I find two of Microsoft’s current ad campaigns interesting. One asserts that computer technology is all about connecting people, particularly synchronously (as opposed to asynchronous stuff like email, file sharing, and wikis). If you replaced the Microsoft logo at the end with Qwaq’s, I think it would fit my company perfectly. They get it.
But now they’re running another series of adds that dismisses search engines in favor of what they call a decision engine. I don’t want Microsoft to make decisions for me, but I sure do want information much closer to real time. On Friday we saw first one green Chinook helicopter go by our office windows, and then another, and then I think a Huey. What’s going on, I asked of the office in general, as they shouldn’t be training on such a windy day over a populated area. So Keith searched. He figured Google was too old-news, so he immediately went to Twitter. Someone had posted that there were helicopters going past their office windows to the nearby San Carlos field for tomorrow’s helicopter air show. This week a fellow on Colbert interviewed the editor of the New York Times. “Here’s today’s paper,” he said to the editor. “Show me one thing that happened today.”
I figure Wide-Area-Networked computer systems have only been around for a little more than ten years. Most of the realtime applications have been dedicated, structured, proprietary systems. But for people to truly connect, to truly work together, they need to be able to pull arbitrary things together in real time — things that the designer of the system did not specifically envision and provide for. Real time arbitrary search(*) is one example, but the general theme is realtime, unstructured, multi-person, multi-media, multi-application collaboration. It’s going to be huge.
(*) When Web search started, realtime search referred to getting answers to a query in realtime. It wasn’t about the age of the underlying information. Now realtime results are the expected norm, and we can safely use the phrase “realtime search” to mean that the information is live.
I don’t know what to call this general application collaboration: multidimensional, multi-facetted, unlimited, live, organic, unconstrained, …
While on the topic of real time and collaboration, what’s your impression of Google Wave?
Thanks.
That’s a great question. I don’t know yet. The intensions and the movie results so far are terrific. But I haven’t looked at the math or the code. What I’m not clear on is what the magic is that would allow multiple people and multiple applications at scale.
For example, how does it play out when you’ve got one or two dozen people on a document, and a couple of them have bad latency? The math of TeaTime (which we use, http://wetmachine.com/i…) is such that no one suffers waiting for the stragglers, and that even the slowest user’s results are not wrong (e.g., non-sensical), just maybe painful for that user to make changes with when there are others also editing. Wave is welcome to use the same magic, and they might, for all I know. Or they might have some other set of tradeoffs. Or maybe they only intend to go after a different problem, such as text editing with a small number of users on well-matched networks. There’s also a set of issues related to bandwidth limitations, for which we have a different piece of magic.
From what I understand, their architecture and protocol are client/server based. I would love to see them abandon their centralized view and adhere more to a peer-to-peer architecture for greater scalability.
While watching the keynote I couldn’t help to think about Engelbart’s demo and of course, Croquet. While I’m not against image based solutions, I keep thinking how much more ahead we would be if solutions like Croquet/Qwak were simply available via a URL and a browser. Which brings me to the point, I think perception of Google Wave wouldn’t be as sensational if people were more exposed to currently available solutions. Is there a technical reason for not providing a “free” limited version (not a trial) of Qwak?
Qwaq — the spelling uses the corner letters of the keyboard — has a free or low-cost Personal Edition in which the number of forums & users and such is limited. Users get the the current product options at the end of their trial period.
However, a bigger barrier might be security. For our customers, we’ve set things up so that you absolutely have to be explicitly invited to an organization in order to see the forums for that organization. There’s no concept of publishing things as world-readable. I think we have a pretty clear concept of unconstrained, organic collaboration — EXCEPT that it’s limited to an organization. I’m not quite sure how to address that while still meeting the requirements of our paying users. But I do think we have to figure that out.