Google is building a Chrome app-based development environment

http://thenextweb.com/google/2013/11/21/google-building-chrome-app-based-development-environment-using-dart-polymer/?fromcat=all

by iamtechaddict • 13 years ago

128 97 13 years ago

thenextweb.com

https://github.com/dart-lang/spark

isaacwaller• 13 years ago

I don't understand Chrome apps. They require installation and only work on Chrome, so they come with none of the benefits of the web. They are slower and uglier than native Windows/Mac/Linux apps. What's the point?

balac• 13 years ago

The main benefits as I see them are:

* Its an easy way to build simple(ish) cross-platform apps using HTML/CSS/Javascript. * A native app platform for Chromebooks * Promotion of your app the Chrome store. * For users their apps are available anywhere Chrome is.

pkulak2• 13 years ago

They are just apps that use the ChromeOS development library (JS, HTML, CSS and some custom APIs). Think of it like Cocoa on OS X. Except they can also run in the Chrome browser on other OSes. I don't understand why people get so riled up about this.

mbreese2• 13 years ago

Is there someway to do standalone installs as well? It seems like that might be against Google's motives in getting everyone to use Chrome as a browser, but it would make it easier to distribute applications in a "traditional" manner.

Plus, it would be nice to have dedicated icons, etc...

archagon3• 13 years ago

I think people get riled up because the humble browser is rapidly turning into this: http://i.imgur.com/XNRtCGh.jpg

_urga• 13 years ago

Chrome Apps as far as I know can only be installed via a proprietary web store. You can't actually package them up as an exe or dmg to distribute yourself. They also don't have access to the system tray - they're not really first-class standalone desktop apps.

bmelton• 13 years ago

Define "installation"? My definition of installation is whether or not I can install it on a PC that is locked down by a system's administrator. While it may (or may not) be possible to lock down Chrome from being able to install apps, I've never been anywhere that did, despite those machines being otherwise locked down.

Otherwise, the benefits are

- being able to develop in just HTML/CSS/JS

- being able to deploy to Windows, Linux, OSX, anywhere Chrome runs

- No infrastructure costs to host / distribute (unless your app has server-side functionality)

etc., etc.

magicalist• 13 years ago

In addition to all the siblings' points, it does look like we're going to be able to have Firefox and Chrome compatible apps someday (I can't find a link right now, but it's been a secondary goal for both teams, Mozilla especially as they define APIs for Firefox OS that will be compatible with the open (but with permissions) web). Hopefully the teams keep working on this.

haberman4• 13 years ago

Just to add a few:

- They work offline, like native apps, but are cross-platform, like web apps.

- You can develop and distribute them without running your own server infrastructure (the Chrome store will distribute them and the run locally).

- They work on ChromeOS

indubitably• 13 years ago

That’s a funny definition of cross-platform. You mean, it’s cross-platform if you use Google’s browser.

And as far as offline goes, that’s an active area of development within the HTML5 spec. It doesn’t strike me as a particular draw given that it’s at least possible, if still somewhat painful, to build truly cross-platform apps with offline capabilities.

Why should I support ChromeOS? ChromeOS is not the web!

wepple• 13 years ago

In addition to the other points made here - it's far simpler to adapt a web application you've already built to an off-line chrome app, than it is to start again in a native language.

mythz2• 13 years ago

The productivity of web-based development, with access to richer API's and local storage, on the reach and cross-platform of Chrome, with the visibility of publishing it on the Chrome Web Store, and only 1 browser to test it on = Win, Win!

trentmb• 13 years ago

What's the point?

It's basically Java, but Google made it. Your Chrome Apps can run wherever Chrome does.

dragonwriter• 13 years ago

They are slower and uglier than native Windows/Mac/Linux apps. What's the point?

They are the "native" apps for ChromeOS that, incidentally, can also run anywhere else* that Chrome OS runs.

* except iOS

bergie2• 13 years ago

With the Chrome Apps PhoneGap wrapper you can package them for iOS too: https://github.com/MobileChromeApps/mobile-chrome-apps

ams6110• 13 years ago

When you put it that way, they sound a lot like.... ActiveX

techzomby3• 13 years ago

Except that ActiveX works only on Windows IE and not built on web tech. Chrome Apps run on ChromeOS, Windows, OSX, Linux, Android and IOS. And they are built using HTML and JS.

dgregd2• 13 years ago

They are slower and uglier than native Windows/Mac/Linux apps. What's the point?

Eat your own dog food. Google developers test new HTML5 APIs.

tluyben2• 13 years ago

Uglier? In general to me that doesn't seem to be the case, at least compared to Win/Lin applications.

taude• 13 years ago

I prefer using my email in the browser over any native email client. And as an added bonus, I get nearly the same user experience across my many disparate devices.

ChuckMcM• 13 years ago

Seems kind of obvious yes? I mean if you're building an OS (ChromeOS) and you want to do native development, you build your tools to work in the OS right? Or am I missing something here? One of the coolest things about early databases was the 'web based' development environment, except for the sql injection attacks of course :-)

mariusmg• 13 years ago

There's no "native" development. It's just HTML and CSS. Don't know why we would need another "IDE" to write html but whatever....

tadfisher3• 13 years ago

The native runtime environment for apps in ChromeOS is HTML/JS/CSS. The term "native" is not always used to refer to on-the-metal environments.

spankalee2• 13 years ago

This actually isn't a cloud-based IDE, it's a Chrome App and runs completely locally and offline, editing a project on the local file system. You won't need to log into anything or store your code on anyone else's servers, though there is git integration.

mythz2• 13 years ago

Great looking forward to this, as I believe this strategy of being able to hack directly on a shared code-base from a browser-based IDE will end up being the most productive environment, especially if it's integrated and can easily consume managed PAAS infrastructure services without the hassle of maintaining or trying to scale them yourself. Having an auto-updated IDE will also be a nice change.

I haven't been excited about web-based IDE's previously as all the ones I've tried have been clunky and slow, but it looks like this is being built with Dart so should offer better start-up and runtime performance given it will be running natively on the DartVM inside Chrome App.

Looking forward to the day that I'm able to login, hassle-free to any ChromeOS PC and have instant access to my working environment. Which will be useful if you ever want to upgrade your workstation, e.g. when it breaks down or a shiny new model comes out.

RivieraKid• 13 years ago

Honestly, I don't see a point. Presumably, it originates from the belief that the the web is the future and that it will finally replace native platforms. I used to strongly believe in that too but now I don't think it will happen any time soon, if at all.

Currently there are two kinds of platforms: the web, initilly intended for documents and native platforms, designed primarily for apps. Both have different strengths. The web: no installation, hyperlinking, decentralization and openess, etc. Native plaftorms: window manager, speed, collaboration between apps, much richer API, centralization (enabling human interface guidelines for example), etc.

A third platform having the best of both worlds would be great but that's just my utopic dream :)

spankalee2• 13 years ago

You can try to get there by bringing the benefits of the web to apps, and by bringing the benefits of apps to the web. We're already seeing this with ChromeOS, FirefoxOS, Cordova, etc. on the web side, and various efforts to streamline app installation, add deep linking, and other things on the app side.

Personally, I think browsers can be a quite good application platform - they're basically a view hierarchy, with a layout and styling system, an event model, and scripting engine. With efforts like the *OS APIs, custom elements, asm.js/PNaCl/Dart, WebGL, canvas, Service Workers... there's a lot you can do, all in a stack that's easily accessible to millions of developers.

zobzu4• 13 years ago

"they’re Google’s way of pushing the limits of the Web as a platform."

Nope, they're Google's way of extinguishing non-Chrome browsers. Only Chrome/Chromium will be supported. That's called vendor lock-in - regardless of being open source.

ahoge2• 13 years ago

Firefox's addons also only work with Firefox. Thunderbird and Komodo also only work on top of the Mozilla platform. Eclipse and Netbeans needs Java. A bunch of apps need Air.

You can make this work inside XULRunner if you want. You only have to write some library which provides the same API as the chrome_gen package (reading/writing files etc) and then import it instead of chrome_gen or just overwrite the package. Everything else is regular web stuff.

Anyhow, using some apps which use Chrome as runtime doesn't meant that you have to use Chrome as default browser. You only need to have it installed. Just like any other runtime you're using.

techzomby3• 13 years ago

These apps don't run on Chrome the browser. They run on the Chrome runtime, and open outside the browser. So, if they're extinguishing non-Chrome browsers, they're extinguishing Chrome browser too - which is obviously not the case.

coldtea7• 13 years ago

Nope, they're Google's way of extinguishing non-Chrome browsers. Only Chrome/Chromium will be supported. That's called vendor lock-in - regardless of being open source.

In order for it to be "vendor lock-in", people should care about those apps in the first place. Which they don't, much. In fact, they are quite niche.

So, yes, they'll might lock a 0.1% of people with that.

Now, if they did the same for Search, Mail or YT I'd be worried.

(Hint: even after all those years, Google Docs, a flagship has an insignificant user base -- and that one is not even Chrome-locked).

reidrac• 13 years ago

"vendor lock-in" and "open source" are incompatible. Google won't stop anyone that wants to add support for other browsers. In fact, anyone could fork Spark.

If you want an example, see OpenOffice.org and LibreOffice.

wiradikusuma• 13 years ago

So this is like Adobe AIR? (FYI Adobe AIR also supports HTML in addition to Flash)

untog• 13 years ago

No. This is a development environment, not a runtime.

ahoge2• 13 years ago

Yes, Chrome Apps can make use of the Chrome Platform API which allows them to do all kinds of privileged things (if the required permissions were given).

http://developer.chrome.com/apps/api_index.html

So... yea, it's very similar. Both are some sort of web-related runtime which can be used for all kinds of applications thanks to additional APIs which make the environment less restrictive.

frodopwns• 13 years ago

Integrate with Google Drive and I'm sold.

patrickaljord3• 13 years ago

Files will use the chrome file syncing api.

TIJ2• 13 years ago

they gonna end up giving such developer access to everyone where we'd be making standalone app every other day in short more power and more apps i am super excited about it.

coldtea7• 13 years ago

Yes, just what we need, more crappy apps.

TIJ2• 13 years ago

how can you be so sure of that and even if i will come up with crappy apps maybe some big professional company comes up with something that you will end up using you never know...;)

nmb• 13 years ago

There will be at least three development platforms called Spark now, the others being http://spark.incubator.apache.org/ and https://github.com/meteor/meteor/wiki/Spark

laureny2• 13 years ago

Before Android, I would have been super excited to read this, but now...?

Seems like it's yet another technology that Android will soon be steamrolling.

daliusd4• 13 years ago

Could you give any reason for your excitement for Android? Android is popular now but that does not mean that it will be popular forever. Fort Model T and Sony Walkman are good examples of things that were popular once.

laureny2• 13 years ago

I never said I was excited about Android nor that it will be popular forever.

I was just observing that Google's strategy regarding Chrome OS makes no sense to me in the presence of Android.

yeukhon2• 13 years ago

Dart meanwhile is Google’s open-source Web programming language, which has an ultimate goal of replacing JavaScript. Polymer is Google’s library for the Web, built on top of Web Components, and “designed to leverage the evolving web platform on modern browsers.”

Go, Dart, Angular.js and now Dart and Polymer. Am I the only one seeing the next wave of confusions?

svachalek• 13 years ago

Because the web was simple and homogenous before this?

yeukhon2• 13 years ago

I will be fair and the answer is no, but at least for the last 2-3 years we have a little stable environment. If it's ruby app we know it's rail, if it's Python we know it's either Django or Flask (or Pyramid). When we think about client side we think of jQuery and pure Javascript. I know there are coffee script lovers and I admit it's great that people coming up with new tools, but we have too much to digest.

There is a certain degree of peer pressure coming from our co-workers and our webdev community - if you are programming in PHP you are doing it wrong or if you are still doing MVC and not single-page app or node.js you are behind the trend. I have little knowledge about what Dart does but from multiple sources people seem to think Dart is Google's alternative to Javascript, which makes me uncomfortable because I start to feel like I am at the second Web/browser war where large organizations like Google is pushing their own version of "Javascript" (back then we got what? javascript, active script? etc?).

bitemix• 13 years ago

Supposedly nothing is stopping angular.dart from working with polymer.dart.

gress2• 13 years ago

I look forward to the day when chrome is the only OS and all the proprietary OSs have finally died out.

72deluxe2• 13 years ago

You don't see Chrome OS as proprietary? It might be "open" but Google surely has a massive stake in it?

Surely the least proprietary mainstream OS would be Linux?

gress2• 13 years ago

I guess the irony didn't come over in my original comment.

busterc3• 13 years ago

Dandy, but Chrome Apps will only ever be marginally more widespread than ChromeOS.

Q. Why? A. Node-WebKit Apps aren't crippled with the sandboxing of the Chrome App API and easily let me target Win/Lin/OSX without requiring the Chrome runtime.

FWIW, I like ChromeOS and would buy a couple of ChromeBoxes for family members, if they had reasonably sized hard drives (200+GB)(SATA) for pictures & videos, and weren't so dependent on Google Drive. Non-tech savvy _NEED_ ChromeOS, they don't need GDrive.

How many of us need yet another IDE for HTML5, CSS & JS? Not me, thanks though. Please Google, get back to autonomous cars and get serious about ChromeOS already.

ghostdiver• 13 years ago

Chrome Apps did not work so far(zero traction), why do they want to waste ever more time on it?

ecesena• 13 years ago

Quick link: https://github.com/dart-lang/spark

nhebb2• 13 years ago

What is the practical use case here? Before this there were Microsoft HTA's and Mozilla's XUL apps, neither of which had widespread adoption among developers.

daliusd4• 13 years ago

Short:

* HTML apps potentially will work everywhere (if you will take care they will at least work on Firefox, Opera and Chrome).

* There are HTML5 tools and knowledge already in place. Plus years of experience in HTML can be used.

Long:

IMHO developers should look into the future here not into today but even today you can do quite a lot. XUL most probably failed because there were no tools, learning material and libraries (I'm not sure about HTA). In addition XUL worked only on single browser. I have managed to write XUL app 7 years ago but I had to invest quite a lot of time and the whole development process was quite slow.

HTML apps meanwhile do not require quite a lot of investment and if you already have some HTML5/CSS/JS knowledge there are only several holes of knowledge to fill. The only problem is to select good tools, find out right information and you can start right now. E.g. if you have canvas based game or static web site with useful information you can create app now: just add application cache's manifest file, add some meta information (in case of Firefox OS create *.webapp file) and you are done.

My background: I don't know HTA. I am familiar with XUL. In addition I'm familiar with QtQuick (QML + JavaScript), Firefox OS Apps, I guess I could learn developing Chrome Apps pretty quickly. I'm familiar with native development for Windows, Linux Desktop, Android and iOS.

nhebb2• 13 years ago

HTA (HTML Application - http://en.wikipedia.org/wiki/HTML_Application) was Microsoft's foray into sandboxed HTML + CSS + JS applications back in 1999.

Grue3• 13 years ago

Chrome is becoming the new IE6. Just yesterday I got notified of some new feature in Gmail... of course it worked only in Chrome. Fuck that.

dgregd2• 13 years ago

Chrome is becoming the new IE6

Web developers hated IE6 not for new features it had. It was hated for lack of bug fixes and no new releases.

IE7 was released 5 years after IE6. Chrome is updated every 6 weeks.

islon• 13 years ago

"Dart meanwhile is Google’s open-source Web programming language, which has an ultimate goal of replacing JavaScript." Good luck on that. On my side if I were to see JS replaced I would choose Lua as a replacement.

toddan• 13 years ago

why not byte code then we all get what we want.

Rafert• 13 years ago

Any relationship to the cloud IDE Brightly that seemed to have been killed somewhere in 2011-2012? The leaked 'Future of Javascript' Google memo from 2010 mentions they expected Brightly to be the first app written in Dart (which was called Dash at the time).

Google is building a Chrome app-based development environment

Related Stories