October 11th, 2007


video of ohio linux fest keynote

The video from my keynote speech at Ohio Linux Fest is now available on YouTube.

There's a little bit of introduction-type stuff at the beginning from Paul Ferris (the primary organizing force behind Ohio Linux Fest). My talk starts at about the 2:00 mark.

Let me know what you think -- but please no jokes about my long hair. I'll get enough grief about that from my parents.

how do you evaluate leadership?

What makes a good leader? This is a question that I am well served to answer. My job description uses that word in lots of places. Usually I answer that it's about being egoless enough to know that you're not always right, trusting the people who are smarter than you to do the right thing, smoothing the path as much as possible, and taking the hits when necessary. For me, the hardest one on that list is smoothing the path, only because I find that even with the best intentions, the best that I can do is often turning a big ditch into a small pothole, and that is more frustrating than anything else.

Here's an interesting wrinkle. There are lots of different people who observe the way I do my job, and all of them evaluate the "leadership of Fedora" using different criteria.

Some folks only care about the end results. "Show me metrics." Some folks look at the short-term changes. "What is new since the last release?" And other people look at it from a pretty long-term viewpoint. "Are you setting Fedora up for sustainability on a 3-5 year timeframe?" Incidentally, if I had to categorize the general attitude of the collective that is all the FOSS blogs, email lists, and websites, I would probably rate it as a "what have you done for me lately?" kind of environment.

I started my Fedora job about a month before Fedora Core 5 was released. Now we're about a month away from Fedora 8. So that's three release cycles. When I think back over the first two (Fedora Core 6 and Fedora 7), I think that a lot of my effort was in the "metrics" and "short-term" areas, though I suppose that you could argue the short-term changes made in Fedora 7 also register on the "setting Fedora up for the long-term" radar screen.

The Fedora 8 cycle has been a lot different for me. In the lead-up to and aftermath of Fedora 7 one of the things that we heard a lot of was that the non-Red Hat Fedora community wanted more day to day responsibility for Fedora, especially in the technical areas. So at least for me, this has manifested itself by being a little bit more passive and quiet than I would say I've been in the past. I'm not sure how "right" that has been -- but I think the that people who evaluate based on metrics, like "how many emails did you send" will probably think that I am stupid, while the people who evaluate on "how well are you setting Fedora up to scale" will probably think "hey, more voices are being heard now more than ever, and when things reach equilibrium again, Fedora will overall be better off." When I look at Fedora right now I see a lot of new leaders emerging, and invariably that leads to an increased amount of intra-project headbutting. I think that you have to adopt the long view and realize that a little bit of that every now and then is both healthy and inevitable.

Of course, the short-term evaluators still get a Fedora 8 with a bunch of new stuff in it. Maybe it's not as dramatic as the world-change that was Fedora 7, but there is a nice balance of new features along the early results of the whole Custom Spins feature. As always the credit for that belongs in the hands of the feature developers, package mantainers, infrastructure team, websites team, docs team, etc.

Once Fedora 8 is out, I do think that the Fedora Board and Fedora Advisory Board need to buckle down and decide what the Big Issues Worth Undertaking are, and more actively cut off non-critical discussions before they suck up lots of smart people's time. I think lately we spend too much time stuck in the minutiae, debating stuff that ultimately isn't critical path for Fedora.