*saves programming*

>*saves programming*
how long is this piece of shit going to hog his language? i'm sick of c++

Attached: 2784075-jon+blow+2015.1+cropped[1].jpg (538x640, 68K)

Other urls found in this thread:

youtube.com/watch?v=xSXofLK5hFQ
youtube.com/watch?v=EVhCUSgNbzo
caseymuratori.com/about
dlang.org/spec/garbage.html
youtube.com/watch?v=8dinUbg2h70
c-faq.com/decl/spiral.anderson.html
twitter.com/NSFWRedditGif

How long did it take him to release a game after Braid?

8 years

likely he will die of a heart attack before releasing it, then

I've started copying Jon's dress style and my programming skills have improved somewhat.

Attached: programming collars.jpg (391x143, 9K)

what exactly is his style? often seems to me he writes C with better polymorphism

>2018
>developing a programming language in cathedral style

Dead before it even began. You need real users and people trying it out in production scenarios to launch a programming language, not hype. But this guy is a video game developer so of course he has the equation backwards.

>referring to cathedral style
Muh nigguh

Attached: cathedral-bazaar.jpg (199x308, 25K)

>blocks your path

Attached: man-month.jpg (1670x2610, 681K)

>You need real users and people trying it out in production scenarios to launch a programming language, not hype
but that's what he's doing? he and his team are making games, engines and development tools with it right now and he's going to open it up more after the summer

Notice how I said it needs real users, not more developers

Only 1 game(and thus an engine) and 1 devtool. It's not DOA but the longer he waits imo the longer he's going to stress himself out with the obvious notion that there's going to be 10x more work post-'release' than right now.

How is it better than rust or D?

Attached: 1487105356281.png (657x525, 24K)

SE books have amazing titles and covers

>rust
easier to work with; if you don't want to write unsafe code then you should become a better programmer
much faster compile times
language features for low-level optimization like structs of arrays and inline assembly
>D
no garbage collection

D is made by people who adore C++ but just did not like all the legacy features
Rust is made by student JS lovers who realized everything with llvm bindings is fast
Jai is made by one skilled guy who dislikes C++ but is forced to use it because its the best viable tool for the job

quick, give me a rundown of this language that he's developing

Switch between Structure Of Arrays and Array of Structures to your memory layout requirements.

Jonathan Blow sort of deserves a MacArthur genious grant desu.

Pshaw, Jon Blow's group of senior engineers are not TRUE users like me, after all they have only been professionally active for most of their lives. He really should be seeking out advice from me and my friends for his own project.

What if you had the features from higher level languages, but instead of at runtime everything was done at compile time and the syntax wasen't bad like Rust's.
What if the language came with its own set of standard tools like Go so that you have a standard way to interact with third party projects without needing to learn their build system, dependency management system, cross compiler setup, etc.

What if the guy behind it was practical and doesn't care to appeal to everyone unless by coincidence.

Look at this testimonial
youtube.com/watch?v=xSXofLK5hFQ

Attached: ironic community.png (633x436, 166K)

what the fuck did I just watch

This one is better
youtube.com/watch?v=EVhCUSgNbzo

there was a programming language I once stumbled up on which compiles to c++ and there were a top down futuristic racing game as an example

I think the developer was from russia

does anyone know what I am talking about?
I can't find that language anymore

>Pshaw, Jon Blow's group of senior engineers are not TRUE users like me, after all they have only been professionally active for most of their lives. He really should be seeking out advice from me and my friends for his own project.

If he wants people to use his language then yes he should seek out advice from ACTUAL USERS. If all his "senior developers" have done is make shitty platform games (aka the most cloned and saturated game genre in history) then who the fuck cares.

>or D
there is a clear focus and purpose

I bet the majority of games you played used Casey's software.
caseymuratori.com/about
This dude is also currently making a game from the ground up with no libs, this means he has experience with everything technical you could need for a game of any type, be it a platformer or something else.

A lot of the people he talks about were ex Microsoft, they know the frustrations of C++ and how it negatively affects design. He basically has an entire group of frustrated lifelong engineers who know what the problems are.

>If he wants people to use his language then yes
I highly doubt that's a goal. Which is the best part, it's a project born out of frustration, not a need to appeal. They see problems in current languages and are implementing solutions to them.

Ultimately, who fucking cares how it's made, as long as it's good. All I ever hear from people is that they regret opening their pre-release projects because there's more chaff than useful content. It's not like he hasen't been open about nearly everything on stream, it's not like you can't email him with criticism.

The majority don't know shit, design by committee is why we're in this mess in the first place.

nothing truly great has ever been achieved through design by committee

this guy has worked in games since forever and has written countless games and engines of different styles. i'll take someone who has already proven that he can get shit done over some armchair LARPer any day of the week

>Bink
It's that crap that never works when playing an older game.

>Widows/WINE failing to maintain compat is his fault
nah
That shit was used in nearly everything, same with granny3d.

The point is that they all have experience and are no slouches, they're certainly not "shitty platform game" developers, they're experienced engineers and game developers, they have chops in both.

>if you don't want to write unsafe code then you should become a better programmer
I'm currently reading about JAI and it says it's strongly typed. Doesn't that mean it is in some ways safer than unsafe rust that will allow for weird type conversions? Although I don't think this is something a game programmer needs.

>no garbage collection
I think the D people are working on this.

dlang.org/spec/garbage.html
>There is currently work in progress to make the runtime library free of GC heap allocations, to allow its use in scenarios where the use of GC infrastructure is not possible.

Although I'm not sure if this includes the standard library or just the runtime library.

t. new programmer so sorry if I'm not understanding things clearly

Attached: 1496115462091.png (612x491, 98K)

>starts stream at 8 AM EST
>doesn't actually touch any code until 11:30 AM
>brews tea 15 different times until it's *just right*
>goes on autistic rants about how OOP is garbage and C++ is garbage, even though his entire language and compiler is written in/based on C++
>goes back and "fixes" already working components of his language for a week
This guy is just too scared of committing. That's why he's so slow to release.

it makes sense when he explains it as the public release of his compiler being compiler version 4, with the previous three already being superseded

oh also i forgot
>made literally one game that wasn't a flop
>people are trying to elevate him to be a new aged john carmack

braid wans't even that good, fight me

Attached: bro_nan.jpg (900x900, 109K)

It's not about strong vs weak typing, problem with Rust is that you're constantly fighting the compiler because security and borrowing and whatnot is a language design construct that takes a lot of time to get used to. Jon wants his language to have as little friction as possible so that you can focus on actually writing code.

As an example, he changed the syntax for enum members to mimic the syntax for constants so that you can easily cut and paste code around when refactoring your program.

explain in what sense the witness, a niche puzzle game that broke even in a few weeks, was a flop

Never played any of those games, they all look like garbage
>ex Microsoft
They will hire anybody who has read a C# book. Not exactly a high bar

>comparing having users and listening to actual customer feedback to "design by committee"
You people are fucking morons.

take it easy Mr. Blow it's just banter

go finish your memelang

>solving the same-concept puzzles for several hours
>there is nothing about why the people turned into stone at all
have to admit the world is visually beautiful

C# is probably older than you are, these people were in before it even existed.

> Not growing an unkempt beard and not cutting your hair following RMS
My beard is the source of my power

Beard is additional weight which requires energy to lift, energy that could go to your brain instead

I think people are more impressed with his technical work and ideals rather than his game designs.

Yes he's making video games and selling those, but the games themselves are probably the least impressive.

Braid is pretty simple in concept, but the technical achievement of making it perform on the Xbox, is dope.
youtube.com/watch?v=8dinUbg2h70

He very much is like Carmack, a good engineer. I don't remember anything about Carmack that was GAME design related, it was all technical marvels like "here's how we made it portable, performant, possible", etc.

Jon is more valuable when viewed as something other than a game designer.

these dubs don't lie

my favorite carmack+romero game design was straight-into-action intro/menus

Before that it was they will hire anyone who read a visual basic book. Plebs gonna pleb.

>this level of denial
Yeah it's highly unlikely anyone there had any merit. I'm willing to accept your alternative now, which team and language has user certified R.E.A.L™ users backing it?

Yes, you're correct, proprietary monopolist developers don't have any merit.
>which team and language has user certified R.E.A.L™ users backing it?
TempleOS and HolyC

Hilarious, truly.

We already have D, though...

No, but seriously, we are reaching the age when any idiot with a stupid pet language can write an LLVM frontend and act like they invented type systems and lambda functions. Users don't care. Your language needs to do something besides remove features from C++.

Why should anyone care about "users" in the first place? Users who like it will flock to it, this isn't a popularity contest where you need to implement x,y, and z just because John Hoogivzafuk likes them, if the language is good people will use it, plain and simple. You have a gorillion languages that pop up and try to please everyone, and nothing comes of them, how about making 1 with a clear purpose and just releasing it when it's ready.

Not everything has to be a communistic, bloated, pile of shit.

I don't know, why don't you ask the guy in OP as he is the one hyping up his bullshit language in order to bait "users" instead of just keeping it to himself

At least he's not Phil Fish

Attached: image1.jpg (1526x849, 225K)

Just removing features from c++ gets you pretty far imo.

What does the OP have to do with any of this? How is some joke post "hype" by any means? Is this your first day here?
Does everything have to be some argument, can we not just joke about technology, can we be a little comfortable for once.

Like yeah Jai is interesting, but I don't think anyone here is seriously championing it since it's not even released yet. Nobody is trying to hype up shit for users, they're more than likely taking the piss because of Jons personality while also discussing the merits of his works in tandem.

>initialised structs
>sane pointer syntax
>sane printf
>no . or -> confusion
>array of structures vs structure of arrays with one change
>1 second compile times
It's everything I ever wanted.

Attached: 1527407764258.jpg (681x681, 236K)

Maybe it's Stockholm's, but I like printf. What's wrong with how printf works in c?

With Jon's language you can just chuck a data structure into a printf statement and it will pretty print it for you with a single format specifier.

It means you don't have to fiddle typing something like printf("(%d, %d)\n", screen.x, screen.y);
anymore, you just say printf("%a"); and it does the same thing.

Attached: 1524965101334.png (1280x720, 396K)

having to manually specify type in statically typed languages is redundant since compiler already has that informantion
plus the format placeholders are fucking crazy if you want to do it correctly in portable way, it was made for 16 bit and all the grammar was added afterwards so the ones with mnemonics were already taken
nobody wants to use char/short/int/long and fixed-size integers are solved with macros because they are typedefed into different integer types across architectures
automated print generation of enums (possible with ugly preprocessor magic) or structs would be comfy but whatever

Attached: Screenshot from 2018-06-08 15-18-39.png (587x643, 83K)

Who would bother with that garbage?
%lld will slways work for 64 bit, %ld for 32 bit since minimun guarantees is all you need when holding values.

Is the stdlib GC and the GC off by default, now? If not, did they substancially improve the GC so I can use it as some kind of better Java?
No? Thought so.
Also, metaprogramming is about as powerless as in C++.

Remedy uses D as a scripting language, but I'm assuming they don't use the stdlib.

Yeah, but gamedevs have a history of abusing things that don't run away as scripting languages.

>they're ready to wait years for overhyped vaporware
You could have mastered Rust several times over already.

Rust is both great and shit my dude.

>but I'm assuming they don't use the stdlib.
Also
>inb4 betterC
If all I wanted was a better C I would have used a better C, like zig.

This, desu.
You could have mastered Ada several times over already.
But wait, it's the same here.

Go has a %v formatter, that determines how to print the object in a standard way, usually by calling object.String(), if implemented, otherwise, typically a pointer or int, or whatever it actually is, in Go syntax.

Am I understanding this correctly, is %a the same?
Because I like that feature a lot. It almost makes the transition from debugging statements to just useful verbose statements, trivial.

%a is for floats in hexadecimal in C so I guess Go devs wanted to stay compatible

Some people have done this before, not with Rust, but with any language that you learn for a while and then never use again. Some people want or need something long term.

I could tell Rust was going to be supplanted soon after I had heard about it. The concept isn't new, it's just that nobody has implemented it yet. Mozilla is the first, but they're certainly not going to be the last.
You might think of it like this,
C is to Rust what C++ is to ??

I've heard nothing but complaints, coupled with Mozilla's bad reputation, leads me to beleive there is plenty of room for improvement somewhere. In my personal opinion, the syntax is not great in Rust, maybe not bad, but if something can do this more gracefully, then I'm all about it.

I don't think it's controversial to say anymore that less lines and less complexity, leads to better programs, especially when collaborating. Some people use "simple" as if it's a bad thing and not a natural progression. If my language is simple, my required effort is low, and if there's "zero cost abstractions" then nobody loses.

It's easier to write, it's easier to read, it's easier to parse, it's literally smaller in bytesize, etc.
And unlike some languages, doesn't hurt you (or your users) at runtime, all the cost is at compile time and the cost is almost nothing with modern compiler techniques coupled with our hardware.
Exciting stuff.

If Rust is anything like C, it will be around forever, literally the rest of history, without being replaced.

But, the thing C has is dependency, control, and familiarity.
If you're going to Rust anyway, you're likely going to want to drop C dependencies if you want their safety, lose control of memory management, and adapt to a new style. It feels like it's already a departure. If someone can come out with something that has the same feature set, but at a simpler cost, it could quickly become the standard, somewhat like C++ but not exactly since C++ is just a superset, with Rust to its successor, the successor wouldn't have any legacy problems, it's still a new approach. It's all advancement in compiled languages, which haven't seen that much love outside of C++ but with C++ you pay a complexity cost in numerous ways. It doesn't have to be that way, we can have useful, simple, and efficient languages/toolsets, not just a subset of that.

>drop C dependencies
Don't be stupid, C is the only way to write portable low-level code, Rust can't replace it because it's a higher-level language

Let's be honest, he's not making any game. He's just milking the audience.

Jai is going to be great.
The compile time program execution is a really great idea.
No more premake or make nonsense would be a godsend.
Apart from that a slightly more advanced and sanitized C is basically what everyone is looking for.
Coupled with the insane compile times, man, this shit is going to be great.

>i'm sick of c++
Then use nim you nimrod.

Can it compile to LLVM IR directly, yet?

Has he embraced C syntax yet, or is he still using his own shitty pointer notation?

Asking unironically, is nim any good?

>implying C syntax is any good
c-faq.com/decl/spiral.anderson.html

>not thriving even under a handicap

He's still doing things his own way because any other way is inferior.
Welcome to Blow's world.

Hey Jon