Yup, if we have to schedule meetings on Friday, starting earlier would be a happy thing for most everyone, I think. By that time those of us with jet lag are probably over it, and Friday is always a weird day at IETF anyway, since only the die-hards are still there, and everybody is tired.
On Mon, Jul 11, 2016 at 12:41 PM, Phillip Hallam-Baker <phill@xxxxxxxxxxxxxxx> wrote:
I think we have rather bigger problems than deciding the start time. But the way we have slid into picking the start time is symptomatic of a broader problem.I find a large number of WG sessions to be very unproductive because they consist of status updates that could be sent in an email or discovered by looking at the tracker. What I am interested in is working through the set of action items that need discussion.Yes, cross area work is nice. But I don't think our current format achieves that and I am not convinced it is as desirable as people think. The purpose of a layers abstraction model is so that people working at one particular level only need to consider their own layer and the interfaces to the layers above and below them. If something in applications depends on something in routing, well something is wrong. And something is really wrong if the interfaces between the layers has to shift more than once a decade or so.One major benefit of the new approach is that I can now schedule meetings in the mornings before everything else starts. That is especially important if you have a 9am Monday meeting and someone isn't going to be at the location in fit state till the Monday of the meeting.Contrawise, I don't see any good reason to delay the start on a Friday. In fact I would prefer to start an hour earlier.So my incremental improvement schedule would be:Mon: 10amTues: 9/10 (slight preference for 9)Wed: 9/10 (slight preference for 9)Thurs: 9/10 (slight preference for 9)Fri: 8 am.