On 25 mrt 2008, at 22:39, David Harrington wrote: > I think asking attendees during registration which sessions they > intend to attend and building a conflict matrix would be the simplest > approach. Of course, attendee conflicts matter less than ADs, chairs, > and presenter conflicts. Actually we pretty much have that today. You can get an auto-updating calendar from the tools page, with only the sessions in it that you select. So analysis of the sessions selected by people who use this tool could be illuminating. A more formal mechanism like this would require more work both to build and to use, while the online blue sheet mechanism I have in mind will be extremely simple (no need to prepopulate it with sessions etc) and actually be somewhat more efficient to use than the existing blue sheets (which, I assume, will continue to exist). On 25 mrt 2008, at 22:21, Steve Silverman wrote: > The Blue Sheets only tell you where someone was rather than where they > wanted to be. I suggest having every registrant, indicate some > number (5?) > of "Primary" WGs and a similar number of "secondary" WGs. It should > be > possible to derive a set of WG "conflicts-to-avoid" from that info. > This > would not be perfect but it would be a reasonable and automated > starting > place. Whether it would be better than the current system is TBD. I think these efforts could be complimentary. > I think > there are just too many WGs and too few slots. But nobody seems to > want > shorter slots, longer meetings, or fewer WGs. The number of wgs isn't all that relevant except to the ADs, because nobody goes to uninteresting wgs. In a way, the meeting could be made slightly longer: on one occassion last year, the RRG met for the entire friday. This was very useful and would have reduced my overlap a good deal this time, but the chairs couldn't get all day friday this time, I assume because the rooms were no longer available friday afternoon. (I'm not advocating making friday afternoon part of the regular schedule, though.) One thing that could help is split the morning sessions in two so more granular scheduling can take place. I'm also unsure why we need a 50 minute break before the plenaries. _______________________________________________ IETF mailing list IETF@xxxxxxxx https://www.ietf.org/mailman/listinfo/ietf