/tech/ - Technology★

It will displace Java simply because it has come to fruition that Java is a horrible, unmaintainable mess. Even good Java is shit. Most Go jobs are snapped up real quick because it's fun to write (if you're an imperative-headed Go guy) and easy

>>983049

*If you're an idiot

>>983049

While java is truly verbose, it has been shaping up since java 8 and it's ecosystem, from build tools to external libraries is vast.

I don't agree that it's an unmaintainable mess and horrible as you say. That impression I believe stems from faulty and misguided overuse of design patterns.

Also while I agree on your point of golang, I don't believe it's enough.

>>983051

I did not realize image boards were popular in India

I wouldn't commit suicide to escape a go job, but it's Rob Pike's language for lesser people. Want to fix up terrible (but style conforming!) code from India? Want to focus on mastering a mediocre performance/resource/safety muddle language right as a new generation of finally-good languages is growing up? Want to have a non-interactive non-dynamic bondage and discipline language without having the most basic of static guarantees that should be your reason for tolerating such a language? Want to know that everything just slightly sucks all the time because the enlightened elite that run the language think that any kind of programmer ergonomics is beneath your station?

>>983053

You're just angry that Rajeesh has a better job than you and a cute wife. India superpower 2020.

>>983067

solar freaking shitting roadways

Golang will forever be for google shill nigger faggots period

>>983059

>new generation of finally-good languages

Please tell me you aren't talking about Rust.

>>983089

Most new languages selling point is that they aim to fix the warts of older/more used languages. What he doesn't understand is that that is not enough and that in this process they introduce their own warts.

>>983073

>walk in street

>without warning, foot sinks into a shit-filled pothole

>trip over and get trampled by the crowd of fellow pajeets

>mfw i realize that india will not be superpower by 2020

>>983093

>Hurr, just because a percentage of the poorest people shit outside, it must be true everywhere.

>t. brainlet

>>983093

You Americans have dog shit everywhere. What is one type of shit vs another.

>>983095

>because a percentage of the poorest people shit outside, it must be true everywhere.

Except it is true everywhere

https://en.wikipedia.org/wiki/Indian_states_and_union_territories_ranked_by_availability_of_toilets

53% with NO toilet of any kind (not even pit).

>>983095

>a percentage

60% of India makes less than $3.20 per day.

>>983097

>2011

>Average is a reliable metric

>>983098

>3.20 is little money in muttland, therefore it is so everywhere

Good job guys, you sure showed me. Now its time for your mandatory trans operation.

>>983101

>>Average is a reliable metric

It is literally not average you fuckwit. I guess they don't teach basic math where you are from. These are absolute numbers.

>hurr durr 2011 does not count

>>3.20 is little money in muttland, therefore it is so everywhere

Yes anon you can afford housing and food for that in India. But you have 20 room mates, and the food is 1000 calories worth of rice. In America minimum wage buys you you're own flat and thousands of calories.

>>983102

*your, fuck deleting posts :^)

>>983102

It literally is not.

Check the link again, more than half states have over 65% toilets and the most populated are above 90%.

Now if it was a weighted average by population, that would be another story. It's like saying a cannibal village in USA has the same weight as New York when deciding cannibal percentage.

>>983104

> toilets and the most populated are above 90%.

Lol no

https://en.wikipedia.org/wiki/List_of_states_and_union_territories_of_India_by_population

>#1 population Uttar Pradesh 34%

>#2 population Maharashtra 36.9%

with no toilets

>>983107

Oh even better #3 population

> Bihar 76.9 without any form of toilet

>>983108

>>983107

With little exceptions (like Bihar), as I said more are less than 35%

>>983104

>half the states are 45% street shitting

<you think this is good

HAHAHAHAHAHAHA

>>983110

I literally quoted you the top 3. The BEST CASE in the top 3 is 34% shitting in the street.

Also again:

>2011

When you should know in the last 5 years there have been major improvements.

>>983112

>34% is the same as 60%

Nice bait

>>983114

>#4 west bengal 45% street shitters

>#5 population Madhya Pradesh 71% street shitters

Even 1% street shitting is too many. Here you are defending half the faggots filling the roads with shits.

>When you should know in the last 5 years

Ah yes major improvements in getting basic things society needs like not street shitting.

>>983111

When you were banging rocks we enjoyed a thriving civilization. A few minor inconveniences are not enough of a setback to our overall superiority and history as a people.

>>983116

We wuzz kangs.

>>983115

Why is getting rid of street shitting something society need?

I don't see you sour about dog shit on the side of the road, or cow shit that grows your food.

>>983117

Yeah but in contrast to mutt black population, who your are parodying, we actually were and will be once again.

>>983119

That explains why you were so easily taken over by a much smaller number of Europeans hundreds of years ago

>>983120

>taken over

>implying it wasn't just a lucrative deal with the EIC and when it stopped being so we kicked them out

Nogrammers ruin another good thread

>>983059

>bill has superior tool

>jake doesn't but is jealous

<th-that's fine... I'm better because I use a worse tool

Your post was a lot of writing that said absolutely nothing.

>>983096

<dogshit

>>983118

Neither of those carry HUMAN INFECTING DISEASES. Enjoy death by cholera you fucking third world shitstain.

>>983148

>superior tool

>jealous

it's go.

nobody is jealous about it

calling yourself a go programmer in public is admitting to be a submissive in Rob Pike's harem of dumb-dumb programmers.

In my opinion the main selling point of Go is that it tries to be "good enough" without being harmful. No other popular language does this because everybody is chasing what the (((enterprise))) wants. I am surprised that Google gives that much freedom to Go maintainers. If I remember correctly, in 10 years there was only 1 instance where Google forced the hand of the creators to implement an 1 anti-feature into the language. Now that Go is gaining popularity every year the language will get new anti-features because of community outcry. If that trend continues sooner or later Go will become what Java is today. I am sad to see this and it is a mistake.

>>983145

In every Go thread on /tech/ there are ALWAYS 1-2 people who try to ruin the thread by shitposting hard. I am surprised nobody posted that Rob Pike quote how Go was made so newbie programmers could pick it up instantly, which (in their opinion) makes everyone who uses Go a retard. I understand that people are biased because they hate Google (for good reasons), but sucks that we cannot have 1 good honest discussion about Go.

>>983233

> in 10 years

in 8 years*

>>983233

>I am surprised nobody posted that thing everyone knows about

>which (in their opinion) makes everyone who uses Go a retard

ddg indexes /tech/ - #1 hit for 'programmer ergonomics' is this thread

I don't think you're stupid. I think Rob Pike thinks you're stupid. And not definitely because of the quote, but rather because of the design of his language and how its designers react to proposals about it: https://grimoire.ca/dev/go

Zig also tries to be "good enough" without adopting every feature in the world. It's also going the gofmt directino. But, it also does basic shit like tell you when you incompletely switch over an enumerated type.

>>983242

>Zig

Wow that looks really interesting. Thank you

>>983267

Ah D. The language where you can have both shitty garbage collection and memory segmentation faults. Worst of all worlds.

<inb4 you can turn it off

<inb4 you can write in memory safe ways

>>983268

><Inb4 you can turn it off

No, if you don't want GC write in C simple as, but some of us don't have that sort of time.

Also you're gonna get segfaults with almost any systems language so idk what you're on about.

>>983267

Is this a joke? None of the D compilers even support OpenBSD. And as for the architecture support the only fully supported are i386, amd64 and arm. Not to even mention that even though language is mature and 15+ years old, there is no community at all.

D E A D

E

A

D

just the fact that it's a compile error to have an unused variable pisses me off, but the retarded shit they do goes far, far deeper than that

here you might like this website http://golang.web.fc2.com

>tor users can't post images

too lazy to update my bypass script. https://ibb.co/eMkwUK

>>983305

They should have a debug switch that disables the variable usage check. Not a terrible idea for the version you're going to ship though.

>>983307

>im about to ship something, ooh wait let's check for unused variables for the rare chance that it indicates a bug

oh what a novel, game changing idea. thank god rob bike has brought us such ingenuity. meanwhile, every language and your dog has been using their "linting" meme-tier QA tools for 10 or 20 years

>>983305

>>983307

It is a bad practice to have unused variables in your code. And if you are you debugging things then you can use

>_ = foo

to use the variable.

>http://golang.web.fc2.com

This looks like some random meme website. What is your point?

>>983312

gr8 b8.

>This looks like some random meme website. What is your point?

if you were a gopher or have observed gophers in practice, you would understand

>>983312

"Bad practice" and "actually causes a problem in the real world" are two different things for a reason.

One particular example that sticks in my craw every time I create a new .go file is the fact that you *must* use shit that you're gonna import. Okay, I'm writing an API wrapper, so I know that I need Strings and JSON. I import those, write some boilerplate, and..

Compile error because I haven't actually done anything with JSON yet. I have to go fucking comment it out.

This is a pain in the ass and doesn't fit how actual developers write code in the real world, nor is it reflective of solving a problem actual developers face in the real world.

>>983088

Yeah fuck Rob Pike and Ken Thompson! They have no idea what they're talking about or doing! It's all about pleasuring Google!

t. anti-Golang retard nigger

Please kill yourself.

>>983333

they don't.

this thread inspired me to go through the entire go-is-not-good collection, and this one's the best: http://www.evanmiller.org/four-days-of-go.html

it starts out with nitpicking, but then at 'autistic gopher' it really focuses in on why Go is such a bad language: Rob Pike like slava is a twat.

>>983045 (OP)

1. Matrix/Synapse/Riot

2. Go-IPFS, Go-TOR and Go-I2P

3. Hydron (replacing >>>/hydrus/)

4. Meguca (replacing Lynxchan/Blazechan)

>>983501

>listing stuff in go

let me try...

influxdb

wtf I can't tell that go is a shit lang anymore

>>983501

>3. Hydron (replacing >>>/hydrus/)

Huh?

>>983501

(Meguca are fags, don't go there.)

>>983514

https://github.com/bakape/hydron

The dev is a faggot though

>>983515

I know, but still Lynxchan are worse fags. Blazechan has a shot.

>>983389

The legs are for the redhead, the shoe on the bottom right is for the ginger. Assuming the blonde is slightly crouching

>>983528

Sounds lik ur mad.

meguca dev is based.

>>983575

I am trying to make everyone here happy, on one hand Hydron is "stunning and brave", on the other hand Meguca is just shit.

>>983379

Pike said those things long before Google. Plan 9 was the epitome of "no local storage", and Go is mega-inspired by Inferno and the other one they wrote.

>>983116

When we were banging rocks, the greeks conquered some of your asses and either raped of sold you into slavery. Then the persians, followed by the hepthalites, then the huns, the afghans, more huns, turks and finally the mongols all had their turn getting sloppy seconds raping pajeets. You are literally the mystery meat of the world genetically

>>983269

>You think d is systems level

>D has a GC

>What is rust

>>983327

>He thinks the real world is the 1 % of time writing the hello world equivalent of the project

>>983656

>comment out part of code and recompile

>Go: HEY YOU'RE NOT USING THAT IMPORT NOW

>can you just fucking c-

>Go: I CANNOT. I WAS INVENTED BY A GUY WHO HAS TROUBLE COMPLAINING ABOUT THE ANNOYING HABITS OF COWORKERS TO THE FACES OF THOSE COWORKERS. ONLY THROUGH MY INFLEXIBILITY CAN ROB PIKE GET HIS REVENGE

>comments out include

>code finally compiles

>now have to uncomment two separate pieces of code to continue

you're lucky Rob Pike didn't have a beef against a coworker who left a lot of commented-out in a project, or you'd have to delete all that code, not just comment it out.

>>983658

>Someone who has never written real software

>>983659

'real world' is not a crucifix and you are not fighting vampires. Compilation is the wrong place to impose code quality standards. An example of a good place, in the real world, would be Bamboo.

>>983667

>It should be an ad hoc mess of 50k incompatible systems

Good idea op. Go back to cuckscript.

>>983658

this.

>>983510

wtf i love gophers now

>>983658

While I too don't agree with some of shit forced on you by go, I believe thats its main selling point. No deviation of style and workflow from the "one true way". It removes the need of individual style guides and standards from each project and every go project follows the same mantra.

>>983658

Go is an autistically passive aggressive language.

>>983692

no it doesn't you dickfuck. that's the entire problem with textual languages. why because gofmt? yes if the entire userbase are people who got memed on and use Go because of blogposts on medium.com and Google, then most of them would indeed use the same formatting tool, which could be done in any language, and will only be a thing in Go until it inevitable stops being a thing. style has never even been an issue. it's for faggots who sit in the blogosphere all day and ended up talking about trivial issues and spending 3 years writing blog posts arguing to eachother over which was is best. if you're reading the style guide before editing some code you should just kill yourself.

>>983575

The meguca dev is a transsexual faggot who bans you if you're not a shitlib.

>>983333

they killed it with Golang 2 though you have to admit

>>983692

Fine. Why is that a benefit, again? Python has pep8 if you actually care about autistically enforcing code styles, and Go is not a better language for forcing this shit down everyone's throat.

>>983045 (OP)

Golang will never replace Java in the enterprise.

>>984311

>Google invented Go

Yup. Rob Pike and Ken Thompson had nothing to do with it and Google did not bow to their expertise to create it.

>>984322

>No company ever creates anything because employees did the work

Ok faggot good luck with your communism.

>>984344

>Properly crediting a work to it's creators' vision rather than saying the Google hivemind created the totality of the Go language is communism

Fantastic logic.

>>984370

>Crediting the product to the company instead of the worker

Literally what communists complain about almost word for word

The only software I've used that involves golang is Hugo (gohugo.io), a website generator. The software easy to understand, but that's just one use case. Not enough for me to form a solid opinion of the coding language.

>>984416

I'm a big fan of the fact that some the trend for this latest generation of languages is both static typing AND no exceptions. Go, and Rust being examples of this.

>>984412

>company > worker

>collective > individual

hmmmmm

>>984420

>companies are communism

wow big brain idea there anon

>>984428

Right wingers are mentally ill. They're incapable of even accurately describing what they're supposedly against.

Leftists understand both socialism and capitalism better than right wing faggots because we consume sources of information outside of artifacted jpegs.

>>984446

> and capitalism

That's because they made up the the term to strawman real economics.

>because we consume sources of information outside of artifacted jpegs.

Yeah like shitty facebook posts in the secret club groups.

>>984446

not books tho

http://esr.ibiblio.org/

wtf I like go now

I had no idea that this awful language could be a Python killer

that changes everything

I program in go for a living right now.

It is amazing.

It is not better than C.

I would say that it is a good replacement for C++/Java/C#.

There are several weaknesses to the language if you try and make it like C.

>>984419

How is having no exceptions a good thing? The whole point of those is that you can't just ignore errors and carry on.

Python, with all of its warts, got exceptions right in the sense of promoting their use not only for "Exceptional, unexpected errors" but for any case when you need to communicate to the calling layer that something happened and that it MUST be handled in a special way.

Without them you're regressed to optionally checking return codes or "if err != nil" in Go.

>>984517

>The whole point of those is that you can't just ignore errors and carry on.

Side channel exceptions are absolutely horrible. Every random function can fuck up in a way that is impossible to tell from looking at the type.

>Without them you're regressed to optionally checking return codes or "if err != nil" in Go.

Much better but still shitty.

>>984517

Exceptions are terrible. It is stupid to just break the whole fucking program because of a small non-important (ever had an empty catch clause? Yes, that's because the error was not so important) error you didn't even know could be thrown there, not to mention handling them try-catch style is even worse, considering you are expected to put try catches inside catches just in case whatever you are trying to fix in the catch clause doesn't work as expected (you can get million-dollar-bugged pretty easily if attempting manual deinitialization of resources in Java, for example).

Letting the programmer check whether the function returned another error parameter via some other channel is also a mistake, since programmers may inadvertently not know an error can be returned from a specific function, depending on the implementation, or semi-conciously decide not to handle it, if they are shit programmers, but at least you don't endlessly nest error handling inside error handling.

As much as I dislike Rust, using tagged unions as return types was one of the few sane choices they did. To either return a value or an error and forcing the programmer to handle it or consciously ignore it is the way to go.

>>984536

>Letting the programmer check whether the function returned another error parameter via some other channel is also a mistake, since programmers may inadvertently not know an error can be returned from a specific function, depending on the implementation, or semi-conciously decide not to handle it, if they are shit programmers, but at least you don't endlessly nest error handling inside error handling.

> Letting the programmer check whether the function returned another error parameter via some other channel is also a mistake, since programmers may inadvertently not know an error can be returned from a specific function, depending on the implementation, or semi-conciously decide not to handle it, if they are shit programmers, but at least you don't endlessly nest error handling inside error handling.

RTFM or if you're working on a large code base use a development environment that allows you to easily go to the definition of a function you want to use. This is not rocket science.

This is why the pajeets can't use anything other than Java.

>>984550

>This is why the pajeets can't use anything other than Java.

Ah yes the language with all those exceptions

>you to easily go to the definition of a function you want to use. This is not rocket science.

Yeah it's more complicated.

>RTFM

Typical no-dev assumes most things have any documentation at all.

>easily go to the definition of a function you want to use

In anything non trivial there is going to be so much control flow for every little thing it is impossible to follow (without wasting a massive amount of time). With non shit languages you can read the literally 1 line of text that is the type and know everything it takes, and everything it can return.

>>984523

The operative word was "optionally".

No, even being able to implicitly ignore errors and have your program continue as if nothing happened is not acceptable. Go really fucked-up there, and also by trying to do cutesy shit with nulls, instead of using Optional types like any modern language.

But yes, I completely agree with the issue of not easily knowing how a function can fail, as that would require studying all the function call chain until you reach the language's "primitives", due to exception bubbling. At the very best you can embrace it and handle what you can (often to fallback to other way of finishing the given task) and just let everything else be rethrown to the client.

>>984536

>To either return a value or an error and forcing the programmer to handle it or consciously ignore it is the way to go.

Sorry for the ignorance, I haven't programmed in Rust that much, but that reads just like exceptions, but with the difference that error return values don't "bubble-up" implicitly. Basically, if some error that you can't handle happens when calling another function, you would then return your own error to the client. The client then would check if it can handle your error, and if it cannot, then it would return its own error to its own client, and so on.

And yes, I've seen try-catches inside catches inside catches, often for fallback measures (e.g. send mail using mailchimp fails? try using the local SMTP server. that fails? use someone's Gmail account, yeah, seriously). But wouldn't that work in a *very* similar way in Rust with nested match expressions that check for Result::err anyways? Or in Go with nested "if err != nil"?

It just seems to me that error handling is ugly everywhere.

>>984554

>instead of using Optional types like any modern language.

They should just use the Either Monad :^).

>, I haven't programmed in Rust that much, but that reads just like exceptions

They are value level in the type. They just have convenience macros to bubble them without nesting the checks.

>>984500

Eric SS Raymond is ready to unfuck Go? WOW

>>983088

Gee, hope you've deleted all those WebMs on your drive.

>>984311

Association fallacy. Google eats their own dogfood, who cares. Go doesn't have the ecosystem, libraries, or thick-client application support of C# or Java and will not be displacing those technologies any time soon.

>>984628

>Association fallacy.

There is no association fallacy. Google uses go you faggot. If they wrote go and did not use it that would be different.

>>984550

>RTFM

I bet you shit on JS type cohercion, even though that aspect of the language is widely and well documented (and may I say sometimes even logical).

Not every programmer is perfect, unlike you. Forcing shit or tired good programmers to rely on their own judgement to avoid catastrophically fucking up just because you can feel hardcore is a bad idea. This mentality is the reason we still have buffer overflows in programs that shouldn't have buffer overflows.

>>984554

Result types in Rust are templated enums. An enum in Rust is basically a tagged union, so when you return a Result type, you are actually returning something that may be an Error, or a the type the actual function should return. So, say, a Result<int> would return a Result with an int if the function was correctly executed, or a Result with an Error (basically, something like an Exception object, but without actually throwing it) if something bad happened.

After calling the function, it always returns, and the programmer is either forced to correctly handle both cases (by using a match expression, which is basically a switch on steroids) or unwrap the result via ? operator (it's been a long time since I have used Rust so exact rules escape me, but IIRC, it means "give me the result if there is any, otherwise return and bubble up the Error if this function can return an Error, for example if it is yet another Result type function, otherwise panic and close the program").

In my opinion, this is the most elegant and logical way to return errors. It is also non-intrusive and transparent enough to let the programmer know when they are probably doing something they could regret.

>>984950

one guy got banned. wow hardcore.

>>984958

People get banned over there all of the time for attempting to post porn and loli in the meadhall threads. I'm not shilling for it but to call it a den of shitlibs isn't correct. People like this anon:

>>984446

>Leftists understand both socialism and capitalism better than right wing faggots because we consume sources of information outside of artifacted jpegs.

Regularly show up and get BTFO. Then they get banned and want to complain because they got BTFO by facts. It states right in the meta thread that the place is a Nazi hugbox but they're more than willing to engage in debate and talk politics if you don't act like an idiot. The anons over there seem well read too because it's one of the few places where I see books constantly getting dumped and the contents of them discussed. It's certainly better than nu/pol/ or /leftypol/.

>>983705

Bullshit. I never moderate /pol/ and only ban on /a/ for shit like frogs and idols.

>>984536

>ever had an empty catch clause?

No actually. An empty catch clause means you don't care that the try clause didn't finish, at which point you might as well not have run it in the first place. In general though, exceptions are useful for control flow; complaining about an empty catch block is like complaining about an if statement not matched by an else.

>>984950

>>984966

Fuck you megucashill you actually have me interested in checking it out.

>The entire reposurgeon suite (not just repocutter and repomapper, which have already been ported) is changing implementation languages from Python to Go.

> ...

>But the real reason that reposurgeon is moving to Go -- rather than some other language I might reasonably think I could extract high performance from – is not either of these demonstrations. Go did not get this design win by being right about Unicode or build protocols.

>Go got this win because (a) comparative benchmarks on non-I/O-limited code predict a speedup of around 40x, which is good enough and competitive with Rust or C++, and (b) the semantic gap between Python and Go seemed surprisingly narrow, reducing the expected translation time lower than I could reasonably expect from any other language on my radar.

http://esr.ibiblio.org/?p=8161#more-8161