D1 is deprecated

1 day ago
D1 to be discontinued on December 31, 2012
Andrei Alexandrescu
Reply
Hello everyone, This is the end of the year, a good time to look both back and forward. Walter and I have had a long discussion about our strategy going forth. But before that, let's take a quick look at this year. By all accounts, 2011 has been a terrific year for D. There has been unprecedented growth in the community; the disreputation of a fragmented, balkanized community is finally becoming a matter of the past; the community flame wars that were the norm in the past have given way to constructive dialog; there's more interest and more talk about D in public and private events; TDPL has been selling steadily; D's brand and major position on the programming languages landscape have become recognizable to many programmers; and most importantly, the community contribution to the compiler and standard library design and implementation has blown off the most optimistic expectations. Going forward, we want to focus on D's core strengths: expressiveness, modeling power, and efficiency. We believe D is a very compelling programming language of this era, and we want to substantiate that belief with equally compelling libraries and applications. In order to increase focus and unity in the language, we are discontinuing support for D1 on December 31, 2012. That's more than one year away, which gives enough time to D1 users to migrate libraries and applications to D2. Phasing D1 away will not only clarify our vision, but also free up considerable time to concentrate on D's two largest issues: (1) quality of compiler implementation and (2) breadth of the standard library. These two matters prevent users from fully tapping into all of D's core assets. They affect expressiveness because code that's supposed to work doesn't or necessitates ugly workarounds; they affect modeling power because bugs prevent full creative uses of the language, and lacuna in the standard library limit the "bricks" to use when building; and they affect efficiency because, evidently, a quality compiler and a good standard library are essential ingredients in writing efficient code. Best wishes for the next year and hopefully many years to come to an awesome community. Let's continue working together to reach D's ambitious potential. Thanks, Andrei

Nice trolling, i expected this since the start when maintenance of D1 libraries started to slow down and being ignored while D2 was getting tons of new code daily. So, staying on D2 is way safer than using the speed and (relative)stability of D1.

D2 leverages core assets and world-class modeling power through expressiveness to provide clients worldwide with robust, scalable, modern turnkey implementations of flexible, personalized, cutting-edge Internet-enabled e-business application product suite e-solution architectures that accelerate response to customer and real-world market demands and reliably adapt to evolving technology needs, seamlessly and efficiently integrating and synchronizing with their existing legacy infrastructure, enhancing the e-readiness capabilities of their e-commerce production environments across the enterprise while giving them a critical competitive advantage and taking them to the next level.

what is D?

>>4
It's broken version of C++

>>4
D is the next Blub if C++ doesn't improve suddenly.

>>4
C/C++ fork. Like Java, but less popular.

>>5
C++, already being broken, is incapable of breaking again. It's a binary state. So really, did you mean, ``It's an unfixed version of C++''?

>>6
Without Google or M$ support, D wont get anywhere. And Google already had some Go, while M$ has tons of crud, like C# and F#, then there is Visual Basic.

>>6
Given your C code, we already know you can't understand C++, nevermind C++11. Although in this particular case, not giving a fuck about C++ or C++11 is not indicative of your utter retardation.

C++check-em

the next Blub wouldn't be a compiled language. It would be...JavaScript. Node.js on the server, JavaScript on the client.
C++ and the entire stack of compiled languages will be legacy tools used only by system/compiler developers.

>>12
FUCK. THAT. SHIT.

>>12
Windows is certainly going to JavaScript-land with Metro.
Linux is integrating javascript interpreters.
Google is already supporting a huge JS infrastructure.
MaxOSX and iPad already support the full range of HTML5 since "flash is obsolete".
2012 - THE YEAR OF JAVASCRIPT ON THE DESKTOP.

>>14
flash is obsolete
That is why they open sourced it in hope to prolong agony.

D is a better C++, for real this time.

>>16
...through fucking around!

>>12
>>14

Except people will still use C/C++ for desktop applications anyways, because it's going to use less memory and will execute in less time. If Adobe wants Photoshop to not lag like a bitch, it's not going to use anything other than C++. Same thing for anything that needs performance. Hell, you can pretty much guarantee any performance demanding game is going to be written in C++ too.

>>18
naw dude naw

>>18
The performance gap is closing fast. In 2008-2009 you couldn't use JavaScript for this.
Now its about the level quarter to half-speed of (unoptimized) C++ and every month the JIT is getting faster.
Games are starting to appear which were impossible.
The new typed arrays make things like the entire Linux kernel to be run at acceptable speed.
Its not going to replace C++ at all places, the apps which depend on raw C++ speed will stay on C++ until the apps become obsolete.
The main problem isn't actually the speed, its system integration. Metro is going to be the flash point: now every code monkey with minimum JS could start writing software and compete with more experienced programmers using compiled languages.

>Metro

To integrate a Metro style app using JavaScript properly with Windows and other apps, Windows provides a set of app-level events, which WinJS wraps for convenience. The generated default.js file subscribes to the app's loaded event and then starts the app:


// default.js

(function () {
    'use strict';
    // Uncomment the following line to enable first chance exceptions.
    // Debug.enableFirstChanceException(true);

    WinJS.Application.onmainwindowactivated = function (e) {
        if (e.detail.kind === Windows.ApplicationModel.Activation.ActivationKind.launch) {
            // TODO: startup code here
        }
    }

    WinJS.Application.start();
})();

Javascript still has no valid support for shared-memory concurrency. Asynchronous timers and HTML5 workers aren't anywhere close to adequate.

Javascript has no programming support for vectorized math which uses SIMD instruction set architectures.

Javascript needs to run in a sand-box called your browser or a sandboxed javascript shell, with no access to your file system except through network protocols.

Javascript is a whole lot of abstraction and unnecessary cruft which gets in your way of actually using your hardware.

Javascript has a fairly ad-hoc assortment of language features and library facilities which have poor cohesion due to the organic nature in which they evolved over the years.

WebGL (OpenGL ES 2.0) is rather limited and some browsers have no support for it.

Javscript is only the future for hipsters and brainwashed college undergraduates.

>>22
shared-memory concurrency
In a scripting language with no (standardized) bytecode format by design? Not needed and not wanted. It would be wasted here, nothing good would come of it.

There were some mumblings about SIMD though. Probably in connection with the types they included for GL or whatever. It's been a while since I've read about it.

JS isn't really a good systems platform. You're not going to get what you're asking of it. Java is much closer and is still a bad choice (you could actually address 90% of the problems with Java.)

I'm always prepare for both D,C, and JavaScript dominance since i know all of them have potential.
Still using the best language for the job:JavaScript for prototype/web apps, C for speed/pure algorithms ,D for more complex stuff where C is too verbose/painful to implement correctly(D is also more debug-friendly).
I know for certain no functional/high-level language will ever get a shred of QuickBasic popularity outside academia.

Also i've should mention PureBasic for user interface apps and rapid development and FreePascal as safer alternative to C within the same speed class, but they're not that popular and have too many flaws/restrictions in the programs they are capable of.

>>24
I don't understand why you care what language is popular, though, since your coding style and mannerisms ensure you are completely incapable of working with others anyway.

>>26
To use latest libraries and enjoy latest features, maintained by Open Sores Community.
What else? Using some obscure language like Anic maintained by a single autist is not scalable or reliable long-term.

>>27
Actually using a small language like lua is easily maintainable by one person in the long term.