(Message tpc:292) /home/mjd/bin/mailpager 292 Return-Path: mjd@plover.com Delivery-Date: Tue Jul 15 03:00:17 2003 Return-Path: Delivered-To: mjd-filter-none@plover.com Received: (qmail 13697 invoked by uid 119); 15 Jul 2003 03:00:16 -0000 Message-ID: <20030715030016.13696.qmail@plover.com> To: gstein@apache.org Subject: State of the Feather Organization: Plover Systems Date: Mon, 14 Jul 2003 23:00:16 -0400 From: Mark Jason Dominus Your OSCON 2003 "State of the Feather" talk was awful. I'm not the only person who thought so; I've talked to at least ten or twenty other people about it and they all thought it was awful. For example, see http://rjbs.manxome.org/debug/entry/2003-07-08 The rest of this letter is going to attempt to be more constructive, because I'm writing to you so that if you ever give a similar talk in the future, you can try to do better. I was going to tell you in person last week, but the opportunity didn't come up. I hope you take this criticism in the way it was intended, which is to say, not personally. The big fault I saw in your talk is that you didn't have the audience clearly in mind. There were two possible audiences. You could have addressed audience #1, the people in the room who were core apache developers, and who were at the conference primarily to hear about Apache. You could also have addressed audience #2, the people who were there to hear about Perl, Python, PHP, and soforth. You talk doesn't make sense for either of these audiences. You spent the whole time talking about the history of the project, the development philosophy, the purpose of the ASF, and so on. This is stuff that audience #1 already knows, and audience #2 just doesn't care about. I walked out at slide #5, when you started talking about licensing, which I think might be the most boring of all possible topics. But I've looked at the later slides on your web site and I see that it didn't get any better. Why did anyone in the room need to know about the political structure of the ASF? Why is this important? What's the significance of the web log statistics? The big thing missing from all of this discussion is who cares about this and why. Unless there's some more pressing purpose, a 'state of the union' talk is supposed to be about what's happened of significance in the past twelve months and what to expect for the next twelve months. I think this is what you should have done. There wasn't anything timely about the talk that you did give; you didn't talk about what new things had appeared in the Apache world in 2003 or what to expect in 2004. You started out the talk by saying that since there were at least a hundred different projects under the ASM umbrella, and there wasn't time to talk about all of them, you wouldn't talk about any of them. I think this was a terrible mistake. You had twenty minutes, so you could have spent it talking about 20 of the most interesting subprojects for one minute each, or 40 of the most interesting projects for half a minute each. Audience #2 would have been likely to hear something interesting--"Hey, I didn't know they were working on that. It might be useful; I'll check it out." Audience #1 would probably have been interested also. I imagine even diehard Apache fans probably wouldn't have heard of all 40 projects. As the chair of the ASF, you must have a uniquely comprehensive knowledge of the software projects that are going on. If not, then you should consider letting someone else give the talk next time. I hope you can give a better talk next year. Please take the length of this letter as a measure of my belief that you can do that. -D.