Is this the future language of video game development? Why or why not?

Is this the future language of video game development? Why or why not?

Attached: rust-logo-640x215.png (640x215, 7K)

Other urls found in this thread:

m.youtube.com/channel/UCCuoqzrsHlwv1YyPKLuMDUQ
youtu.be/qr9GTTST_Dk
twitter.com/AnonBabble

Video games don’t care much about safety. So no.

What applications is safety necessary in? Wouldn't you want safety so the game doesn't crash?

I only know a little JavaScript so forgive me for my newbie question

Creating games which don't crash is not a hard problem and if they do crash it's also not a problem (if you know what I mean). Safety matters more for applications where you care about security servers, operating systems etc.

The real future of video game development is Jai (atleast for indie developers who go from scratch)
I have no doubt.

Attached: jonathan-blow_2332371.jpg (1439x810, 76K)

jai need compile for templeos. a perfect match

Hell no, the safety shit is even more annoying when you are doing something that doesn't even care about safety. Jai is the future of video games.

>Wouldn't you want safety so the game doesn't crash?
not when it comes at the expense of having to use shit like the borrow checker in Rust
If a game crashes the world doesn't end, people don't die

Yes. The game dev world and Rust are both controlled by transsexuals.

Game stability is Paramount to UX

stability is not valued to the point where people are going to use shit like Rust to make things more stable
honestly most of the crashes in a game if it's not programmed badly come from shit outside the games control like video drivers anyway

you would think that but in reality the most important part that the game isn't shit as in it's a good game. which is much harder to make than a game which doesn't crash.

Have you used much Rust? Once you understand what's actually happening, it's a delight to write in. Nobody is really suggesting rewriting games in Rust. You can keep all your C/C++ tools and libraries, and simply use rust going forward.

I disagree with you. A fun game that runs badly will lose to a fun game that runs well. Pubg is a great example of a fun game totally dogged by terrible performance, frequent crashes, and all sorts of memory leaks. The first game that appeared to seriously challenge it for genre control completely ate its lunch

>Once you understand what's actually happening, it's a delight to write in
yeah I don't really believe you, I'm of the opinion that if it smells like shit and looks like shit then it probably is shit, there's nothing "delightful" to me about having to follow a bunch of stupid rules to manage memory when I can do it myself
Pubg being written in Rust wouldn't make it any faster

Wait for jblow to put out jai. The mans autism will provide incredible performance

jai doesn't do much C++ doesn't
looks like a pointless language

There's no profit in rewriting EVERYTHING to rust.

Attached: the literal devil.jpg (327x327, 41K)

Smells like vaporware. Think I'll keep on writing games in Rust, while you sit around hoping for Jonathan Blow to get his shit together

Attached: Screen Shot 2019-04-10 at 18.32.05.png (2866x1496, 223K)

It compiles 10X faster and jblow is expecting even better optimizations past that. Aesthetically incredible as well. I am a dickrider though.

>the expense of the borrow checker
user, you do realize that it's checked at compile time

Jon is ACTIVELY developing the language and building a game as a proof of concept with it. Check out his autism channel m.youtube.com/channel/UCCuoqzrsHlwv1YyPKLuMDUQ

Expense as in expense on the programmer

That's a significant benefit but you have to be fucked in the head if you think C++ syntax is asethetic

well obviously if you're going to make a game the performance should be good for the game to be good. all I'm saying is that when you're making a game, making a good game is the greater concern than a crash free non memory leaking pristine executable (which rust doesn't guarantee anyway)

why hasn't he released it already? why is he building all his own shit instead of using LLVM?

Enjoy your bucket, I guess. And I'm shocked that you don't think that plugging memory leaks and being more resource efficient would make pubg run faster.

And I am ACTIVELY developing games in Rust, because Jai isn't available to the public in any form for some bullshit reason

#include
std::string(foo); // yes, this is valid


a e s t h e t i c s

I'm not waiting merely anticipating with excitement and use C++ for the time being like any sane person.

memory leaks don't slow things down unless the memory builds above a critical mass or you're actually processing the things that are leaking in which case it's not a memory leak and it's not something Rust would fix.
I don't know what enjoy your bucket is supposed to mean

they have an llvm backend for release builds.
he does his own "shit" because it's faster an is very pedantic when it comes to compile time.

If he's using LLVM then why is it taking him so long? It makes language development easy, I'm going to make a language for my next game project when I've finished my current one

>they have an llvm backend for release builds.
Well, at least he's slightly less retarded than the Go devs

>I do not enjoy x. I don't understand how people can like x when I don't like x.
A-user...

yeah someone stated an opinion, what are you going to do about it?

>Rust
>Jai
>Future of video games
Well, I'm no game developer, but video games were the place where using OOP made a lot of sense, and those languages do not support it.

>jai doesn't do much C++ doesn't
Wrong, it allows arrays of structure to be effortless turned into structures of arrays greatly increasing performance. It's features also don't revolve around the performance killing objected oriented paradigm that makes memory access significantly slower. It is shaping up to be much better for games.

>If he's using LLVM then why is it taking him so long?
he's developing a game first the language second and the game developement is still actively shaping the language he isn't on a schedule.

first of all turning things into a SoA is really niche as far as performance increases go and second of all he removed it from his language anyway
and it's not like you can't make structures of arrays in any other language

Sounds like exactly the type of developer I'd like behind a language I'm using
"yeah I'll fix it next year i'm busy doing something else"

>OOP makes sense
>For software where accessing your data fast is extremely important
Game devs hate OOP, having all your data behind a million pointers kills performance.
>He removed it anyway
Citation needed. Other languages make it a pain to do it.

Attached: serveimage(38).jpg (645x729, 57K)

Game devs love OOP
sick of you retards with absolutely zero experience in the field acting like you know what you're talking about
I'm not gonna comb through his youtube vids to find a citation for you, you can if you want

>devs hate convenience

...

ITT: niggas arguing about when language to start with instead of just writing something and seeing if it pans out

Maybe pajeet devs that like releasing slow games that require a super computer to run. I'm sure the guys who made pubg love OOP.

he's independent and it's his project so yes he can do whatever the fuck he wants. but he said in the past that they'll open source it when it's finished. since he's shall we say very opinionated when it comes to open source development.

Rust's structs/methods/traits allows for OOP style programming.

99% of all games made since around 2000, good or bad, use OOP

using C++ != using OOP

You serious? OOP is the bane of game dev. All those fucking pointers and shit you need absolutely demolish performance.

they use C++ and they use OOP
again why are you talking shit about something you know nothing about

More like "I want this to actually work and be useful to other people. I'm going to get it done at my own pace"

He doesn't want it to be useful for other people, that's the point, if he did he could have got something workable out ages ago

Samefag.

Oh, glad you could join us JB, I appreciate your insight. How's the game coming along?

literally every language that isnt C or C++ is made in C or C++.

Based C.

Rust is just yet-another-front-end-for-LLVM. which is made in C

there are two types of people the ones who try to innovate. and the ones who lack the ability to do so and try to keep others down.
which one are you user?

The syntax is beautiful but its not meant to be.
> making every value immutable unless explicity stated, rather than the other way around (or better, do it like scala [var, val]).
> forcing safety unless you explicitly disable it, rather than the other way around.

He's streaming right know just ask him there. he doesn't browse this zoomer imcel Jow Forumsentoomen board.

>Once you understand what's actually happening, it's a delight to write in
>Pointers isn't allowed, so everyone are using indices in arrays to do the same
Yes very delightful

>or better, do it like scala [var, val]
That does look better, but in Rust you must also declare if you're passing a variable as mutable or not to a function.

myFunc(&mut myVar)
This makes sense
myFunc(&var myVar)
This doesn't. Are you declaring the variable again or what?

>it's checked at compile time
But it enforce program limitations that needed to be solved at runtime cost

Does God talk to him yet?

keep it the former way with passing variables, but do var and val for declaring variables.

he's very sane not sure if he's religous

I have no reason to talk to vaporware developers.
The only meaningful thing for him to do is publish his git repo, until then his language is nothing but hype.

God only has 1 voice, Terry told us this.

I remember once he said he didn't believe time was real

what makes you believe it is?

>Wouldn't you want safety so the game doesn't crash?


This is way less important than the loss of expressive power you could gain from something else that isn't based around borrow semantics. Games would more greatly benefit from languages like D and Jai, which have far more powerful meta constructs.

no, compile times will significantly bottleneck development

I program mostly in C++ and don't really know Rust but I thought this talk was pretty cool:

youtu.be/qr9GTTST_Dk

>Is this the future language of video game development?
No.
>Why or why not?
Video game development happens in engines with scripting languages.

Attached: the cat but alien.jpg (420x416, 19K)

>Not only do I want to have to spend all my time in low level hell, I want to fight a retardedly designed borrow checker telling me I can't do anything!
No thanks

>I used Unity once now I am an expert

If they wanted to write a game in whatever language they wanted, they'd be engine devs, not game devs. The overwhelming majority of successful games use canned engines. The overwhelming majority of successful games that don't regurgitate the same engine over and over again because engine dev is hard fucking work and you typically don't want to redo it for every game.

Attached: 1549898835613.jpg (300x300, 19K)

The majority of devs are not indie Unity devs, the majority of devs work at large companies were they use in-house technology where most of the programming takes place in C++ and there's no clear-cut seperation between game and engine

If you're a large company and you're not reusing work, you're not a large company for long. You bet your ass that engine is being reused for multiple projects and/or multiple generations of projects. No (intelligent) company throws away (at least) thousands of hours of R&D time for no reason, and no professional dev is going to reinvent the wheel, especially when said wheel was already invented here.

Attached: 1554141485487.jpg (1035x1199, 140K)

who said anything about throwing away code or reinventing the wheel? Do you think an 'engine' is a discrete piece of software that is made once and it's done with? A engine is a nebulous term for the lower end of the code behind a game, and it's always changing and being updated through every project to support new features and to be more efficient and expressive

There has to be at least some clear separation between engine and game, or else subsequent generational uses of the engine would require cleaving out the game parts from the engine parts in order to reuse it. Calling it nebulous with no clear-cut separation is either kicking the can down the road or duplicating effort, both of which behaviors are liable to get you fired. Having a separate engine that you improve upon on a game-by-game is not that different from being able to grab the UE4 repo and changing that. It is effectively canned, no matter if the can is being provided to you by your company or by Tencent. Also it depends on what you mean by "large company", since console developers having their own internal engine that they share between different game teams and occasionally third parties isn't unhead of historically.

Attached: 1549770738922.png (640x603, 369K)

>>he can only name unity game engine
Game end yourself.

There's a seperation, it's just not as clear cut as you would be led to believe by using indie engines like Unity and (now) Unreal Engine. A large amount of code that goes into a game exists in a state where it's not general purpose enough to be in the engine as as standard but it's not high level enough that it's something gameplay scripters write in Lua. Commerical engines aren't walled gardens like Unity where you just write scripts ontop, they're C++ libraries which you write your own C++ code ontop of, and if that code is general purpose enough eventually it might be folded into the core set of tools and become part of the engine

No it's Java. Now shut up niggers.

Basically you have the engine and the game content. If you license an engine, chances are great your game will be bulletproof from the first day.

If not, well you design and build a safe engine. Just google "gmod source developers hate". Like Half Life 2 came out, all the other games came out on the engine, but thanks to Garry Newman and Gmod, the Valve dev team had to keep the engine patching for several years.

(because when you gave complete freedom to users... Its apparently not too hard to break it.)

How is it possible to make a game in Rust when it can't express mutable state without monads?

Attached: 1554488588020.jpg (700x400, 150K)

If the language hinders productivity, it will not be adopted. The people with money care about time to market

It can though? Are you thinking of another language, user?

Does it matter?

Rust doesn't exactly prevent crashes. In fact, part of it's (memory, type, and primitive data race) safety mechanisms are just "crash". Ie, Rust by default does bounds checks on indexing an array, and in the out-of-bounds case panics (essentially an exception). Panics can be caught at the thread or process boundary, and should always represent an otherwise unhandleable error, like indexing an out-of-bounds offset. If they're unhandled, that's just a crash.

You can of course instead choose to use the unsafe version of the indexing operation, or the Option / Result returning version instead, to assert your own safety checks or explicitly anticipate / handle the error case. Ie, a lookup request handler would probably use the Result/Option version to pass back the result to the requester, any case where a bad index means a bug in your code you'd use the panic-ing version, and for optimizing a particularly hot piece of code you might use the unchecked version, wrapping it in an unsafe block and including a comment justifying why it's necessary and how you've validated the pre/post conditions.

I'm a big fan of Rust, but you should be careful to understand what exactly Rust tries to guarantee and how.