About dart's popularity (again, sorry!) and swift

[Kasper Peulen]

There is a lot of talk about Dart being not popular enough. I think most people here love dart, and everybody thing he knows how Dart can get popular, including me ;) But of course, most of this is just pure speculation and guessing...

I think it is also easy to confuse your own wishes for the future of dart with the things that really makes a language being picked up by the mass.

I think that Dart would get much more popular, if it pushed the tooling support to be closer to what IntelliJ provides for Java or Visual Studio provides for C#. But yeah, to be honest, that is just what I personally would love. I have no idea what the "mass" wants. I once saw a poll that showed the NotePad++ is still the most popular editor...

So therefore let's look some actual data for once, the stackoverflow developer survey.

http://stackoverflow.com/research/developer-survey-2015

One result that got my eye is "the most loved language of 2015". It turns out to be Swift. It is not the most used language of course, but people that use Swift love it, and programmers want to start coding in Swift.

Okay, so that is just a fact: Swift is loved. So, considering Dart 2.0 and breaking changes etc. I think it would be smart to very consider at what Swift is doing. Because they do something right it seems, not only to attract the mass, but also being loved.

One thing that Swift may does very well. Is that it looks very familiar, but with all the familiarity it is never completely the same as you used to, there seems in every feature always something that is a bit better then you are used to from older languages. At least it pretends to be. For example:

1. You can just declare variables as you used to, but you don't need a semicolon anymore.

2. You can declare variables as constant (in dart we would say final). 

But instead of writing the standard 5 letter word const, you can now write a 3 letter word let. That lines up nicely with var.

3. You can just write if/else statements as you used to, but you don't need parenthesis anymore.

1. if name == "world" {
2. print("hello, world")
3. } else {
4. print("I'm sorry \(name), but I don't recognize you")
5. }
6. 4. You can just write for statements like you used to, but instead of writing for(int i = 0; i < 5; i++) you can now write:
   
7. for index in 1...5 {
8. print("\(index) times 5 is \(index * 5)")
9. }
10. 5. You can just write switch statements like you used to, but you don't need break statements anymore:
    
11. switch someCharacter {
12. case "a", "e", "i", "o", "u":
13. print("\(someCharacter) is a vowel")
14. case "b", "c", "d", "f", "g", "h", "j", "k", "l", "m",
15. "n", "p", "q", "r", "s", "t", "v", "w", "x", "y", "z":
16. print("\(someCharacter) is a consonant")
17. default:
18. print("\(someCharacter) is not a vowel or a consonant")
19. }

5 . 6. You can write new instances of objects, but you don't need the new keyword anymore:

1. let someResolution = Resolution()
2. let someVideoMode = VideoMode()
   
3. let vga = Resolution(width: 640, height: 480)

S

YSo well okay, I think you see the pattern. To be honest, all this little sugar, is in practice not that important. In my eyes, in daily programming, having a very good analyser, and debugging experience, is much more important for productivity. However, I can understand that things like this sells very well. If there are going to be breaking changes in dart 2.0, I think it may be smart to consider what Swift is doing here. Swift is the most loved language of 2015. Dart wants to get a little bit more love, maybe learn a bit from a winning team?

a

[Lasse R.H. Nielsen]

It may just be syntactic sugar, but it's *good* sugar :)

/L 'I can haz syntax sugar?!?'

--
For other discussions, see https://groups.google.com/a/dartlang.org/
 
For HOWTO questions, visit http://stackoverflow.com/tags/dart
 
To file a bug report or feature request, go to http://www.dartbug.com/new

To unsubscribe from this group and stop receiving emails from it, send an email to misc+uns...@dartlang.org.

--

Lasse R.H. Nielsen - l...@google.com  

'Faith without judgement merely degrades the spirit divine'
Google Denmark ApS - Frederiksborggade 20B, 1 sal - 1360 København K - Denmark - CVR nr. 28 86 69 84

[Günter Zöchbauer]

IMHO your points are valid they are quite minor.

Using Swift as reference for how a language has to look like to be loved is quite misleading.

I don't know the Apple world and from what I have seen Swift is quite nice, but I think Swift mostly is loved because it allows you to develop for Apple in a better language than ObjectiveC.

Swift doesn't have to fight for it's place. Apple said Swift is the new programming language, and so it is. Apple users usually do what Apple tells them to do :p

I guess there is an even higher number of people who love JS than who love Swift (maybe not percentage of users ;-) ).

IMHO a bigger problem is that Dart is great for people who work only with Dart, but not quite compelling for people in projects where several languages are used.

js and dev_compiler are a big step to make Dart a good fit in projects with existing JS code.

I don't use anything else on the server but I guess there is also room for improvement.

[Jan Mostert]

If Dart was the language of choice for Android, then Google+Dart would have the same leverage as Apple+Swift.

[Filip Hráček]

+1 to Günter: Swift is great, but it's popular because of the reference point (Obj-C). Introducing Swift was a big jump DevExp / productivity jump for iOS devs.

Also, Swift is a more dynamic language than Obj-C, while Dart could be seen as less dynamic than JavaScript. When you remove the bounds from a statically typed language with manual memory management (Obj-C), you'll get a lot of love from the majority of developers. When you add bounds (a.k.a. sanity) to an (arguably overly permissive) scripting language (JavaScript), majority of devs won't appreciate it. Dart (without the tooling) is only appealing for senior devs. 

I personally agree that Dart is going to be successful through tooling.

Thanks for pointing out the SO dev survey. I know the Dart team is looking at data like this. To me, the most interesting part of the survey was [a] Windows (all versions) having 54.5 % of dev market share, [b] all of the most popular IDEs being text editors (NotePad++, Sublime, Vim, ...), [c] tabs winning over spaces, and [d] side projects getting as many hours of a devs week on average as it does.

— Filip 

[Filipe Morgado]

I think Swift's strength is also Apple's commitment. People know Swift is the way to go on iOS/OSX.

And Dart drowns in a see of uncertainty, with no public commitment being shown at all.

Also, Swift designers were a lot bolder, and yet, Swift is still familiar.

From top Google strategists to Dart's language designers, maybe some could use a pair of balls.

[Joao Pedrosa]

Hi,

Businesses just move way too slowly. I was just reading an article about migration to Windows 10 and how 250 million PCs are still on Windows XP, many of which are used by businesses.

A language that targets businesses has to appeal to that snail pace target. Even though many developers wished that weren't the case.

Apple stuck with Objective-C for decades despite other languages showing more "promise." Many developers are still developing with Objective-C even though they were supposed to be using Swift.

Even developers are willing to criticize new technologies as the newest "hippy" stuff. If it doesn't get use in their workplace, it's "hippy." :-)

Lastly, notice that a language's tool-set may work best when it provides the basics or bottom-up. If it is easy to develop in, it will gain traction. Now, if the tool-set is top-to-bottom, the requirements may exceed the expectations, budgets or attention span.

E.g.: command-line tools still matter. :-)

Swift borrowed a lot of love from languages like Ruby. The anti-hero type of languages.

Cheers,

Joao

[Gen]

I think the ambition (or rather dream) for Dart has changed tremendously:
From "conquer the world, replace Javascript and offer great tools" to "satisfy internal clients and give it as open source for those who like it".
Depending on the requirements, Dart is already the best choice for web developers because of async/await, the collection library and maybe Observatory.
Some day Dart will rival Java and C# and Typescript (.d.ts files).

Some more trends:
TypeScript + Dart: https://www.google.com/trends/explore#q=TypeScript%2C%20%2Fm%2F0h52xr1&date=1%2F2013%2036m&cmpt=q&tz=Etc%2FGMT-2
TypeScript + Dart + Swift: https://www.google.com/trends/explore#q=TypeScript%2C%20%2Fm%2F0h52xr1%2C%20%2Fm%2F010sd4y3&date=1%2F2013%2036m&cmpt=q&tz=Etc%2FGMT-2

[Mike]

Although this 'sugar' is not critical, at least it is something positive that improves the developer experience.

Putting Dart into Chrome would over time have made Dart very popular, but that isn't going to happen. The idea of having a good, free, IDE (Dart Editor, Chrome Dev Editor) was great. Maybe we could enhance Dartpad to the point where it could be useful for basic programs (proper imports, project file structure, launch in browser) so there is a better path for those wanting to explore or learn Dart. Providing an easy way to use Dart on Google App Engine would be great too, but we're not there (yet?).

[Jim Simon]

I'd still love to see a native json serialization mechanism that's as simple and easy to use as JSON.parse and JSON.stringify are for JavaScript.  I understand that since type annotations are for documentation only, this is currently hard to accomplish without a lot of extra information at runtime (code bloat).  I'm hopeful that Dart 2.0 or the new JS interior library can resolve this somehow.  I personally believe that this is one of the few areas where JavaScript still has Dart beat.

On Oct 19, 2015 12:11 PM, "Mike" <michael.s...@gmail.com> wrote:

Although this 'sugar' is not critical, at least it is something positive that improves the developer experience.

Putting Dart into Chrome would over time have made Dart very popular, but that isn't going to happen.  The idea of having a good, free, IDE (Dart Editor, Chrome Dev Editor) was great.  Maybe we could enhance Dartpad to the point where it could be useful for basic programs (proper imports, project file structure, launch in browser) so there is a better path for those wanting to explore or learn Dart.  Providing an easy way to use Dart on Google App Engine would be great too, but we're not there (yet?).

[Srdjan Mitrovic]

https://www.dartlang.org/articles/benchmarking/

On Mon, Oct 19, 2015 at 11:30 AM, Gen <gp78...@gmail.com> wrote:

But the results are much different when lim and i change.

lim = 999 and i = 10:
Dart2js hashmap: +/- 3800 milliseconds
Javascript object: +/- 640 milliseconds

lim = 99 and i = 100:
Dart2js hashmap: +/- 120 milliseconds
Javascript object: +/- 210 milliseconds
 


Am Montag, 19. Oktober 2015 20:13:02 UTC+2 schrieb Gen:

I made a small test and I was wrong about dart:collection being useful. The dart2js version is much slower than the comparable Javascript/Typescript version.

Dart code compiled with dart2js : +/- 20 000 milliseconds on my computer

import "dart:collection";

class A {
  int x = 0;
}

generator() {
  return new A();
}

main() {
  Stopwatch stopwatch = new Stopwatch()..start();
  HashMap<String, A> map = new HashMap();
  String key = "";
  int lim = 9999;
  for (var x = 0; x < 1; x++) {
    for (var i = 0; i < lim; i++) {
      key = key + i.toString();
      map.putIfAbsent(key, generator);
    }
    key = "";
    for (var i = 0; i < lim; i++) {
      key = key + i.toString();
      map.remove(key + i.toString());
    }
  }
  stopwatch.stop();
  print(stopwatch.elapsedMilliseconds);
}

Javascript/Typescript: +/- 870 milliseconds on my computer

var A = (function () {
    function A() {
        this.x = 0;
    }
    return A;
})();
function generator() {
    return new A();
}
function main() {
    var process = require("process");
    var map = {};
    var key = "";
    var lim = 9999;
    console.log(map[key] == undefined);
    for (var x = 0; x < 1; x++) {
        for (var i = 0; i < lim; i++) {
            key = key + i.toString();
            if (map[key] == undefined) {
                map[key] = generator();
            }
        }
        key = "";
        for (var i = 0; i < lim; i++) {
            key = key + i.toString();
            delete map[key];
        }
    }
    console.log(process.uptime());
}
main();

--

[Gen]

I do not understand your message.
I tested the output of dart2js and not the Dart VM. If I limited myself to server programming then I would use Dart anyway.
The processes had been long enough to measure in milliseconds and not microseconds.
I removed my posts because the performance of the Dart version is probably related to String creation and management and not to the HashMap implementation.
I did not do further tests to analyze why the Dart version was so much slower when very long strings as keys had been created.

[Srdjan Mitrovic]

Sorry for not being clear. Both V8 and Dart use a JIT and need some time to reach their peak/standard performance. If you do not account for that, your measurement could vary a lot due to the strategies when the optimizations kicks in. 

Cheers,

- Srdjan

On Mon, Oct 19, 2015 at 1:06 PM, Gen <gp78...@gmail.com> wrote:

I do not understand your message.

I tested the output of dart2js and not the Dart VM. If I would limit myself to server programming then I would use Dart anyway.

[Gen]

Thanks for the clarification.
I assume by time you mean number of iterations.
I made all tests with nodejs 4.2.1 and used dart2js from SDK 1.13.0.7.

[Jack Murphy]

The talks for version 2.0 sound nice - but i'd take a stable language and tooling before anymore syntax changes. 

Goal:

Make the tooling so good that no one wants to use other languages

On Mon, Oct 19, 2015 at 4:29 PM, Gen <gp78...@gmail.com> wrote:

Thanks for the clarification.
I assume by time you mean number of iterations.

I made all tests with nodejs 4.2.1 and used dart2js from SDK 1.13.7.

[Gen]

I fully agree with the importance of tooling.

But I think at least the following things will still divide programmers and languages in the next decade:
- Access to the native API (OS, GUI, filesystem, sockets,...) or not.
- Type system. I guess most programmers including myself prefer statically typed languages. I see no good use for dynamically typed languages and programs.
- Garbage collection or not.
- Concurrency model.
- General performance and memory consumption.
- Available libraries.

[Dan Schultz]

IMO, Dart hasn't broken out of its niche because it hasn't provided any material benefit over solutions that already exist. If you look at many of successful languages of the past, they become popular because they made a big leap in dev productivity, or they're popular because of decree. Ruby became popular because of Rails, Java became popular because it provided a higher level language than C/C++ (less memory management, garbage collection, etc), and ObjC/Swift are popular because it's dictated by Apple.

Adding some sugar to a language helps, but it's not going to win mind share.

For web development, my impression is that developers are looking for functional languages, libraries and tooling. Look at the buzz around React and the new comer Elm. If only Dart 2.0 could catch this crowd with language features like record types, pattern matching, non-nullable types and polymorphic methods...

[Filipe Morgado]

It's worth noting that Swift is a first class citizen in Xcode, in a "use it or die" philosophy.

Dart has absolutely no official backup and not even a word of acknowledgment from the Android teams.

And it takes at least half an hour for a newbie to setup a decent development environment.

I don't see a strategy here.

If Android Studio supported apps based on Mojo IPC with, optionally, the Flutter engine with, optionally, the DartVM... THAT would be something to look forward to.

[seb mitchell]

If you have an android device, there is a demo app for flutter in atom https://atom.io/packages/dartlang

[Daniel Joyce]

Google killed dartium in chrome and typescript stole the majority of its thunder wrt being a more strongly typed language for web development because of its backwards compatibility with JavaScript is simpler and it works. It also generates very clean code.

--

For other discussions, see https://groups.google.com/a/dartlang.org/
 
For HOWTO questions, visit http://stackoverflow.com/tags/dart
 
To file a bug report or feature request, go to http://www.dartbug.com/new

To unsubscribe from this group and stop receiving emails from it, send an email to misc+uns...@dartlang.org.

--

Daniel Joyce

The meek shall inherit the Earth, for the brave will be among the stars.

[Sascha Schubring]

Dart runs already on Windows/Linux/Apple and has an incredibly good documented framework and development environment.

As soon as you have a development suite to generate Apps for both Mobile platforms, for what I can predict, the war is over. Because the community around Java / Javascript / Dart ist super active and for what I have seen the code quality is good and usable. For almost everything, you find a Dart port or solution. So any Java developer loves that.

Java is enterprise. To compete with that is really, really tough. Since google autocompiles tools to dart (for what I've seen), now ask your self what that means!

If the universe states: we are running on Dart now!

... ok.

[Gen]

Since google autocompiles tools to dart (for what I've seen), now ask your self what that means!

What does that mean ? And what tools are compiled to Dart ?
Does it mean that Google is actually preparing to replace Java by Dart ?
This makes sense after rumours that even Oracle considers Java old technology or at least one technology platform among others for cloud services.
http://www.javaworld.com/article/2988180/core-java/insider-oracle-has-lost-interest-in-java.html

That the Google Ads team uses Dart is already reason to trust that Dart is important for Google.

[kc]

On Monday, October 19, 2015 at 12:54:04 PM UTC+1, Filipe Morgado wrote:

I think Swift's strength is also Apple's commitment. People know Swift is the way to go on iOS/OSX.

And Dart drowns in a see of uncertainty, with no public commitment being shown at all.

Also, Swift designers were a lot bolder, and yet, Swift is still familiar.

Quoted for truth.

 

From top Google strategists to Dart's language designers, maybe some could use a pair of balls.

Quoted for LOL's. 

But seriously Google need to give a client side roadmap - especially to break down the barriers between web and mobile.

How many client side runtime's -  V8, ART, Dart, Go - does Google need ?

K.

[Lex Berezhny]

But seriously Google need to give a client side roadmap - especially to break down the barriers between web and mobile.

How many client side runtime's -  V8, ART, Dart, Go - does Google need ?

Let's take this one at a time...

V8 - If you build a web browser you need to be able to run JavaScript. If Google made Chrome without a JavaScript VM nobody would use it. The roadmap for web browsers is mostly predetermined, or rather, is not solely up to Google to determine. Not much room to innovate other than making things faster, which Google did with V8.

ART - This one is mostly due to momentum. Android has to keep moving forward and competing. Even if Flutter might be on the horizon they can't just freeze all Android development, that would be a disaster. We still have many years of Java on Android and this path needs to work well (just as Objective-C still has many years with Apple until more people switch to Swift).

Dart - Seems like the current long term bet for Google.

Go - Not client side runtime. My understanding is that this was designed for server side applications.

[kc]

Opportunity:

1. Flutter

2. Dart 2.0 - more expressive in terms of syntax/semantics with less boilerplate/ceremony.

Deliver 1. and 2. together to give a far more faster, fun but structured way of developing Android (and iOS) Material Design apps. (Developing Java Android apps is awful imo).

Tooling is important. But most important - the semantics and performance of the runtime with Flutter and the integration in the Android platform. Third parties can create tools but they can't do this.

The Dart runtime model is attractive because it's so Smalltalk-ish simple:

// named params - no overloading - msg resolved at runtime - types are annotations only (no runtime effect)
obj.msg(param1: 1, param2: "two");

But can this runtime model with GC/JIT compete with ART (no JIT/AOT) and Swift (no GC) on mobile with Flutter.

K.

[kc]

On Tuesday, October 20, 2015 at 3:24:05 PM UTC+1, Lex Berezhny wrote:

But seriously Google need to give a client side roadmap - especially to break down the barriers between web and mobile.

How many client side runtime's -  V8, ART, Dart, Go - does Google need ?

Let's take this one at a time...

Lets do that...

 

V8 - If you build a web browser you need to be able to run JavaScript. If Google made Chrome without a JavaScript VM nobody would use it. The roadmap for web browsers is mostly predetermined, or rather, is not solely up to Google to determine. Not much room to innovate other than making things faster, which Google did with V8.

No. See Strong mode and SoundScript. Also WASM.

https://developers.google.com/v8/experiments?hl=en

 

ART - This one is mostly due to momentum. Android has to keep moving forward and competing. Even if Flutter might be on the horizon they can't just freeze all Android development, that would be a disaster. We still have many years of Java on Android and this path needs to work well (just as Objective-C still has many years with Apple until more people switch to Swift).

No. The ART runtime could migrate to a new toolkit like Flutter. And ART really isn't Java anyway - there's no Java byte code interface - the interface is still Dalvik. So Google could bless something like Kotlin and evolve in it's own direction via Jack and Jill.

 

Dart - Seems like the current long term bet for Google.

The problematic word here is seems. If there's an memo as entertaining and informative as the infamous one from 2010 giving a current roadmap I wouldn't mind seeing it.

 

Go - Not client side runtime. My understanding is that this was designed for server side applications.

Some 'experimental' work is being done in this area on Android.

https://github.com/golang/mobile

Google loves experiments. Apple, MS and Facebook all seem rather more focused.

K.

 

[Matthew Butler]

On Tuesday, October 20, 2015 at 11:24:05 AM UTC-3, Lex Berezhny wrote:

Go - Not client side runtime. My understanding is that this was designed for server side applications.

Go started off planning to be a server side language. That was the initial design. But it has been becoming a general purpose language. Similar to Flutter, Go has an experimental mobile shift as well.

https://github.com/golang/mobile  

[Daniel Joyce]

Go was supposed to be a server side and even systems programming language. Google pushed it as a language suitable for creating operating systems. Then they backed off a bit. Now people mostly treat it as a 'more performant python' server side language.

--

For other discussions, see https://groups.google.com/a/dartlang.org/
 
For HOWTO questions, visit http://stackoverflow.com/tags/dart
 
To file a bug report or feature request, go to http://www.dartbug.com/new

To unsubscribe from this group and stop receiving emails from it, send an email to misc+uns...@dartlang.org.

[Lex Berezhny]

On Tue, Oct 20, 2015 at 10:46 AM, kc <kevin...@gmail.com> wrote:

V8 - If you build a web browser you need to be able to run JavaScript. If Google made Chrome without a JavaScript VM nobody would use it. The roadmap for web browsers is mostly predetermined, or rather, is not solely up to Google to determine. Not much room to innovate other than making things faster, which Google did with V8.

No. See Strong mode and SoundScript. Also WASM.

https://developers.google.com/v8/experiments?hl=en

Removing features from a language so that it can be faster on a specific runtime but still run on other JS VMs is not the kind of groundbreaking innovation that will change things.

Roadmap for V8: make JS code run faster in the browser.

This was the original goal and this is still the goal of the projects you mention.

There was a time when V8 being fast was innovative. Somehow I don't think there will be new breakthroughs on that front to cause everyone to stop what they are doing and pay attention again.

ART - This one is mostly due to momentum. Android has to keep moving forward and competing. Even if Flutter might be on the horizon they can't just freeze all Android development, that would be a disaster. We still have many years of Java on Android and this path needs to work well (just as Objective-C still has many years with Apple until more people switch to Swift).

No. The ART runtime could migrate to a new toolkit like Flutter. And ART really isn't Java anyway - there's no byte code interface - the interface is still Dalvik. So Google could bless something like Kotlin and evolve in it's own direction via Jack and Jill.

Are you saying that the .dex format makes no assumptions about the input language being Java? And as a result the subsequent tools in the chain?

Java is the official language for developing Android apps. ART is just another step in the optimization of that tool chain.

I don't think this is ambiguous or confusing. The roadmap is pretty clear here.

 

Dart - Seems like the current long term bet for Google.

The problematic word here is seems. If there's an memo as entertaining and informative as the infamous one from 2010 giving a current roadmap I wouldn't mind seeing it.

Okay, here is something from less than two months ago:

https://github.com/dart-lang/dart_enhancement_proposals/blob/master/Meetings/2015-09-02%20Language%20Meeting.md

That pretty clearly states the top goal is mobile app development.

 

Go - Not client side runtime. My understanding is that this was designed for server side applications.

Some 'experimental' work is being done in this area on Android.

https://github.com/golang/mobile

I don't think you can even begin to compare the investments Google is making in Flutter vs the golang mobile project. Not to mention that there doesn't appear to be any mention of UI framework for go on mobile where as for Flutter you get a full dedicated UI.

On the Dart side you see the core language team putting mobile app development as the #1 priority. Are there any meeting notes from golang stating Golangs strategy to put mobile app development as the #1 priority?

[Sascha Schubring]

@Gen

its very simple

go here:
https://pub.dartlang.org/search?q=google

its just my guess.

After the sad fight with oracle (I rather had seen Java in hands of IBM or Google), I find Dart really has the power for a big change. Its not only, that it can replace Java but additionally has so many good things.

Think of what we are arguing about. A couple of years ago I doubted anything like that could happen. IBM+Oracle  its not the company names but what they are doing. Java applications are so complex. Business intelligence, distributed, embedded etc. and now you bring it to the cloud in a sandbox. GWT was a nice Idea but with Dart you save compile time and get documented use cases, repository etc. Anybody could do anything with that. Its not only a good OOP language you can create super large projects with. It is a killing machine!  I only recognized Dart after 3 years of development history! But we are in 2015 now and Java has a 20 years history. Compiler technology, design patterns etc. we are on a completely different level today. So maybe people create stuff like that for fun, but... I don't think so.
Dart is on par with Java 8. This is serious. I am just attending this conversation because its so ridiculous, I can not resist.  Talk Dart to the death, so that it can arise like a phoenix. "Phoenix Script" would actually be a nice naming for Dart. JS and Java reborn in Dart.

[Alain Ekambi]

"GWT was a nice Idea but with Dart you save compile time and get documented use cases, repository etc."

Sorry but GWT still is a nice idea. Wayyy advanced than dart. And  since we dont have a dart VM in the browser Dart will need a compilation step for web apps too.

--
For other discussions, see https://groups.google.com/a/dartlang.org/
 
For HOWTO questions, visit http://stackoverflow.com/tags/dart
 
To file a bug report or feature request, go to http://www.dartbug.com/new

To unsubscribe from this group and stop receiving emails from it, send an email to misc+uns...@dartlang.org.

--

Alain Ekambi

Co-Founder

Ahomé Innovation Technologies

http://www.ahome-it.com/

[Daniel Joyce]

If we get Strong mode / Soundscript, async, and es6/es7 features such as collections in the next year or so, then what will keep Dart compelling? Typescript ate up a lot of the reasons for adoption. 

[Sascha Schubring]

"GWT was a nice Idea but with Dart you save compile time and get documented use cases, repository etc."

Sorry but GWT still is a nice idea. Wayyy advanced than dart. And  since we dont have a dart VM in the browser Dart will need a compilation step for web apps too.

I hated the compilestep in GWT, its not the same like dart2js translation. That was my no. 1 reason to avoid use for inferior projects. You could easily use GWT for any project, but not if you need to get things done real quick.

Not sure what you mean by "advanced". GWT Designer and Compile Targets such as Android are not so much of an advantage. Servlet RPC Communication also even if helpful is rather done with websockets directly, since we are in 2015 not 2012.

In total you can subtract a lot of advantageous boilerplate, how ever I should not lean to far out of the window with that statement.

[Jan Mostert]

GWT was a great idea for its time when browser incompatibility in terms of JS was at an all-time high, things are much better now and I'd argue GWT is semi-obsolete, unless you enjoy that horrendous long latency between making a one line change and waiting, refreshing and hitting permgen then restarting the devmode or needing Java developers to do the designer's work.
It just took forever getting anything done in GWT and then on large projects, compile time was insane.

Dart on the other hand got rid of the boilerplate and the latency GWT had and compile time is a small fraction for similar sized projects.

[Alain Ekambi]

I guess u have not checked GWT in a while. Devmode is gone.

Nothing against Dart just don't see it solving something GWT has not.

[Sascha Schubring]

 
@Alain

Its true. I used it 2011 last time, when I wanted to compile Jphysx to JS.

Dart is focussed on Web Development as it is supposed to be done these days. The language design is advantageous as well as the packages. Server side may still not be competitive with GWT but its not just about GWT there, its about the Development Stack. Application server, distributed session and objects etc. a lot of things you would need to solve somehow if you wanted to fully avoid Java. So don't get me wrong :)  I just love how you do things in Dart. Its much more light weight.

If google can use Java with any further problems, they can kill Dart. If they wanted to kill Java, they would absolutely have the option! At least I am convinced of that and if I was in charge, I would get rid of Java in favour of a more intelligent virtual machine with a language design that makes things possible. And also because I don't like oracle. And it would be strategically advantageous, if they wanted to promote the chrome browser and system. Because then, if you have stuff running better in Dart, people will switch. Like they preferred chrome in favour of other other browsers but then when chrome sucked people went back to FF. Thats how it works. I assume with Dart things will run much more stable. You even can export to Dart within Flash Pro. So anything you would need to get Dart to the top exists.

I have to learn where swift fits into this...

[kc]

On Tuesday, October 20, 2015 at 5:29:32 PM UTC+1, Lex Berezhny wrote:

On Tue, Oct 20, 2015 at 10:46 AM, kc <kevin...@gmail.com> wrote:

V8 - If you build a web browser you need to be able to run JavaScript. If Google made Chrome without a JavaScript VM nobody would use it. The roadmap for web browsers is mostly predetermined, or rather, is not solely up to Google to determine. Not much room to innovate other than making things faster, which Google did with V8.

No. See Strong mode and SoundScript. Also WASM.

https://developers.google.com/v8/experiments?hl=en

Removing features from a language so that it can be faster on a specific runtime but still run on other JS VMs is not the kind of groundbreaking innovation that will change things.

Roadmap for V8: make JS code run faster in the browser.

This was the original goal and this is still the goal of the projects you mention.

There was a time when V8 being fast was innovative. Somehow I don't think there will be new breakthroughs on that front to cause everyone to stop what they are doing and pay attention again.

ES7/SoundScript/WASM do look pretty innovative. V8/MS teams are collaborating.

 

ART - This one is mostly due to momentum. Android has to keep moving forward and competing. Even if Flutter might be on the horizon they can't just freeze all Android development, that would be a disaster. We still have many years of Java on Android and this path needs to work well (just as Objective-C still has many years with Apple until more people switch to Swift).

No. The ART runtime could migrate to a new toolkit like Flutter. And ART really isn't Java anyway - there's no byte code interface - the interface is still Dalvik. So Google could bless something like Kotlin and evolve in it's own direction via Jack and Jill.

Are you saying that the .dex format makes no assumptions about the input language being Java? And as a result the subsequent tools in the chain?

Scala, Kotlin, Celyon all target Java byte code.

 

Java is the official language for developing Android apps. ART is just another step in the optimization of that tool chain.

No reason it couldn't be something like Kotlin with the eventual goal of liberation from Java given the issues with Oracle. This isn't a strategy Google have chosen but it does make some sense.

 

 

Dart - Seems like the current long term bet for Google.

The problematic word here is seems. If there's an memo as entertaining and informative as the infamous one from 2010 giving a current roadmap I wouldn't mind seeing it.

Okay, here is something from less than two months ago:

https://github.com/dart-lang/dart_enhancement_proposals/blob/master/Meetings/2015-09-02%20Language%20Meeting.md

That pretty clearly states the top goal is mobile app development.

I would like to see a comment from Hiroshi Lockheimer rather than a comment in a meeting.

 

 

Go - Not client side runtime. My understanding is that this was designed for server side applications.

Some 'experimental' work is being done in this area on Android.

https://github.com/golang/mobile

I don't think you can even begin to compare the investments Google is making in Flutter vs the golang mobile project. Not to mention that there doesn't appear to be any mention of UI framework for go on mobile where as for Flutter you get a full dedicated UI.

Well maybe it's time for Google to focus and give dev's a firmer roadmap. MS are back in the game.

K.

[kc]

On Tuesday, October 20, 2015 at 9:19:27 PM UTC+1, Daniel Joyce wrote:

If we get Strong mode / Soundscript, async, and es6/es7 features such as collections in the next year or so, then what will keep Dart compelling? Typescript ate up a lot of the reasons for adoption. 

Flutter went with Dart rather than V8/ES+. That's the only reason I'm remotely interested in this project. You have mentioned JavaFxScript - that's the vibe Dart w/Flutter should go for. Fun but structured with decent tooling. Make it really easy to turn out Material Design apps quickly.

K.

[Yegor]

 

But can this runtime model with GC/JIT compete with ART (no JIT/AOT) and Swift (no GC) on mobile with Flutter.

JIT or AOT? Why not both? ;)

[kc]

On Saturday, October 24, 2015 at 7:51:17 PM UTC+1, Yegor wrote:

 

But can this runtime model with GC/JIT compete with ART (no JIT/AOT) and Swift (no GC) on mobile with Flutter.

JIT or AOT? Why not both? ;)

Is this a fresh approach by the Dart VM team? For iOS?

The Dart VM is way too much of a black box. 

Some info on how the the Dart VM is engineered - especially to interact with the Flutter engine (with a nice surface syntax) - would help make the sale.

The V8 engine team have a tech orientated blog:

http://v8project.blogspot.co.uk/

K. 

[Bob Nystrom]

On Mon, Oct 26, 2015 at 5:22 AM, kc <kevin...@gmail.com> wrote:

The Dart VM is way too much of a black box. 

Some info on how the the Dart VM is engineered - especially to interact with the Flutter engine (with a nice surface syntax) - would help make the sale.

I don't think many of the VM folks follow misc@, but if you ask on vm-dev@, they might see it.

Cheers!

– bob

[kc]

I think it's something that whoever is leading this project should consider. Seeing how the lang/VM/Flutter design and implementation hang together would help sell Dart.

You managed to do a good job of giving an design overview and the niche/market for your Wren lang project.

K. 

[krupal shah]

First of all Sorry, If I am posting to this older thread.

But; IMHO all the points here are valid nad Dart should feel like a modern language rather than it still looks like Java.

I think Dart would be a great fit for Android developers. Look at this and this. Now swift has been coined as a programming language for Android.

Oracle seems to continue suing Google for Java. And to be fair, no Android Developer is quite satisfied with Java.

We all need a modern language with a modern look and feel. In many cases; Dart has these feel but again it just looks old in many cases.

Dart has a great standard library which Swift can never build. Dat has nice Asynchrony support which is must requirement for Mobile Applications but Swift does not have it.

There isn't anything innovative in Swift while Dart has optional typing, mixins, futures and streams, method cascades and many things that many modern programming languages doesn't have. Apple just decorated all old idioms with their beautiful words and it worked.

We don't want another cross platform framework like Flutter. Android framework is intelligently written and all we need to support Dart for building native Android applications.

Optional semicolons, Removing unnecessary brackets, var, val nad let, powerful pattern matching...these are easy things to implement. Dart is still in 1.x and going to be in 2.x soon. Why can't Google just decorate Dart to feel like a modern language and make it default programming language for mobile applications?

Well, think about one programming language for all - mobile apps, web, server and IoT devices. 

Thanks. 

On Monday, October 19, 2015 at 8:45:45 AM UTC+5:30, Kasper Peulen wrote:

There is a lot of talk about Dart being not popular enough. I think most people here love dart, and everybody thing he knows how Dart can get popular, including me ;) But of course, most of this is just pure speculation and guessing...

I think it is also easy to confuse your own wishes for the future of dart with the things that really makes a language being picked up by the mass.

I think that Dart would get much more popular, if it pushed the tooling support to be closer to what IntelliJ provides for Java or Visual Studio provides for C#. But yeah, to be honest, that is just what I personally would love. I have no idea what the "mass" wants. I once saw a poll that showed the NotePad++ is still the most popular editor...

So therefore let's look some actual data for once, the stackoverflow developer survey.

http://stackoverflow.com/research/developer-survey-2015

One result that got my eye is "the most loved language of 2015". It turns out to be Swift. It is not the most used language of course, but people that use Swift love it, and programmers want to start coding in Swift.

Okay, so that is just a fact: Swift is loved. So, considering Dart 2.0 and breaking changes etc. I think it would be smart to very consider at what Swift is doing. Because they do something right it seems, not only to attract the mass, but also being loved.

One thing that Swift may does very well. Is that it looks very familiar, but with all the familiarity it is never completely the same as you used to, there seems in every feature always something that is a bit better then you are used to from older languages. At least it pretends to be. For example:

1. You can just declare variables as you used to, but you don't need a semicolon anymore.

2. You can declare variables as constant (in dart we would say final). 

But instead of writing the standard 5 letter word const, you can now write a 3 letter word let. That lines up nicely with var.

3. You can just write if/else statements as you used to, but you don't need parenthesis anymore.

1. if name == "world" {
2. print("hello, world")
3. } else {
4. print("I'm sorry \(name), but I don't recognize you")
5. }
6. 4. You can just write for statements like you used to, but instead of writing for(int i = 0; i < 5; i++) you can now write:
   
7. for index in 1...5 {
8. print("\(index) times 5 is \(index * 5)")
9. }
10. 5. You can just write switch statements like you used to, but you don't need break statements anymore:
    
11. switch someCharacter {
12. case "a", "e", "i", "o", "u":
13. print("\(someCharacter) is a vowel")
14. case "b", "c", "d", "f", "g", "h", "j", "k", "l", "m",
15. "n", "p", "q", "r", "s", "t", "v", "w", "x", "y", "z":
16. print("\(someCharacter) is a consonant")
17. default:
18. print("\(someCharacter) is not a vowel or a consonant")
19. }

5 . 6. You can write new instances of objects, but you don't need the new keyword anymore:

1. let someResolution = Resolution()
2. let someVideoMode = VideoMode()
   
3. let vga = Resolution(width: 640, height: 480)

S

YSo well okay, I think you see the pattern. To be honest, all this little sugar, is in practice not that important. In my eyes, in daily programming, having a very good analyser, and debugging experience, is much more important for productivity. However, I can understand that things like this sells very well. If there are going to be breaking changes in dart 2.0, I think it may be smart to consider what Swift is doing here. Swift is the most loved language of 2015. Dart wants to get a little bit more love, maybe learn a bit from a winning team?

a

[Smit Joshi]

"Also, Swift is a more dynamic language than Obj-C, while Dart could be seen as less dynamic than JavaScript. When you remove the bounds from a statically typed language with manual memory management (Obj-C), you'll get a lot of love from the majority of developers. When you add bounds (a.k.a. sanity) to an (arguably overly permissive) scripting language (JavaScript), majority of devs won't appreciate it."

This is only an excuse. Isn't it?

On Monday, October 19, 2015 at 1:29:33 PM UTC+5:30, Filip Hráček wrote:

+1 to Günter: Swift is great, but it's popular because of the reference point (Obj-C). Introducing Swift was a big jump DevExp / productivity jump for iOS devs.

Also, Swift is a more dynamic language than Obj-C, while Dart could be seen as less dynamic than JavaScript. When you remove the bounds from a statically typed language with manual memory management (Obj-C), you'll get a lot of love from the majority of developers. When you add bounds (a.k.a. sanity) to an (arguably overly permissive) scripting language (JavaScript), majority of devs won't appreciate it. Dart (without the tooling) is only appealing for senior devs. 

I personally agree that Dart is going to be successful through tooling.

Thanks for pointing out the SO dev survey. I know the Dart team is looking at data like this. To me, the most interesting part of the survey was [a] Windows (all versions) having 54.5 % of dev market share, [b] all of the most popular IDEs being text editors (NotePad++, Sublime, Vim, ...), [c] tabs winning over spaces, and [d] side projects getting as many hours of a devs week on average as it does.

— Filip 

[Marcello Dias]

This is the kind of thread that will never die.

Here my two cents,and maybe I myself won´t agree with me in the future.

Why DART was named DART?

At least I think because it was a promisse of clean conde,running in a clean VM.

See this in Youtube

https://www.youtube.com/watch?v=5AqbCQuK0gM

Now I see myself clapping every Lars speech and

Larry Page or Sergey Brin clapping every Anders

Speech.

Believe me or not,I don´t thing that Anders itself believes in a word that he said.

I remember having some talkings in the NET with people that have choosen TypeScript over DART.

Their arguments:You can have TypeScript and JavaScript Libraries in the same project,all browses VMS run javaScript.

My arguments:Performance is the most important thing in Internet where resources are lower and more expensive than Desktop Applications.

I Can use Dart both in the Client and in the Server.

JavaScript code is garbage and should never had been written,it was written in 20 days,and it is really bad for what could have been done 20 years ago.

It is like having a religious ex- seriall killer living in your house.

The other browsers will have to acept DART or become turtles.

I Don´t care how many years google would take to do this,I don´t mind having two browsers,one to run stupid old javascript code,and one to run corporate code.

But Google said We will never have a DART VM,thats really disapointing.

At first i thougt that´s pure marketing.

But even for the market point of view,that was really bad,I think it is affecting the language aceptance and users Growth.

Being revolutionary was one of DART´s most appealing reasons.

What I´m traying to say is:Its harder to Sell Dart to your Boss today than it was in 2014.

“All men should keep their word, kings most of all. “

I think thats from Game of Thrones

Google is not acting as a King, at least in my opinion.

In fact I don´t produce Code rigth now,I left my job as a CIO ,I know have time to code again,Just a Starter with DART.

Marcello

[tomaszkubacki]

Reality is JS won the game - however it's dev nightmare to maintain large projects in it.

There is still contest on which lang compiled to JS will have best dev experience for non-small projects.

And here Dart and TS will compete. I consider Dart to be nicer lang than TS, but it seems TS is more popular at the moment.

I think https://github.com/dart-lang/dev_compiler could change the game

[Marcello Dias]

tomaszkubacki

>>Reality is JS won the game - however it's dev nightmare to maintain large projects in it.

Every season there are new games,specially when we talk about applications.

>>And here Dart and TS will compete. I consider Dart to be nicer lang than TS, but it >>seems TS is more popular at the moment.

IF Global economy helps and people decides to rewrite their applications, we re leaving kind of a new wave with Web Components. For what I´ve seen so far, Polymer is so ahead of other initiatives, that those who offer better tools to work with it will win the game. What I could not imagine(MS and Apple would never do that) is that Angular would choose TypeScript over Dart. In my head in 2016,both Angular and Polymer would prefer DART. In my head in 2016,we would have DART running in Chrome without a transpiler. Google is probably the most democratic and fair company of the world. Marcello DIas

--
For other discussions, see https://groups.google.com/a/dartlang.org/
 
For HOWTO questions, visit http://stackoverflow.com/tags/dart
 
To file a bug report or feature request, go to http://www.dartbug.com/new

---
You received this message because you are subscribed to the Google Groups "Dart Misc" group.

[Marcello Dias]

Forget to say that Material Design is in that new wave also.

[Song Lusic]

Swift's strength is Apple's commitment. Who wants to learn/use a language which might die at any time soon?

Google has abandoned many projects in the past and decided not integrate Dart VM into chrome. Google never commits to Dart, even Golang.

It doesn't seem that Dart and golang will die soon, but who knows? People are just not confident about the future of Dart, including me. That's all.

With Apple's commitment, Swift has a brighter future. At least swift will not die in about 20 years and Apple will use it as a primary language for sure.

Even if Dart will not die soon, Google can use any other languages in its projects. For example, TensorFlow chooses python.

I am currently using nodejs for the server side and plan to switch to another language. 

Comparing golang and swift, even if golang is more suitable and matured in the server side, I am still considering swift (vapor/perfect framework).

Even if swift server side framework is not production ready and in a very early stage, swift does have a brighter future.

The reason why I come across this thread, it's because there is a news about google fuchsia OS which has dart and flutter builtin. This may give some brighter future to Dart if fuchsia succeeds. 

PS: I knew many people use Dart internally and Google AdWords uses Dart. I read Filip Hracek's post about Dart. But google still doesn't commit anything about Dart and Golang.

[Günter Zöchbauer]

I wouldn't expect Swift to vanish anytime soon but I would expect Apple to cease to support other platforms anytime if they don't get the benefit out of it that they expect.

[Bob Nystrom]

On Sat, Aug 20, 2016 at 12:58 PM, Song Lusic <sgo...@gmail.com> wrote:

With Apple's commitment, Swift has a brighter future. At least swift will not die in about 20 years and Apple will use it as a primary language for sure.

In the past twenty years, Apple canceled Dylan, HyperTalk, and Object Pascal. They certainly aren't putting much effort into Objective-C these days, either.

Every large technology corporation with a long history has a big pile of terminated projects. Death is part of the technology lifecycle.

Cheers!

– bob

[tatumizer-v0.2]

> Death is part of the technology lifecycle

Of any lifecycle. By definition

Sorry to break bad news.

[Lex Berezhny]

Does this mean lifecycles are unethical? Also, does it mean we're all going to die? This discussion has taken a new turn :-( What is the Dart team doing to fight lifecycles?

[Mike]

OMG does this mean that if I use Dart, I'll die?

[Steve Lympany]

Eventually

[Filipe Morgado]

If it wasn't for lifecycles, I guess Adam and Eve would still be by themselves, punching cards :P

[Lex Berezhny]

Oh, the humanity! (pun intended)

On Aug 23, 2016 7:10 PM, "Filipe Morgado" <pix...@gmail.com> wrote:

If it wasn't for lifecycles, I guess Adam and Eve would still be punching cards :P

[JosephLi]

Please don't forget NewtonScript!!!

[Filipe Morgado]

Pffff ... Who needs programming languages anyway?!?

We should just all sit down and wait for AI to be able to compile our natural language.

[Günter Zöchbauer]

On Wednesday, August 24, 2016 at 12:51:40 AM UTC+2, Mike wrote:

OMG does this mean that if I use Dart, I'll die?

Sure, but if you don't, you will die unhappy ;-)

[Bob Nystrom]

On Tue, Aug 23, 2016 at 4:26 PM, JosephLi <josep...@gmail.com> wrote:

Please don't forget NewtonScript!!!

Oh, you're right!

You know, NewtonScript is actually a pretty interesting language. It was, I think, the first prototypal language targeted for real industry use. It's also the only language I know that directly bakes the concept of nesting UI widgets into the language semantics by having hardcoded support for two kinds of inheritance/delegation: "is-a" and "part-of".

– bob

[kc]

On Tuesday, August 23, 2016 at 6:11:13 PM UTC+1, Bob wrote:

On Sat, Aug 20, 2016 at 12:58 PM, Song Lusic <sgo...@gmail.com> wrote:

With Apple's commitment, Swift has a brighter future. At least swift will not die in about 20 years and Apple will use it as a primary language for sure.

In the past twenty years, Apple canceled Dylan, HyperTalk, and Object Pascal. They certainly aren't putting much effort into Objective-C these days, either.

Need to mention Apple pre vs post Jobs.

Jobs canned Dylan and the Newton. Then Apple backed Objective C plus frameworks - avoiding the siren song of Java on OS X and JS on iOS. Developers thus had a platform to build on.

Now Apple have a new language - Swift - which offers an evolutionary path from the Objective C frameworks. 

So developers want to get a feel if Google is giving backing to Dart the same way Apple backed Objective C and now Swift.

(An irony is that Dart is more of what some devs were after - Objective C without the C - then Swift).

 

Every large technology corporation with a long history has a big pile of terminated projects. Death is part of the technology lifecycle.

Some Google tech dies before it was born. Chrome Dev Editor - which should have been something like Atom.

K. 

[Marcello Eduardo de Oliveira Dias]

Polymer.Dart seems to be an abort.

Death before official init of life.

Or I guess it is an elephant pregnancy.

2016-08-25 12:14 GMT-03:00 kc <kevin...@gmail.com>:

On Tuesday, August 23, 2016 at 6:11:13 PM UTC+1, Bob wrote:

On Sat, Aug 20, 2016 at 12:58 PM, Song Lusic <sgo...@gmail.com> wrote:

With Apple's commitment, Swift has a brighter future. At least swift will not die in about 20 years and Apple will use it as a primary language for sure.

In the past twenty years, Apple canceled Dylan, HyperTalk, and Object Pascal. They certainly aren't putting much effort into Objective-C these days, either.

Need to mention Apple pre vs post Jobs.

Jobs canned Dylan and the Newton. Then Apple backed Objective C plus frameworks - avoiding the siren song of Java on OS X and JS on iOS. Developers thus had a platform to build on.

Now Apple have a new language - Swift - which offers an evolutionary path from the Objective C frameworks. 

So developers want to get a feel if Google is giving backing to Dart the same way Apple backed Objective C and now Swift.

(An irony is that Dart is more of what some devs were after - Objective C without the C - then Swift).

K.

 

Every large technology corporation with a long history has a big pile of terminated projects. Death is part of the technology lifecycle.

[Ованес]

Not a serious conversation on serious subjects.

[Ованес Ованесян]

But Dart is already quite popular.
Or you need to Dart became the most popular?