I relocated from Los Angeles to Connecticut a few months ago, and a few of my geekier friends joked that I had to meet Joel Spolsky and Paul Graham before I came back to California.
Joel is in the midst of his 21-city FogBugz World Tour and one of his first stops was in New York City, where I saw him demo FogBugz 6.0 two weeks ago. In fact, in the picture at the top of Joel’s post about the session, you can barely see my head peeping out over the guy with the black shirt and white stripes on the left side. Those stinking paparazzi never leave me alone.
The demo went well; it wasn’t spectacular, but it was a good 40-minute overview of FogBugz’s main components: a wiki, forums, bug tracking, and scheduling. But it didn’t need a big flashy presentation – the application itself is seriously impressive.
I don’t know when FogBugz became a complete project tracking and support tool for small software shops, but it’s no longer just a bug tracker. The wiki looks really sharp – the WYSIWYG editor is completely custom, and the AJAX everywhere makes for a fantastic user experience. Joel said they spent between 2 and 3 person-years developing their AJAX library from scratch since they encountered too many issues with third party libraries.
I’ve been using FogBugz on Demand (their hosted solution) for the past several months and I hadn’t even noticed there’s a wiki (intended for writing specs and documentation), forums, group email (perfect for a support team serving external customers), and the awesome evidence-based scheduling piece.
In evidence-based scheduling a completion date is constructed from developer estimates, but the completion date is actually a range of dates with probabilities that each date will be met. So you can look at the calendar and say “We have a 50% chance of finishing by October 31st, and a 90% chance of finishing by November 23rd.” In addition, the schedule is based on the developers’ past estimating accuracy, so the better their estimates (compared to their actual time spent), the tighter the completion timeframe. Bad estimators means a wider timeframe.
It’s the right way to handle the task of determining a completion date, and now that they’ve implemented it I’m amazed no one has done it before. Some form of evidence-based scheduling is going to find its way into every project management tool over the next few years. If you’re not happy with your bug tracking or project management application you should take another look at FogBugz.
The capstone of the morning was during the casual Q&A session after the official meeting. Most people left quickly and I found myself in a room filled with about 10 other attendees and the entire staff of Fog Creek Software.
After talking to a couple Fog Creek developers I wound up having a 15-minute conversation with Joel about consulting vs. products, leverage, and SaaS. He’s as smart in person as he sounds in his blog, and it’s obvious he’s given a lot of thought to the issues we discussed. Having read Joel’s blog since 2001 it was cool to finally meet him face to face.