Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
Google Upgrades

Google Releases Dart 1.1 161

Posted by timothy
from the is-it-on-target dept.
rjmarvin writes "Google released version 1.1 of its Dart open-source web programming language today, with new features and improved tools. The Dart Editor is updated with improved debugging, code implementation and more descriptive toolkits, and new UDP (User Datagram Protocol) and documentation support command-line and server-side Dart applications. Google also highlighted benchmarks such as the Richards benchmark, where Dart 1.1 is running 25% faster than JavaScript, as part of the larger competition between Dart and JavaScript in creating more complex applications in the web development space."
This discussion has been archived. No new comments can be posted.

Google Releases Dart 1.1

Comments Filter:
  • 25%?? (Score:2, Insightful)

    by Anonymous Coward
    It doesn't seem much of a speed advantage to lure developers away from the ubiquitous JavaScript.
    • by Anonymous Coward

      My thoughts exactly -- 25% is near nothing. I'd guess that you could achieve the same or more just with built-in native support for common programming patterns, either by detecting those in the JIT compiler or, where necessary, by providing new built-in JS functions for the script writers to use.

      • by Nerdfest (867930)

        There is also the advantage of not Dart being a much nicer and consistent language o use than JavaScript, to the point that it would probably be worth it even if it were a bit slower.

        • by Anonymous Coward

          Why do they advertise the performance then and not the clean-ness and features of the language?

          • by dwater (72834)

            I think because there would be a natural assumption that using dart would actually be slower...they're saying that isn't the case, and in fact it is a bit faster.

            • by dwater (72834)

              someone else mentions :
              > And it appears to be a misquote of TFA too: "Dart’s Javascript output continues to shine. Performance on the Richards benchmark is 25% better than the first release, making runtime comparable to the original JavaScript."

              So, it didn't even mean what is said...it is just 'comparable' after all.

      • 25% is near nothing, hey? Then I suppose you wouldn't mind donating 25% of your net monthly income to charity? Help make the world a better place.

    • Re:25%?? (Score:5, Insightful)

      by lgw (121541) on Thursday January 16, 2014 @03:45PM (#45979455) Journal

      So how long till Google drops this project too? I'm all for new approaches to code that runs in the browser, but I'm a bit hesitant to invest in any technology stack from a company with such a history of dropping projects.

      Is there any sort of non-google dev community supporting Dart itself? Or is it completely dependent on Google at this point?

      • Re:25%?? (Score:5, Informative)

        by JavaTHut (9877) on Thursday January 16, 2014 @04:55PM (#45980065) Homepage

        Dart is really just the evolution of Google's GWT efforts, which they've been pretty good about supporting long-term and cultivating community contributions while also making a lengthy migration path to Dart

      • Re:25%?? (Score:5, Informative)

        by MochaMan (30021) on Thursday January 16, 2014 @07:03PM (#45981127) Homepage

        Dart team member here. The Dart project, like Chromium, is being run as a fully open source project [google.com] accepting patches from Googlers and non-Googlers alike. We've also begun the ECMA Standardization [chromium.org] process, meaning that like JavaScript we'll have a open standard that anyone can implement to. In terms of Dart users, here's a list [dartlang.org] of some. Hope that answers your questions!

        • by nurb432 (527695)

          It really doesn't answer the question of what happens when Google proper loses interest and pulls funding. Does it survive on its own, or vanish?

          If its was a more established language, with decades behind it, there would far less risk in choosing it. But that is always the chicken/egg problem with stuff like this.

          • by swillden (191260)

            It really doesn't answer the question of what happens when Google proper loses interest and pulls funding.

            Google's purpose in creating Dart was so that Google could use it to build its own apps. So if Dart is better than Javascript for that purpose (it is), and if Dart can get enough penetration into the browser market that Google can actually use it for its own services (debatable), then Google will continue supporting it.

            So, rather than asking what Google is going to do, you're better off asking what Microsoft, Apple and Mozilla are going to do.

            • None of them seem to have expressed any interest in including support for Dart in their browsers. Or rather, they have expressed negative interest in supporting Dart.

              • by non0score (890022)

                Which one of them have expressed interest in each of the language the other is proposing? Let me give you a list:

                • Apple: Doesn't care, would rather you program in Objective-C.
                • MS: TypeScript, same issue -- who else do you think is adopting that?
                • Mozilla: LLJS, and how far along is that? Sure, there's asm.js (we're talking about speed here, yeah?), but that's a compile target and not something you'd want to write by hand from scratch -- that and it's not exactly a new language.
                • Google: Dart, and who knows, it m
          • by MochaMan (30021)

            It really doesn't answer the question of what happens when Google proper loses interest and pulls funding. Does it survive on its own, or vanish?

            There's really no possible satisfying answer to that question without knowing the reasons for the hypothetical future loss of interest/support on Google's part. One could ask similar questions about any language with equally unsatisfying answers.

            The project is open source and headed for ECMA standardization - both of those are both very positive from the point of view of future continuity. The best anyone can answer is that if such a situation ever should come to pass, the project is in the best possible po

        • by lgw (121541)

          Well, my actual question was whether the non-google contributors had reached critical mass, such that the project could practically be sustained without Google.

        • by AuMatar (183847)

          Question for you- why compile to Javascript rather than a new interpreter directly in Chrome? The last thing I want to do is write in one language, compile it to another, then need to debug across language boundaries. Its a maintenance nightmare and a complete non-starter for me.

          • by Xest (935314)

            Because Javascript is the only ubiquitous cross browser application engine?

            Of course they want to go further than that (and afaik Chrome can execute it natively?) but getting Microsoft, Mozilla, and Apple to accept that maybe Google, their arch enemy, kind of has a point and hence implementing a Google born technology, is an impossible task.

            It took long enough to get Microsoft to accept HTML5 and that was with Apple, Mozilla, and Google all working together with it.

            It's a chicken and egg scenario, Google ne

          • by MochaMan (30021)

            Question for you- why compile to Javascript rather than a new interpreter directly in Chrome?

            To allow for the widest possible cross-browser app compatibility today. Dart currently support all modern browsers (back to IE9) - forcing users to use Chrome would divide the web, which is the last thing we want to do. That said, we're working on getting the VM into Chrome, but really that comes down to an added performance boost. We expect most developers will target JS to ensure cross-browser compatibility. It's entirely possible (encouraged, even) to deploy both.

            The last thing I want to do is write in one language, compile it to another, then need to debug across language boundaries.

            The good news is that you can write and d

        • Thanks for providing some information.

          In terms of Dart users, here's a list [dartlang.org] of some. Hope that answers your questions!

          I'm not sure this is making much of a case for uptake, though. I took a quick look through that list, and the only non-Google company I really recognize on there is Adobe.

    • Re:25%?? (Score:5, Informative)

      by rlwhite (219604) <rogerwh@gmail.REDHATcom minus distro> on Thursday January 16, 2014 @03:58PM (#45979583)

      And it appears to be a misquote of TFA too: "Dart’s Javascript output continues to shine. Performance on the Richards benchmark is 25% better than the first release, making runtime comparable to the original JavaScript."

      • MOD PARENT UP.

        That is correct. They are saying Dart is 25% faster than earlier Dart. Now almost as good as JavaScript. They are NOT saying it is 25% faster than JavaScript.
        • by JavaTHut (9877)

          Although in some benchmarks the Dart VM is 25% faster than JavaScript (and much more in other benchmarks). The article quotes are a mess. Just look at the actual benchmark numbers at https://www.dartlang.org/performance/ [dartlang.org] for a good idea of what's actually being claimed.

          • Because that site would be completely 100% objective, and certainly would never pick specific benchmarks that might indicate a specific language is better than others.

            Actually, I'm surprised they claim it's only 25% on these hand-picked [one of which was specifically coded for it] benchmarks.

            • by non0score (890022)

              I can understand your sentiment where you may not trust these benchmarks to be unbiased. However, I am fairly certain if they are, it'd be pretty easy to prove. Do you have any sites that show to the contrary?

              Furthermore, one of the major points of all the "new web languages" is to add (optional) typing, which is where the biggest speedups come from modern JS engines -- type inference and unboxing. So to strip them of these features is akin to saying "you're allowed to benchmark C against Python if you made

  • by Anonymous Coward

    Make it so that link in the article is broken. See who notices.

  • Broken Link (Score:4, Informative)

    by rjmarvin (3001897) on Thursday January 16, 2014 @03:34PM (#45979343)
    The first link is broken, it loops back to the submission.
    • Re: (Score:2, Funny)

      by Anonymous Coward

      Maybe it's gone the way of Google Reader already...

    • by rjmarvin (3001897)
      Strike that, it actually goes straight to a 404 error page. Might want to get on that, guys.
    • by Anonymous Coward

      Should be:

      http://sdt.bz/content/article.aspx?ArticleID=67599&page=1

      aka

      http://sdt.bz/67591

      • Get your shit together, the top link of a prominent story remains broken.
        • Yeah, seriously. It could use a little editing too. The submitter apparently wrote that it is 25% faster than Javascript, when the article says that Dart 1.1 produces 25% faster Javascript than Dart 1.0.

          Wait, you're the submitter. Why did you write that it's 25% faster than Javascript?

          • by rjmarvin (3001897)
            I suppose the most honest answer would be, whoops?
            • by gbjbaanb (229885)

              either that or "yay Dart, I love Dart, let me hype it up as much as possible with some slightly vague claim that I can say was a mistake".

              Pah. Let me know when they make NaCl more ubiquitous in browsers.

        • Get your shit together, the top link of a prominent story remains broken.

          This is simply to get you ready for the event that after you get interested in Dart and maybe learn and build something with it, Google will EOL it and cancel the project.

      • by xandroid (680978)

        ...and by rust of course i meant dart...

        • Nice freudian slip to ignite a Chrome vs Firefox flamewar. :)

          Thought bubble: Suppose firefox-nextgen (implemented 99% in rust) includes

          where rust-web is a subset of rust for which FF provides a rust console (repl). Code compiles directly to 'safe' native code without the need for an interpreter nor the overhead of garbage collection. As a Mozilla-proprietary open source language (i.e. like vbscript and Dart, no other browser adopts it), it could offer performance gains due to being a typed 'systems language

  • by Billly Gates (198444) on Thursday January 16, 2014 @03:34PM (#45979349) Journal

    Everyone here would be screaming bloody murder and all MS is trying to sabotage the web again?! But if Google does it then it is cool and innovative.

    I am tired of chrome not implementing W3C standards without using the -webkit to get it to work properly. I am not the only once concerned it is the next IE 6 [pcmag.com] but thankfully there are only a few sites which only work well in Chrome.

    Mozilla Firefox is catching up and has the fasted DOM according to tomshardware and ASM.JS looks to be rather interesting. Unfortunately it is agaisn't Google's interest to support it as they want a closed ecosystem similar to IE 6 and activeX before it.

    I still use Chrome as Firefox is still behind in a few areas, but even IE is catching up and I find both IE and Firefox to use less ram than Chrome.

    • by Nerdfest (867930)

      Is Dart an open language spec? I've assumed it is, but may be incorrect. That's usually what detracts from Microsoft's attempts ... they try to lock you to Microsoft. If this is just as locked, it's just as useless. If it's open, can be forked, etc, if Google goes Microsoft-like, then it's a great idea.

      • by TopSpin (753) on Thursday January 16, 2014 @04:13PM (#45979709) Journal

        Is Dart an open language spec?

        The language spec is CCA 3 and ECMA standards tracked. The source code is BSD.

        Javascript was not an immaculate conception of Berners-Lee, Torvalds and Stallman. It was a product of Mozilla, blessed by nobody and foisted on the world via the defacto browser of the day. It is also more than flawed enough to justify some competition.

        The <script> tag has a "language" attribute for a reason, the curmudgeons of Slashdot notwithstanding.

        • by narcc (412956)

          It is also more than flawed enough to justify some competition.

          How so? Are you familiar with the language at all?

          The tag has a "language" attribute for a reason

          Not any more. It's been depreciated for a while now.

          • by Anonymous Coward

            Deprecated in favor of the type= attribute, which does the same thing.

          • It is also more than flawed enough to justify some competition.

            How so? Are you familiar with the language at all?

            I'm 90% sure that the reason Javascript was built as a prototype-based language [wikipedia.org] is because that was the easiest way to create the interpreter.

            Javascript's scoping rules are horrible, though. Overall the language is fine, if you avoid certain parts, but it's not a great example of anything.

            • by narcc (412956)

              I'm 90% sure that the reason Javascript was built as a prototype-based language is because that was the easiest way to create the interpreter.

              Possibly, but the reason doesn't seem relevant. Being accidentally well-designed isn't exactly a strong criticism!

              Javascript's scoping rules are horrible, though.

              Sorry, what's wrong with them? I've heard that before, but I've yet to hear an actual reason. I've looked, but can't even begin to understand this complaint.

              • ? You like Javascript but not OOP??
                • by Megol (3135005)
                  Why do you assume that a /. poster are logical?

                  BTW the quote isn't from Dijkstra and is fundamentally wrong (OOP originated in Norway).

                  • by narcc (412956)

                    BTW the quote isn't from Dijkstra

                    Yes, it is.

                • by narcc (412956)

                  You like Javascript but not OOP?

                  It's not complicated. Just get past the pitifully superficial and I'm sure you'll be able to puzzle that one out.

                  So... no answer on the alleged problems with Javascript's scoping rules? (I'm not too surprised. It turns out that that particular myth evaporates the instant you look in to it. I guess that's why it's not as popular now as it used to be.)

          • How so? Are you familiar with the language at all?

            Can't speak for him, but I do. Dart has for example a Future class which actually immediately tells you what is going on even when you have derived your own version of it, when Javascript has as many solutions to the concurrency problem as there are programmers. I often think of Javascript as a write-only language, while Dart code actually opens itself up rather well to studying.

            I think that most importantly Dart seems to know what it is and what its purpose is. Javascript was excellent when the web was new

            • by narcc (412956)

              but since jQuery came around it appears more like a tool for inventing infinite ways to shoot yourself in the foot.

              Ugh... Try not to judge javascript on the basis of that abomination.

      • by rlwhite (219604)

        From the spec: (https://www.dartlang.org/docs/spec/latest/dart-language-specification.html#h.jn6bj1irtqj1)
        "Except as otherwise noted, the content of this document is licensed under the Creative Commons Attribution 3.0 License, and code samples are licensed under the BSD License."

      • by MochaMan (30021)

        It's started the process [chromium.org] of ECMA standardization.

    • Any replacement for javascript is a step in the right direction. If they dropped dynamic types altogether it would be a much better alternative than it already is.
      • What I would like to see is a virtual machine. Yeah it will add bloat and waste cpu etc.

        But a VM will increase security and any language can run inside it. So once can use python for example or make up his or her own language and have that JIT compile if it is not cached and run etc. Kind of like people usuing Java to run python with jython but more vm than sandbox like.

        I see Google's native client and NACL and this as just that. They are making Chrome into a ChromeOS virtual machine where it is all a Googl

    • by Anonymous Coward

      I wish Google was less pushy with their tech. Mozilla's making them look worse and worse as the years go by. Google makes a bunch of iffy languages for the web while Mozilla makes one very promising systems language, and upstands Google by showing they could just incrementally improve Javascript at a faster rate. One of them wants the world to change to their model, the other is working to change the world from within.

      I get the impression that Google just wants people to use their tech. With the advent of a

    • "Unfortunately it is agaisn't Google's interest to support it..."

      Are you talking about Firefox or W3C? Because I am pretty sure Google is Firefox's single largest contributor of funds.

    • Where are these "cool and innovative" posts you speak of?

    • If MS did it, it would be totally proprietary and *only* work on MSIE.

      • by spongman (182339)

        Yeah just like Typescript which is open-source and distributed as a node package.

        Oh, wait...

    • by non0score (890022)
      I think every new version of Chrome they remove more of the webkit prefixes (and I don't think they add any more now?). I thought Google isn't even the "owner" of webkit, and Apple is? I'm also guessing this is one of the reasons why Google forked and went their own way.
  • by schneidafunk (795759) on Thursday January 16, 2014 @03:36PM (#45979359)

    My understanding is that Dart will not be really useful until it has native browser support on all browsers. I have not used it, so please correct me if I'm wrong. I'm curious to know if anyone who has experience with it can explain the benefits.

    • Re: (Score:2, Interesting)

      by Billly Gates (198444)

      There is also ASM.JS that Firefox is on the bandwagon with but of course it is against the interest of the 2 organizations to support the opposite as they want to dominate.

      Even if MS supported one or both in future versions of IE both Chrome and Firefox still hold too much a grip before anyone could adopt. It is frustrating as I do not trust Dart as it is highly tied to the chrome native platform and may have patents and licensing issues.

      • by non0score (890022)
        I just read your other post, and you start to sound like someone who is uninformed. Chrome is the only other browser other than FF that "supports" Asm.js (technically speaking, all JS engines support Asm.js, so I don't know what you mean by "against it"). You can even run Epic Citadel on Chrome, and performance has been steadily improving.
    • by LWATCDR (28044) on Thursday January 16, 2014 @03:46PM (#45979469) Homepage Journal

      dart does have a compile to javascript option.

    • Huh? Dart compiles to javascript. Eventually native support would be nice, but since I can't see Microsoft being keen to implement anything by Google, Javascript generated from Dart source will remain dominant for some time. If, as the article says, that the generated Javascript is at near native Javascript speeds, I might be awfully tempted to give Dart a try. I know Javascript has its defenders, but man oh man I find it a difficult, even painful language to do anything useful in.

    • I'm curious to know if anyone who has experience with it can explain the benefits.

      I'm sorry I don't have much experience with it but the benefits should be obvious: it's not javascript.

  • by erice (13380) on Thursday January 16, 2014 @03:44PM (#45979441) Homepage

    I was wondering how it could be 25% faster than javascript when it compiled into javascript so I checked out TFA.

    Performance on the Richards benchmark is 25% better than the first release, making runtime comparable to the original JavaScript.

    So it has 25% faster javascript output. It is not 25% faster than javascript.

    • by Shados (741919)

      Dart has an actual VM of its own. Its probably what they're benchmarking. It can be cross compiled, but it doesn't have to.

    • by JavaTHut (9877)

      See https://www.dartlang.org/performance/ [dartlang.org], in some cases they actually are claiming that the cross-compiled javascript performance outperforms the standardized implementation of an algorithm written originally in Javascript (e.g. the DeltaBlue and Tracer benchmarks)

    • If you follow the links in the article you'll see a chart [dartlang.org] with "dart", "dart2js", and "js v8". The native Dart implementation is 25% faster than JS. The dart2js conversion is slightly slower.

  • Why Dart? Why not a language agnostic runtime and then have Dart target that?
    Then when some new (or old) language wants to run in the browser you don't have to update your browser for it.

    I don't have to upgrade my CPU to run a new language.
    I don't have to upgrade my OS to run a new language.
    Why should I have to upgrade my browser?... its time that browsers have a nice interface that any code could hook into.
    How about LLVM or something as a standard?

    I think Google is already doing this with Native Client...

    • The purpose of dart is probably to push chrome as the leading platform (which browsers are, nowadays). All you ask goes in the opposite direction.
      The LLVM as standard would be a great idea. Java ideals done right.

  • Make JS cooler, why start something new to fix an old problem?

    • by dingen (958134)

      JS is already improving very nicely. ECMA5 is a huge improvement, and it's getting more consistent with each version.

      There's really no need to invent a whole new programming language for the browser. JS does its job just fine already, it just has some legacy issues which are being fixed over time.

      There's no way Dart or whatever language will be "perfect" or even close to something like that, so going for this approach serves no purpose at all.

  • I have not seen any comments as to how well Dart works with JQuery and JQueryUI, I have found these tools allow me to actually make javascript and client side programming actually work.
    • by Shados (741919)

      jQueryUI makes sure your client side programming does NOT work =P

  • Slightly offtopic but Ceylon will run on top of the JavaScript runtime, so this an alternative to Dart.

The tree of research must from time to time be refreshed with the blood of bean counters. -- Alan Kay

Working...