I've been thinking a lot this week about how to talk about technical topics, for a few reasons. First, the PyCon 2011 call for proposals is out. Second, as the organizer of the Boston Python meetup, I'm frequently talking to members about possibly presenting, and this week I was considering an especially complex proposal, and trying to wrestle it into a digestible form.

Lots of people, when presented with the possibility of doing a technical talk, will decide that they aren't skilled enough at speaking. It is scary doing your first presentation, but you should know that everyone is bad at it before they try it. I was a nervous wreck before and after my first conference talk, but now I look forward to them. You aren't as different from the people you see up there now as you think.

Another big self-barrier is "I don't have a good topic." This is also not an insurmountable hurdle. If you don't have a topic now, you can probably find one easily. Here are some ideas for finding a topic:

• Jack Diederich suggested the heuristic of looking back a year and asking yourself what you had wished you knew back then. Tell others about it.
• Alternately, think about what it is you hope to learn in the coming year. One way to learn it well is to suggest it as a PyCon talk. Giving talks don't have to be just teaching experiences, they can also be learning experiences.
• Have you built something you think is cool? Tell the world about it.
• Even if you haven't built it, is there something you use that you think deserves broader exposure? Explain why.

If you are going to talk about something you (or something else) has built, you have to figure out how to make it applicable to other people:

• If your project is useful to other people, your topic could just be the project itself: "How to Use MyProject to Fooble your Baz Better", or whatever.
• But you may have a project with a small audience, or maybe it's only interesting to you. Was there a special challenge in building it? That can be your topic, so long as you can generalize the problem so that others can appreciate it. If you've built a connector so that Django can query your 40-year-old legacy database, you don't want your topic to be "How to Read AncientDb Records," you want it to be "How to Build a New Backend for Django."

This last point is the most important: put yourself in other Pythonistas' shoes, and ask yourself why they will care about your topic. Don't get me wrong, I'm not trying to dissuade you. If anything, I hope you'll do this and find the story you can tell that will interest more people than you thought.

Propose a PyCon (or Python meetup!) talk, you'll be glad you did.