You are reading a MIX Online Opinion. In which we speak our minds. Joshua Allen Meet Joshua Arrow



Web, Interrupted

Jun 8, 2011 By Joshua Allen

The Web is an eternal débutante. Since 1995, when Marc Andreessen predicted that the Web would make Windows and Linux obsolete, people have presumed that Web standards would become the client platform of choice. Instead of just Web pages, the Web would be used for applications. For those who love the Web, the last 20 years were an emotional rollercoaster. Every few years, it looks like Web standards will finally be crowned Queen of the ball, but then something happens to dash our hopes.

Microsoft took the first steps, introducing DHTML and Active Desktop in 1997, and AJAX and HTML Applications in 1999. When the Internet bubble burst in March of 2000, momentum faltered. Many developers and investors became convinced that the future belonged to proprietary platforms. Investment flowed into alternatives like Java, Flash, and .NET.

By 2002, AJAX began to rise again, and the “Web 2.0” movement predicted the imminent victory of Web standards. But now Flash was the plugin of choice and looked increasingly hard to beat. If Web standards were going to win, it was hard to see how.

Apple as Grey Knight

Hope first came with the iPhone. When the iPhone was released in 2007, there was no SDK for developers. Steve Jobs reassured the developer community: “You don’t need an SDK. You should be developing all of your apps using Web 2.0”. Web developers rejoiced! Web standards finally had a foothold, at least on phones. But it wasn’t long before Apple introduced the App Store with a proprietary SDK, and no support for apps written in Web standards.

Apple pulled the same stunt again when the iPad was released, saying that HTML5 was more than enough to make up for the lack of Flash. Still stinging from the bait-and-switch of the iPhone, some developers pointed out the gaps in iPad’s support for HTML5 applications. Two years later, the App Store does not offer first-class support for Web client applications for iOS devices.

Google Equivocates

Google launched the Android Marketplace in October of 2008, supporting only Java apps. Once again, the Web got short shrift. Google softened the blow by announcing ChromeOS a year later. ChromeOS was to be a brand-new operating system, built on Web standards. While Apple decided on a proprietary model for iPad, Google assured us that their future tablets would be based on ChromeOS with Web standards at the core.

But things didn’t happen that way. There are now more than 100 different tablet devices for Android with no Web-based tablet in sight. On phones, Google doubled down with Java. Google inked deals with Mediatek and MStar (who make the chipsets for 75% of the world’s cheap feature phones) to base their chipsets on Android. Before the Mediatek deal was done, Android already outshipped iPhone in units, so the combined numbers are staggering. It’s ironic that Palm’s WebOS was quietly dying on the vine, killed by the flood of iPhone and Android units, while Apple and Google were shipping proprietary platforms and paying lip service to Web standards as a platform.

ChromeOS is now relegated to PCs, a sector that Google says is “irrelevant”. Not exactly a ringing endorsement for the Web platform. For the scenario of the future, it is Java all the way.

Eternal Spinster?

Will Web standards always be an also-ran? Are Web developers forever stuck in the world of Web pages, forced to switch to proprietary frameworks to build bona-fide applications?

Frameworks like Sencha and PhoneGap are working hard to fill the gaps, enabling you to build applications using HTML5 and CSS3, which you can package up to look like native Android or iOS applications on tablet or phone. These frameworks provide shims between the Web browser and proprietary platforms, so you can build apps instead of just Web pages. But third-party shims are always in a precarious position. Not only are they at the mercy of the platform owner’s app store policy, the platform owners are investing exclusively in competing alternatives. When the company that controls the operating system is placing all of their eggs in a Java or Cocoa basket and can eliminate your distribution at any time, you shouldn’t bet on your Web standards shim ever having majority.

So, what’s your take? Are major platform companies brazenly pandering to Web standards to lure people to their proprietary platforms? Or do they sincerely believe the talk about Web standards and are simply waiting until the platform is mature? Will we see a day when all major platform companies invest in Web standards as their primary application platform? And more importantly, should the Web win?

Follow the Conversation

11 comments so far. You should leave one, too.

Carl Seide said on Jun 8, 2011

Go ahead and use HTML for applications if you love kludges and want to die like HTA and WebOS.

John Allsopp said on Jun 8, 2011

A little rule of thumb I've used for, oh about 20 years now.

The web always wins ;-)

Kazi said on Jun 8, 2011

I very agree. Html supports audio and video playback in v5. It took about 18 years to implement. It should have been there in v0.1.

Joshua Allen said on Jun 9, 2011

@John - Sounds like a good rule of thumb to me!

Charles said on Jun 9, 2011

Nice post, Joshua.

Let's see how the future unfolds as the Web waits for no one...


James said on Jun 11, 2011

web always wins++

the proof is in the pudding: i downloaded an app to consume a website.

now if IE would just go away already...

karstenj said on Jun 15, 2011

I think the salient point of Joshua's piece is the questioning of Google and Apple's commitment to the web, given the traction each company has with their own proprietary platforms (iOS and Android).

Mike Steinert said on Jun 16, 2011

This statement seems to be a bit of a red herring:

"the Web [will] make Windows and Linux obsolete"

It might be more believable if rephrased:

SAS will make some (or most) native software on the desktop obsolete.

Supporting audio and video playback in a web browser (native platform software) requires device drivers. Device drivers generally require some sort of kernel, e.g. Linux or Windows. The article you linked to seems to support this idea.

Running a web browser on an embedded device can be a monumental challenge. Where performance is critical you will always see native APIs. This doesn't mean that you can't mix native APIs with web content/applications where it is appropriate, e.g. the WebView widget in Android.

There are also security concerns when you fire up a web application. How can I be certain what the application provider will do with do with my data?

I think there will continue to be places where native applications are useful. This doesn't mean that Web standards are a failure. Do you really develop for the iPhone because of standards? Perhaps to a certain degree. I suspect it also has something to do with the large number of potential customers with an iPhone in their pocket.

Aaron Weyenberg said on Jun 18, 2011

I don't necessarily see native and web as adversaries. However, this is an economic issue. Apple can fill their coffers with money a lot faster by attracting more and more developers. John is right. The web wins in the sense that it's going to persist, but it's business incentives that are a new part of the game.

Kevin Tamura said on Jun 18, 2011

Nice article Josh but I believe you're creating a bit of a false dichotomy here between native and web. There no need to choose one over the other as both can be used to compliment the other.

However I agree that the web as a platform holds much promise. It has been jilted time and time again but that is more do to the lack of maturity and the inconsistent support, or the lack there of, from browser makers.

The future holds great promise, it's a matter of if it'll live up to it's potential.

Joshua Allen said on Jun 19, 2011

@Aaron, @Kevin - Great points! I agree that the two needn't necessarily be adversaries.

But there is an enduring partisan chauvinsm that will continue to exert influence -- Web chauvinists will push the Web, and native chauvinists will push native. Much like the Montague might say, "I could never make peace with a Capulet!".

I'm a Web chauvinist, so I think there are objective reasons that the Capulets are better. But I can also easily make peace with the Montagues and even motivate myself to produce hybrid Web/native apps. Whether that's ultimately a tragic attitude remains to be seen.