« April 10, 2003 | Main | April 14, 2003 »
April 11, 2003
Grand Central Connects Web Services
Imagine a world with no Postal Service nor Federal Express. Whenever you wanted to send a document to someone, you'd have to find out how they accepted documents, find a trustworthy carrier to transport the document, and then negotiate delivery mechanisms, notification requirements, and security measures. This is the current situation of Web services. Sure, transport is available and there are some basic standards about envelopes and how to look up an address, but there's not much else. In fact, unanswered questions about security, reliability, trust, and coordination abound. [Full story at InfoWorld...]
One of the best parts of doing this review was my interactions with
Grand Central CTO John McDowall. Its clear that John's primary objectives are guaranteed message delivery and 100% uptime. In addition to these primary objectives, his engineering team works hard to ensure that the network delivers some other important features:
- Security. The network provides authentication and authorization services on three levels: HTTP with password authentication, HTTPS with password authentication, and HTTPS with a host-verified SSL certificate from Verisign.
- Logging. The network logs details about every transaction that goes through it. Each transaction is assigned a unique tracking ID. These logs can be reviewed online, accessed as a web service, or exported for use elsewhere. This is useful for debugging but, more importantly, provides for monitoring and auditing. Grand Central maintains these logs for a year with a standard contract.
- Interoperability. The network supports SOAP via HTTP, of course, but also allows connections to and from partners who use FTP or even SMTP.
The Grand Central engineering team follows some important principals in building the network:
- Multi-dimensional scaling. The network is a distributed collection of servers that can be scaled for performance and reliability in small steps.
- Clean service oriented architecture. Grand Central creates the features in the network as Web Services and treats things like alerts and events as first class citizens---that is they're just SOAP messages that can be fed into the network and treated like any other SOAP message. This creates enormous power through extensibility and service re-use.
- User self-service. The network is designed so that all of the features can be accessed and configured by users using a web interface without interaction with the humans at Grand Central. Future extensions will expose these interfaces as web services so users can control their network programmatically.
In speaking with John, the answer to many questions is "when customers demand it, we'll deliver it." This kind of customer driven product management has served a number of companies well. Success in this strategy depends on having a solidly built foundation on which to hang features and Grand Central has the basics done.
05:15 PM | Recommend This | Print This
Singularity Hints
So, you'll probably accuse me of living in a cave somewhere, but I was talking with Scott Lemon this morning at a Utah Venture breakfast and he mentioned "the singularity." The look on my face must have told him that I had no idea what he was talking about, so he explained.
The Singularity is the point where superhuman intelligence spontaneously emerges from our various technological activities. The idea was put forward by Vernor Vinge and taken up by Ray Kurzweil. I had an opportunity to hear Kurzweil speak a year or so ago and I have to say it was fascinating. This is how Kurzweil describes it:
The Singularity" is a phrase borrowed from the astrophysics of black holes. The phrase has varied meanings; as used by Vernor Vinge and Raymond Kurzweil, it refers to the idea that accelerating technology will lead to superhuman machine intelligence that will soon exceed human intelligence, probably by the year 2030. The results on the other side of the "event horizon," they say, are unpredictable. We'll try anyway.
Now, I've got a really big practical bone in me, so I tend not to worry about this kind of thing too much. Even so, I see things that make me think that if we're not headed for this kind of singularity event, we're at least headed for an era of near ubiquitous data .
- Already, I believe that people who write and read blogs, use IM, and even email are at a decided advantage over those who don't in terms of not only what they know, but how they think about the world. Maybe this is self congratulatory or self delusional, but I believe it nevertheless.
- Things like OnStar and Grid Sensors are already creating a world where information is being collected in ways we never thought possible and Moore's Law will just make them cheaper and hence more ubiquitous.
- Improvements in business intelligence tools and concepts like virtual networks of demand will provide even more of the right data to people at just the right time. In fact, I noticed just this morning that Sam Ruby is blogging about aggregators that are much more intelligent.
Will all these trend add up to a singularity? I don't know. I can't even bring myself to speculate. I do know, however, that they have significantly changed how I work, what I think, how I interact with people, and even the people I know in a very short time. When I talk to people who haven't made the jump, I feel like I did in 1993 when I was trying to explain the WWW to people.
When I gave a guest lecture to the Rollins eBusiness Center, I created a list of keywords that I think are important indicators of technology trends in the near term:
- Personalized
- Peer-based
- Decentralized
- Collaborative
- Connected
- Converged
- Presence-enabled
I'll probably expand on this next week when I talk to the BYU CS department. Scott is one of the guys I talk to regularly who gets it. Dave Fletcher is another. Most of the time I can hardly wait to wake up in the morning to see what cool new thing the day will bring. Its rare that I'm disappointed.



