I love my apps!

I have an iPad and a Droid. I used to have an iPhone (before I decided I wanted my phone to also be able to make calls). I love apps! They're efficient and fast. Websites on mobile browsers can be difficult to manage. The apps can connect with internet data, but do it with a much improved user experience. No doubt. When it comes to mobile at least, a well-designed app beats a well-designed website 99% of the time. It's a new paradigm today. An interesting read is on O'Reilly, where Mac Slocum interviews Ken Yarmosh on app dominance.

But does this browsers don't matter anymore? David Card seems to think so.

Browsers don’t matter anymore….

...Subsequent competing browsers offered the promise of a similar platform: An application that with the rise of web apps and media could act as a user’s primary UI.

TODAY’S PLATFORM DELIVERY VEHICLE

But today’s platform is the web itself, as browsers and even operating systems have been rendered less important. Companies that deliver mass-market APIs for consumer apps, like Google, Facebook and Apple, don’t depend on specific browsers for distribution. Neither do enterprise suppliers like IBM, Oracle, SAP and Salesforce.com. Even Microsoft can’t depend on Internet Explorer to establish its standards or businesses. Silverlight and Bing underscore that fact. All that’s to say that the excitement about RockMelt arises from the potential of establishing a new browser, but it feels like that potential is based on an outdated model.

Setting aside for the moment that, for front-end developers like me at least, browsers do matter — especially the bad ones that require extra work just to get them to render things properly — I disagree with his assessment on three counts.

1. The issue has never been requiring specific browsers — at least not since the pre-dot-bomb days when lazy developers would test only for Internet Explorer and the rest of us could just twist in the wind.
2. Advances in the web, however, have required excluding certain browsers. Google has famously dropped support for IE6, for example. Web technology is improving rapidly, especially in the JavaScript realm — all the more so in how JavaScript benefits from HTML5.
3. There is indeed a race happening between Google Chrome, Firefox and Safari, and Internet Explorer is working to catch up. (And let's not forget the snooty Opera, either.) Why are people even using Chrome? Because it's fast, its JavaScript rendering is fast! Safari, now already on version 5, is working to make itself more extensible. Firefox 4 (currently in beta) has a new fast JavaScript engine. And IE9 seems poised to (cough) actually support web standards, including CSS 2.1! I'd say the browser wars are just starting to heat up.

But my biggest disagreement with all this is with the supposition that, somehow, the proliferation of apps today spells the long-term demise of the browser. Yes, it makes sense today to create apps, because (x)html/css/js still can't match the power of client-side Objective-C etc. It's near impossible to create the full experience of a Flipboard or Pulse with just a webapp running today's html/css/js rendered on today's browsers.

But as handheld devices get more powerful, and web browsers get better (especially for handhelds), and HTML5 (or whatever supersedes its heir-apparent status), CSS 3+ and JavaScript continue to improve, I predict a trend towards web-served code rendered in browsers to create the handheld app experience.

Why?

Because apps can get to be a pain. In talking about new apps launched this week, Robert Scoble wrote recently how difficult it can be to gain adoption:

My wife? It’s hard enough to get her to try any new iPhone app, much less one that only lets her share photos with a close group of friends. She says that’s what she uses Facebook for.

I love apps, but I am finding the pain of installing and signing up and connecting and all that ends up being a barrier for me. Do I really want to download this new app? Do I really want to go through the 15-30 minutes to set it up, sign up for the service of whatever, connect with my existing networks, assess whether this thing is even worth all the effort? I mean, as easy as it is these days, installing apps is hard!

And that will count when competitor offerings are webapps where you don't need to install anything, you just use your browser, with the UI dynamics executed client side but the data and data management processing happening server side.

In other words, apps are a stop-gap — just a way of bypassing the limitations of the browser. But when browsers get better, things will change.

Browser don't matter? Ha! Ask that again in 5 4 3 2? years.