Stickis and BlogRovR: the conversation is everywhere

Tonight’s VLAB event, fashionably web 2.0-focused and just as fashionably named “Is Ajax where the action is?”, was a blast. Om Malik did a fantastic job of moderating the panel: Scott Dietzen of Zimbra, Larry Augustin of VA Software, Sam Ramji of Microsoft, and Pradeep Tagare of Intel capital, all with lots of witty, nuanced insights to share, both on Ajax as a technology and on business prospects in the web 2.0 enterprise and consumer ecosystems. I expect a fair amount of sound bites to pop up on Om’s blog and elsewhere.Here’s a question I haven’t heard addressed, though: this might come across as heresy in our Ajax-crazed microcosm, and I am as guilty as the next web 2.0 startup CTO of worshipping at the XHR temple, but here it comes: when are we going to end the madness, end the desktop taboo, and let web 2.0 apps break out of the browser sand box? When are we going to reclaim the desktop? Or are we just going to let Google push the existing desktop vendors into irrelevance by interleaving its all-encompassing applications between the OS and the browser?

Look, there’s room for more on that quick launch bar than just Firefox. I know it made sense at some point to move away from native, client-server apps, and deliver instead all our applications and services in the browser; but isn’t it time we revisited the assumptions and the tradeoffs?

While I have heard and have agreed historically with the arguments touting the no-download, no-installation, no-administration beauty of browser-based applications, let’s take a closer look at a few things:

• Given the number of Ajax runtime toolkits and development environment plug-ins being worked on and nearing critical mass, and given the number of browsers developers have to worry about, I am not convinced developing for a handful of (should I just go ahead and say two?) desktop environments is less cross-platform than developing for “the” Ajax-stack browser.
• As Om pointed out tonight, Ajax isn’t suited for dial up consumption. The XML or JSON data that goes over the wire may indeed be better suited for a slow connection than a traditional HTML site, yet since an Ajax app is a not only data but also a full-blown app in JavaScript, it does have to get downloaded. Except it happens (modularly on-demand, of course, but what’s to stop us from doing the same for a desktop app?) every time you use it. That’s a lot of bytes pushing around the JavaScript source code of your widgets.
• While we’re less and less frequently offline, there’s still something to be said about storing and manipulating data locally. And now we’re starting to try and figure out how Ajax apps can access local drives.
• Aren’t we proudly re-inventing a lot of the wheels, err, widgets that have been around as native components on our desktops for a while? Could there be some measure of “new frontier” excitement fueling this, the fun of building some very cool stuff all other again? Better, sure, but all over again. I liked the mash-ups in Zimbra, but how are they all that different from the Microsoft smart tags that have been around for years? Is this about Ajax or about building a developer community to contribute mash-ups?
• Our operating systems (Windows and Mac alike), our anti-virus software, our browser extensions, heck, even our browsers keep updating themselves. Why not have all our apps live-update themselves and be done with it? It’s not as if there were no down-time on our broadband connections or no space on our ever larger hard drives.
• Just because a lot of people first discovered the Internet via a web browser, do we need to perpetuate the myth that only a browser can consume TCP/IP, HTTP, XML, or JSON? Do we need to keep equating the transport with the container? Of course not: lots of examples are already in the wild, from desktop widgets to the Google desk-bar, etc. In many of these, an HTML rendering engine is used, so I am not even saying we have to give HTML up. But we can break the no-download taboo, we can live outside of the browser sandbox.

Is this all due to the fear of the download? I see the no-administration argument fading because of self-updating, self-healing applications. Sure, there’s security; the browser is viewed as a safe place to hang out: as long as you’re just looking, you don’t get hurt. That’s not entirely true, with all the popups, the malware that gets pushed our way from the web, leveraging browser vulnerabilities or users’ inattention. And the distinction, again, is blurring. Web 2.0 apps want access to your drive. They are pushing the limits of the sandbox. But in my view, falling back to the browser as a container is more a reflection of current holes in how security and access are currently managed (poorly) in the OS rather than how superior a paradigm the browser represents.

I don’t think web 2.0, or the next web as I like to call it, is all about Ajax. It’s about the richer conversations, the richer exchanges, between humans and between services, that we strive to enable and foster. That means decentralized services and content, that means mash-ups. But that doesn’t mean it has to happen in a browser only. What do you think?

Trackback URI | Comments RSS