We need a web browser designed for the cloud
January 8, 2010
Over the Christmas holidays I’ve been thinking about the web 2.0, social web, the cloud, or whatever you might call it, and what’s bothering me about it right now. Not being in control of your data is one issue, but I’m not going to talk about it here. Mostly because it’s orthogonal to my other big beef I have with “the cloud”: usability. Let me explain.
Right now webapps are many mostly disconnected apps that are just loosely connected by a browser. Now I know you can have some inter-app connectivity nowadays, mostly thanks to feeds (e.g. Twitter & Facebook). But it’s not really what you get on the desktop, is it. There’s no drag’n’drop from Flickr to Gmail like there is from iPhoto to Thunderbird, for instance. I’m not suggesting that we implement drag’n’drop for the web, but I do wish there was a way to connect apps better, in a way that makes sense for the cloud. Right now we’re copy-and-pasting URLs or downloading stuff first just to upload it again. That feels a bit 1990s to me.
There’s another thing about cloud apps, and this one actually feels very 1980s to me, back when DOS was ruling the PC world: they all have their own UI. Indeed WordPress and Google Docs are about as similar to work with as Lotus 1-2-3 and the Norton Commander. You might now be thinking that surely there must be room for branding and that different apps demand different interfaces. Yet on the desktop there is surprisingly strong demand for coherence. Rightly so I say. Consistency matters. It boosts productivity because all people naturally form habits. Good UIs exploit this human trait.
I realise it’s going to be a long way before the makers of web apps can agree on common UI paradigms, if at all. It’s much more likely going to happen if you bring the cloud to a specific platform. The iPhone is a nice example: the Facebook and WordPress apps follow the platform’s UI paradigms just as much as the Email client the OS ships with. Of course, that platform on the desktop is the web browser. So while I think web apps themselves can still improve a lot in terms of usability, both the goal of integrating different web apps as well as pure practicality suggest that the browser is a good place to start improving the “cloud experience.”
This can and probably should start with small things, like authentication and identity management. As a technical basis, OpenID is great, yet it’s not available everywhere and when it’s available, it’s not as user friendly as it could be. Why should I have to sign in anyway? iPhoto works straight away, too.
Others have realised this as well. There’s Flock, the “social web browser.” It makes a nice effort of somewhat integrating web apps like Gmail, Twitter, Facebook, and Flickr. But its closed-source nature has prevented it from being adopted by the billions and getting much community support in the form of extensions, like Firefox upon which it is built. Mozilla themselves, however, are doing very cool stuff that might bring Firefox up to speed:
- Ubiquity makes much copy-and-pasting unnecessary and for services like searching, translating, URL shortening, etc. you don’t even need to open a new tab and go to the respective web page anymore, Ubiquity lets you do it on-the-fly. This is the right thinking, but it needs to be built into the browser much, much more. Even more than what Taskfox, where the URL bar was replaced with a ubiquity-like command bar.
- Weave is many things, but what I like most about it is the identity management part that integrates OpenID with the browser and provides a quasi single-sign-on for all other sites. This is exactly the kind of seamless cloud app interaction we need!
- JetPack isn’t so much end-user technology as a developer platform for building Firefox extensions as easily as Greasemonkey scripts. That means it’ll be much easier to integrate web apps with the desktop app that is the browser, including for instance building more coherent UIs around existing web apps.
All this sounds very exciting. In fact, I think an awesome platform for the cloudy web would be a cocktail of Flock, Taskfox, Ubiquity, Weave and JetPack. Unfortunately that all sounds a bit far fetched for Firefox 4 and that’s not even due till next year. Plus, things like the Home tab challenge give you an idea of what they’re actually planning for Firefox 4. A “Home tab”? Seriously? And then there’s Raindrop which aims at integrating various communication channels, e.g. email, twitter, etc. The prototype is impressive. Sadly however this takes the whole integration effort on a tangent that I don’t consider fruitful. First of all it’s restricted to messaging, and secondly it’s designed much more like a web app than what the browser really is supposed to do: provide desktop integration.
So the question that intrigues me is why not take what awesome components Mozilla have produced and turn them into a cloud browser. Why not go mix that cocktail?
On a last note, I think it would be interesting to see how Google is tackling this problem at all. After they have built a brand new browser from scratch, but apart from impressive speed and stability it’s still a very conventional browser. It’s hard to tell what they’re cooking up next, but so far neither Chrome OS nor Google “yet-another-cloud-app” Wave have blown my socks off.