Dave,
Unfortunately, the ultimate and practical meaning, of these kinds of conclusions about venue selection, is that we do not place productivity as a high priority. We have a collection of other priorities that take precedence, for a collection of reasons. This means that the impact of face-to-face meetings, on productivity and quality, is almost entirely a matter of luck.
I think some of the goals beyond immediate productivity are also important, such as financing for our activities or boosting future productivity. Anyway, I agree with you that face-to-face meeting results are somewhat random in terms of people turning up or the most needed people being there. We have, however, experimented with various meeting details (e.g. Paris timings) and adapted our behaviour based on what we learned. We are tracking what number of participants we are getting in, and presumably optimizing a little bit towards larger crowds. We are changing our locations to distribute access problems and travel costs a bit better. But you seem to talking as if you knew where the IETF needs to meet in order to optimize productivity. Please tell us where :-) Is our key contributor (draft author) distribution the same as the one in participating in meetings, or different? We are choosing our meeting sites according to the meeting participant distribution, I think. What change has there been in the key contributor participation across meetings in different continents or meetings in different times of the year? Have we experimented by holding meetings for a long period of time at the same location and determined what effect it had on the number of contributors? --Jari p.s. I'm working on a little script to extract author information such as location data from drafts. If I get it accurate enough I can post some statistics about contributors. Or maybe someone already has this data. _______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf