SuitWatch

Lessons from the pressure cooker


I've been on a mission lately: To see what IT workers in the pressure-cooker conditions of political campaigns might teach IT professionals everywhere about the resourceful use of Linux, free software and open source development methods. What works best? What doesn't? How do you develop and apply "solutions" to problems all over the country with widely varying participants and circumstances? How do you embrace workers and volunteers accustomed to other platforms and methods? And how do you get stuff done when the deadlines are absolute and the costs of failure are extreme?

My plan from the beginning has been to visit (or at least talk to) the headquarters of leading candidates who were known to be using Linux and other open source platforms. Notably those included the Howard Dean and Wesley Clark campaigns, although I've also heard from other campaigns (such as Bush's and Edwards') whose techies say they're using plenty of open source stuff too.

My timeline was on the long side, since the final story would run in the June issue of Linux Journal, which has an early March deadline. By that time, the primary season would be well underway and the pressure cookers would be hissing at full volume.

Since Burlington is just a few hours' drive from New York, I decided to hit the Dean campaign first, right before the Iowa caucus primary (Monday) and Linux World Expo (Wednesday-Friday). The timing seemed propitious as well as convenient. Also, the Dean campaign was (and still is), by all accounts, successful at pioneering extremely effective use of the Internet -- for organization, fundraising, and much more.

As a reporter I have to be nonpartisan. Still, it's hard not to like a campaign that liked you before you ever heard of them. That's what happened with the Dean campaign, which is run by Joe Trippi, who used to work for Debian co-founder Ian Murdock <http://ianmurdock.com/>. Joe urged the campaign to follow the lead of its own growing constituency, and to practice what he calls "open source politics". Many members of the team are also fans of The Cluetrain Manifesto <http://cluetrain.com>. In fact, one of my Cluetrain co-authors, David Weinberger <http://www.hyperorg.com/blogger/> is the campaign's Senior Internet Advisor. While I was there last weekend, volunteer staffer Tom Limoncelli asked me to sign his copy of Cluetrain. Some of you may know Tom as an author of books on network and system administration. He's helping design the system plumbing required to handle a Niagara of incoming e-mail.

Through the weekend, I heard Cluetrain quoted a number of times. There was my "markets are conversations" line, and David Weinberger "hyperlinks subvert hierarchy." But the one that made the most sense of the campaign was Chris Locke's "networked markets get smarter faster than most companies". Exactly that principle, I heard, applied in electoral politics today. That's why the campaign was building or applying technologies that embraced their own networked markets.

Sitting in the server room, IT Manager Harish Rao waved at racks of machines and said, "Very little in here is not open-source-based. The accounting system runs on Microsoft, for example. And we have a NAS -- network attached storage box -- that's proprietary, but also Linux based. The back end is completely accessible. We could have built our own box and saved a little bit of money, but I'd rather have my people working on what the box is supposed to do, which is important backups."

Rao's team is adrenalized to make choices that are pragmatic, resourceful and either inexpensive or free. "We recognize two things in our development here," he says. One is that pretty much everything we do has been done before, by somebody, somewhere. The other is that we're here because of our supporters." And those supporters help. While the organization is busy developing a "voter file" system to send out to the field offices, it's also embracing the systems already in use by some of those offices.

The organization also takes advantage of outside commercial products where they make sense as well. Some are open source, such as BestPractical's RT request tracker <http://www.bestpractical.com/>. Some are not. While I was there, Userland founder Dave Winer drove up from Boston and installed Channel Dean, "an RSS feed that's like a clipping service for people interested in US politics". It's based on Userland's Manila <http://manila.userland.com/>, a commercial product that runs on OS X and Windows.

"We're very pragmatic," Harish Rao said. "We do what works. You need a balance. I think for most back-end services and infrastructure, open source is the way to go. For the desktop, we believe in choice. Use what you want. Most of the services you're accessing are off the Net anyway. If you're used to Outlook, hey, it does some good things. So we take a practical approach. But we're members of the open source community, and that's our context."

So is the pressure-cooker environment. "Everything around here is a scaling issue," Harish said. "Even going to the bathroom. Everything here has to do with scaling. We have to do a really good job of doing a lot with a little, in a quarter of the time. We've been able to do that."

How?

"I believe in three principles," he said. "First I always make sure I hire people I can trust 100%. Second, I always try to hire people who are smarter than I am. Third, I give them the independence to do as they see fit as long as they communicate about it to their other team members. We've had a lot of growing pains, a lot of issues; but we've been able to deal with them because we have a high level of trust, skill, and communication."

Of course, a political campaign is a special case. Ir self-selecta people who are at the high end of the motivational bell curve. No pointy haired bosses or lazy Wally characters from the world of Dilbert. Can you expect the same inside less extreme organizations?

I think so. Most enterprises have at least some high-pressure jobs that require high levels of motivation. The extreme examples forge our ideals. Seems to me that the Dean For America experience applies.