programming, following a set of directions without understanding what it is exactly that they're doing.

Who cares?

[Guspaz]

Seriously, does it really matter anymore? You want to write in JavaScript? Do it. You want to write in something else? Most other languages can be compiled down to Javascript without any major performance penalties. Basically, these days you can do whatever you want in whatever language you want, so what does it matter in the end?

Not what it was designed for

[Anonymous Coward]

The problem is JavaScript was only really designed to do small simple things, but just like the shitpile that is X86 it's been extended and kludged in ways that I'm sure the creators never intended or imagined and nowadays it is being used to run game engines, database engines, emulators and virtual machines!

The fact that you NEED a browser with a good JIT to visit JavaScript-heavy pages without responsiveness turning to molasses just shows how out of control things have gone.

It's only because we have such a ridiculous excess of CPU power that people can get away with this; It's bad enough that the web is so polluted with crap that it would take several minutes for the average web page to load over a modem or non-G3/LTE phone, but that we need accelerated or highly clocked CPUs and craptons of memory just to be able to render and run many pages makes me want to scream!

Re:Not what it was designed for

[Anonymous Coward]

What always gets me is all the code that has to load and be in memory in order to run some Javascript that compiled would be a cache line worth of machine code.

That's why I use NoScript of FireFox

[Paul Fernhout]

http://noscript.net/

But it still is a bit of a kludge, compared to your broader insightful point. Chrome has something similar, but I don't think it is as good.

Re:That's why I use NoScript of FireFox

[Joce640k]

Chrome's Javascript blocking is a joke. It's almost as if Google doesn't want you to block it....

Strong types

[MichaelSmith]

Please give this language strong types. When you have megabytes of code and thousands of functions you can not rely on run time type checking. Its just too dangerous to do that.

Re:Strong types

[Anonymous Coward]

You misspelled static typing.

Both Python and Ruby are strongly typed.

Optimization keywords

[Anonymous Coward]

IMHO the problem with all these dynamic languages is that they don't have optimization keywords. You can "monkey patch" everything. You need a way to define "locked" classes in the language. Maybe the regular syntax for defining a class, but "LockedClass" instead of "Class" or something like that.

The Ruby guy I know always drops down to C after profiling the parts that aren't fast enough in Ruby. Ruby stays the glue, and the he uses whatever native code mechanism it has for the parts that need to be fast.

If the language itself had optimization keywords though, he could just stay in Ruby. C programmers hardly ever reach for assembly these days; but all the dynamic guys are reaching for C or C++ all the time after profiling.

That leads to the question--Should we introduce optimization keywords to dynamic languages, or dynamic constructs to compiled languages. Maybe the answer is "yes". Of course you can shoot yourself in the foot with dynamic code. You can write really convoluted math equations too; but nobody blames mathematical notation for papers that are incomprehensible gibberish. They blame the author, and rightly so.

The choice of advertisers we are doomed with

[Anonymous Coward]

Java can go diaf

QWERTY

[Tablizer]

It's the QWERTY of programming languages: you can't toss it because everyone would have to toss it at the same time for the transition to work, and that requires either sorcery or an apocalypse.

Re:QWERTY

[Anonymous Coward]

Actually the flip side of the metaphor fits better (both ways) - what alternatives there are haven't been demonstrated to be any better, but people keep claiming it despite that lack of evidence.

Re:That's why I use NoScript on FireFox

[Paul Fernhout]

I had something like "NotScript" for Chrome in mind as opposed to Chrome's block all JavaScript option. But "NotScript" for Chrome does not have all the same features of NoScript (which inspired it) for FireFox. Related: http://hackademix.net/2009/12/...

I've been writing javascript since 1997

[Anonymous Coward]

And I shit all over your gay nigger framework

Amazing JavaScript stuff by same sql.js author

[Paul Fernhout]

Lua: http://kripken.github.io/lua.v...
"Lua is implemented in portable C. It is possible to run C compiled to JavaScript at speeds approaching that of a native build (using the asm.js subset of JavaScript), which means that you can in principle run C code that happens to implement a VM at high speed as well. Of course this is theoretical until it is actually attempted - that is the point of this project."

A Sql.js demo: http://kripken.github.io/sql.j...

3D, just amazing Doom-clone: https://developer.mozilla.org/...
"BananaBread is a 3D first person shooter that runs on the web. It takes the Cube 2: Sauerbraten engine, which is written in C++ and OpenGL, and compiles it using Emscripten into JavaScript and WebGL so that it can run in modern browsers using standards-based web APIs and without the need for plugins. The project has several goals. First, to serve as a testcase for running a demanding 3D game in browsers: Having a working testcase lets us try out new browser features and to profile performance in order to make browsers faster. Another goal is to prove that games of this nature can run in JavaScript and WebGL, which many people are skeptical about. Finally, all the code in this project is open (and practically all the art assets), so others can learn from this effort and use this code to create their own browser games. The latest update of this demo uses asm.js for additional speed, and WebRTC for multiplayer."

The author's GitHub site, where there is a tool to compile LLVM output like from C to JavaScript: https://github.com/kripken
https://github.com/kripken/ems...

By others (MineCraft-like): http://voxeljs.com/

It's been said JavaScript is much better than we deserved... It's great to see all these advances. And I think you are right, the next two years will see the further spread of all this.

My own JavaScript experiments towards a social semantic desktop, with the idea that you could have a simple backend and do most of the heavy lifting of processing and displaying information locally in the browser.
https://github.com/pdfernhout/...

Re:Amazing JavaScript stuff by same sql.js author

[Anonymous Coward]

www.lua.org/

Re:jslinux

[Anonymous Coward]

Linux Kernel Running in PC Emulator Written in JavaScript...!
http://www.geekosystem.com/linux-kernel-pc-emulator-javascript/

Re:Javascript FTW

[Anonymous Coward]

var myList=[5,10,1].sort(function(a,b){return a-b});
var myList2=[5,10,1].sort(function(a,b){return b-a});

um, what did you read?

[Anonymous Coward]

The article was about Sam Newman cracking on the JavaScript is a platform (ffs), but somehow all the comments are about SQL vs yo mamma and "I've worked in silicon valley, sine before you were a twinkle in her daddies eye"..

41.7 ms timebase of film

[tepples]

In motion pictures, each frame is displayed for 41.7 ms. How can lip sync even be expressed to a precision of less than one frame?

Re:41.7 ms timebase of film

[jbolden]

The sound has to be clicking at a minimum of 0.05ms and generally is more like .02ms. The eye is filling in the picture and the sound needs to be synchronized with the position of the lips as it is filled in. So for example if a character starts speaking and the frame on the screen has his lips closed that's going to be noticeable. If the lips are varying their positions in line with the sound then the brain will accept the illusion.

When JavaScript is your hammer...

[jeffb (2.718)]

I use JS as a carpenter would use his hammer...

And most posters here feel that, to paraphrase the old saw about C++, "When JavaScript is your hammer, everything looks like a thumb."

Nope, not repulsive

[Anonymous Coward]

I've been programming for 30 years. Started in 6502 assembly. Have programmed in more languages than I can remember. I love JavaScript, and I use it professionally every day. Is it the best language? No. Am I still waiting for the one new programming language variant that is as awesome as JavaScript, but does everything better? Yes. But I'll happily keep using JavaScript until the better variant becomes a global standard.

And yes, I write very good code.

Re:Nope, not repulsive

[leptons]

I started on 6502 and I've been coding for 35 years. I love javascript and have written 100s of thousands of lines of javascript. I don't really need it to change to get useful things done with it.

Lua is okay too

[Anonymous Coward]

Lua lets you do some pretty amazing things using tables and meta-methods, but at the end of the day, I keep wishing it did things more like JavaScript.

"fellow coders"

[Anonymous Coward]

Actually, you have a pretty good point. I've spent about five years really honing my skills in JavaScript, to the point of reading all the books on design patterns in JavaScript, object-oriented programming in JavaScript, etc. There really are many good ways to use the language in a variety of different programming contexts. However, professionally I work with several people who haven't written much beyond simple event handlers in JavaScript, or threw together some JQuery stuff without realizing how JavaScript works. And there's pretty much no way to get any of those people, under tight deadlines, to spend any more time learning about the language to not implement an ugly solution.

Sort it out

[Anonymous Coward]

Are you objecting to the fact that the default sort method casts arguments to strings during the sorting?

Lowest?

[Anonymous Coward]

JavaScript really has nothing in common with Basic.

That's a little harsh...

[ralphtheraccoon]

That's a little harsh. Javascript, as as language, isn't *that* bad. It's got plenty of annoying things, and plenty of warts, but as a language, it's workable with.

The problems you mention are mostly solvable. "use strict" solves a bunch of them, and then use jslint (or similar) to catch the bugs your compiler normally would.

If you really want a compiler, and to get away from some of JS's more annoying bits, then switch to coffeescript (which compiles to JS).

Debatable

[Anonymous Coward]

There are very good reasons for "mobile code" to be a high-level programming language on par with (say) Algol. Low-level approaches such as NaCL have to perform extensive low-level operations to ensure nobody can break out of the execution sandbox. A type-safe high-level language just needs some sort of "private" specifier (as in C++ classes) to ensure containment of untrusted code. In addition to code controlling resource access, which all approaches require.

Non-type safe languages like C and C++ are a major enabler of the "cyber warfare domain". They are a regression from the days of Algol and their use should be discontinued. At least in a rational, peaceful world.

We just don't necessarily require Java or JavaScript for this purpose. A "completely memory safe Pascal" would fit these requirements, too. And it could be compiled into highly efficient machine language "just before execution". We could have type-safe multithreading, too.

Please refer to my language "Sappeur" for more on this (URL in another posting on this discussion page).

Garbage collection could be done using reference counting and synchronous destructors. That is an easy approach to enable (soft) realtime capability. Which is critical for high-fidelity GUI programs.

Frank Gerlach

Oh really ?

[Anonymous Coward]

Maybe you could be more specific instead of throwing mud. Or is it F.U.D. ?

All the things I wanted to technically achieve (memory safety, efficiency, security, type support for multithreading) with Sappeur are achieved. It might not look pretty, but it is effective.

I'll leave "making it pretty" to people without an opinion and rational thought.

Re: The hipsets need to go. Now.

[Anonymous Coward]

I can't claim 'decades in the industry' only a half decade at best. I'm a niche developer, DSP Engineer to be specific, but I'd like to think I have enough insight to judge a bad dev from a good one. It's not that difficult to be honest. I've never met any of the types of characters that you describe in this article. Not a single time. Most of these people won't even make it to a face to face interview and even if they did they'd typically be speaking to a good dev, someone who could see through the bs facade they'd be throwing up.

Most of the devs I work with come in with tassled hair, wearing clothes their wifes/fiance/girlfriends purchase for them. They eat virtually whatever food/beer you put in front of them as they're generally too busy to be picky about these things. They don't complain about the state of the world as it pertains to androgynous behavior/sexism, they complain about Microsoft not following standards or that someone neglected to use std::unique_ptr somewhere.

Javascript, Ruby on Rails and even GNOME 3 are all tools. All of which were developed in the face of what appeared to be some shortcoming(s) of their predecessors. The tools are only as good as the person utilizing them. Clearly you've had the misfortune of working with some imaginary people who are very inept at using these tools, and hence your rant.

Pascal, Wirth, IMPLEMENTATION

[Anonymous Coward]

Nothing stops a well-funded and experiened software engineer from building a highly optimizing Pascal compiler which can restart after compilation errors. It was just that Mr Wirth focused on the concept, not an American-style implementation.

C and C++ are rotten on a conceptual level, while Mr Wirth's Pascal compiler might be not polished yet.

The FreePascal guys seem to be quite close to C++ already:

http://benchmarksgame.alioth.debian.org/u32/pascal.php

GNAT Ada, another Algol-style language, seems to be even close to C++:

http://benchmarksgame.alioth.debian.org/u32/ada.php

Fortran (which is also close to Algol than C++) seems to be highly competitve, too:

http://benchmarksgame.alioth.debian.org/u32/benchmark.php?test=all&lang=ifc&lang2=gpp&data=u32

Frank Gerlach

Google's Money

[Anonymous Coward]

Both Chrome and Firefox got massive Google investments to "polish the turd". Think of 1000s of millions.

What they do is to retro-actively try to identify "commonly used types" during code execution time and then emit optimized code for this. Of course they always need to have an interpreter handy in case something "seldom" happens.

If they had strong typing in the first place, this kind of nonsense would be not necessary. A simple recursive-descent Pascal compiler without optimization will outperform this "clusterfuck" (my judgement) approach. Runtime of the compiler would be negligible (in the order of seconds for 100k LOC) and could be amortized over many program runs.
And that Pascal compiler would be about 1/100th of the Lines Of Code of the JavaScript optimizing JIT. That implies 1/100th of the cyber warfare "opportunities".

But given the utterings of one very senior Google person, these "opportunities" might be part of the Grand Plan. Imperium needs to dominate information, ya know.

Frank Gerlach

Indeed

[Anonymous Coward]

Why do you need proper French food if you can get a BSE Burger, courtesy of Dame Margaret Thatcher. She removed all those "too expensive" regulations which kept the BSE infectant jumping from generation to generation. When you feed cow meat to cow, you really don't need to use so much precious energy, do you ?

Boy, given the Cyber Warfare and Crime Domain, we need more frenchiness, not less. Otherwise, people will go back to paper, courier and lockbox file. And we can go to McDonalds, preparing unhealthy food.

Frank Gerlach
Gäufelden, formerly French occupied zone

I'll bite

[Anonymous Coward]

Many of those "nice" and "free" things from America (C, Unix, JavaScript,Google search, facebook and Firefox) have turned out to be

A) destroying technologically more solid competition like Pascal, MPE or Ada

B) being an enabler of cyber war and cyber crime

C) a frontend/honeypot for NSA-GCHQ collection systems.

D) economically destroying any competitor. Objective is monopoly for one player.

Don't tell me Bell Labs or Google never had an agenda. Just read Mr Schmitt and then you know the agenda - more power to NY and Washington. Forgive me if I dislike this as I have no representation in Washington. I am merely a Vasall of third rank.

Frank Gerlach
Gäufelden

Re:I'll bite

[Anonymous Coward]

Subject: "I'll bite"

You mean as in, my post was food for trolls, so you're biting?

Oh, the irony (or would that just be experienced option) of someone from Germany complaining about American technology being used as an enabler of crimes
http://en.wikipedia.org/wiki/I...

Thank you Apple

[frankenpc510]

for killing the browser plug-in. It's not that javascript was a good option. It's the ONLY option.

Javascript is your enemy

[Anonymous Coward]

believe it

Speaking as a JS dev with no relevant degree...

[Pherdnut]

You sound very threatened. Like those Java devs (not all Java devs, but I've run into lot of these guys) that they've been producing by the metric ton for about oh, twenty years (or decades if you wish) who deeply resent having to learn anything new after graduation. As any JS dev can tell you, that's a really bad hangup to have in web technology, which I'm sorry to report to you has started to touch on just about everything.

You really don't need a compiler or static types to write decent maintainable code and believe me, I've seen enough Java and C# code bases from hell to tell you that they certainly don't protect from writing complete procedural spaghetti horseshit wrapped in meaningless classes either. If people would just be bothered to understand what the whole point of OOP was in the first place, they wouldn't sweat the lack of a compiler when doing anything non-trivial in a different style of execution environment.

Have you ever learned a lisp-like language or gotten very far with C#'s first class functions? If you haven't, give those a whirl and then come back to JS and stop throwing tantrums every time it doesn't behave like Java or C or C++ just because it looks like them.

Hottest toys you mean.

[Pherdnut]

I don't know about you but I won't be trusting anybody to maintain their down-compiler as JS continues to evolve. That would be short-sighted to say the least.

RE:Repulsive

[Pherdnut]

Speaking as a JS dev who's had to put with some pretty ugly Java, I'd be happy if the average Java developer would content themselves with simply actually understanding what OOP is supposed to do for them rather than assuming that the 18-method-calls-to-retrieve-some-static-data-for-a-basic-web-service mess they've made is actually OOP just because it's all wrapped in classes by default.

See also node-webkit

[Pherdnut]

For cross-platform desktop OS apps with node handling the file I/O and JS and web technology handling all UI concerns inside a window. Only the Python community does cross-platform as well, IMO.

Or you could learn to write maintainable in JavaSc

[Pherdnut]

I'm sorry, but the whining in this industry is getting pretty epic. Learn the !@#$ing tool. It's not Java or C#. You CAN write maintainable code without static types. JS and Python devs have been doing it for some time now. And believe you me, it would really help C# and Java devs with their primary languages if they learned how to.

JavaScript is paradigm-free

[Pherdnut]

I would actually disagree with the statement that it's paradigm free. While I respect C# as a feat of language design, one thing I don't like about its implementation of every language feature it can get its hands on is that it doesn't have a very strong opinion and there's just a crap-ton of syntax you have to learn as you work with it (although given the alternative I'd rather have first-class funcs in any language at the expense of bonus syntax). JS encourages OOP and functional used together and has strived to do a lot with a little syntax. Strictly imperative isn't really possible if you're using it for web UI as much of the DOM relies on event handlers and you can't really escape the topics of closure or OOP when you start doing non-trivial stuff from there. Functions can be classes or methods or in many cases both.

Done

[Anonymous Coward]

Today we'd have little trouble designing a better language than JavaScript for what we currently do with it, but our best bet to get something like that working is to build a language that compiles to JavaScript and then hoping browsers start building VMs for that language directly, skipping the JavaScript step.

Done

PHP

[Anonymous Coward]

PHP is a testament to that truism.

PHP: The Good Parts would be a 1 page book