May 30, 2009

Around the first quarter of 2008 I got interested in comet-enabled applications because of the richness and immediacy they deliver to the increasingly real-time web. The list of use-cases I could envision leveraging comet for in the world of ECM / BPM were endless - that just made the research all the more worthwhile.

Back in March, one of the languages and frameworks that made comet enablement really simple was Scala and Lift. Back then, they were on version 0.7 and I had emacs going and getting the hang of the Lift. Jorge Ortiz, a Scala-Lift expert, had an excellent sample application called LinkShare on the Scala blog which I extended with some of my own customizations.

LinkShare leverages Lift’s CometActor trait which is an extension of Scala’s Actor trait to perform the magic of having your links show up in near real-time on other browsers and vice-versa. This example was a powerful glimpse into the web browser’s role for content-centric collaboration. While the ideas kept exploding in my head, there was just no bandwidth for me or my team to possibly make a credible attempt at building out a collaborative platform using Scala-Lift at the core.

After the Google Wave Early Developer Preview video I saw today. I am glad we didn’t. If this doesn’t keep the SharePoint team awake at night, along with the IE9 team - then nothing will. Predictably, it was a buggy demonstration - but that comes with the jagged edge that they are playing on.

Even with two years worth of development and the relative simplification HTML5 brings to ‘cometifying’ your application via Web Sockets and the years worth of AJAX experience the team had with Google Maps AND the enormous cache of operational knowledge they have with XMPP (GTalk) - they still haven’t cracked it fully.

What they have done though is shown how powerful the browser can be and how much room for improvement exists in the content collaboration space. I also love the fact that they have built a context-aware spellchecker leveraging their massive cache of indexed web content. Spelly is a living example of “great content creates great systems”. Yeah ok I plugged my book ;).

The other area of encouragement is that this is a 3/4 Google product - which is usually a success indicator @ Google. Google generally executes on 2/4 aspects and loses the plot for stuff like Knol. The four aspects being APIs / Technology / Product / Execution. Google Search is APIs + Technology + Execution. You can’t say product because it really does one thing exceedingly well and doesn’t have the bells and whistles of a Cuil or a Bing. And that’s fine with me.

Wave delivers on APIs / Technology / Product - but the verdict is still out on execution. For this to “win”, it needs to be able to get deployed on private clouds and it needs to get buy in within government agencies at the Federal / State / Local level. If Google is going to expect all Wave shared objects/content to be deployed within their public cloud infrastructure - the impact it can have will be dampened due to data privacy and control issues from enterprises.

It’s nice that Tom and Alice can add photos to the same album simultaneously and see their changes occur in near real-time. But it is nicer to be able to jointly draft a contract scope document, fire off an intra-tweet to my Laconica instance and kick off the next step in my highly optimized collaboration WaveFlow. I want to hold business conversations with Wave and that means I want private or atleast bordered infrastructure.

And yes - I know a ton of other products do live collaborative editing and authoring of documents - that’s NOT what makes Google Wave unique! It’s the seamless interplay of a variety of components, the excellent potential of the extension, embedded and robot APIs that make it compelling.

I look forward to Wave maturing into a beta release later in the year - they are onto something, here’s to hoping they execute the right strategy!

Previous post Zachary Quinto a.k.a. Young Spock Spotted The ImageWork team was having dinner @ Acqua last night and we spotted Zachary Quinto - who plays Spock in the new Star Trek film at the place. When Next post OECD - Open Data Done Wrong The Organization for Economic Co-Operation and Development, a Paris-based group of 30 countries with market economies released its annual labor