On 1/25/21 5:08 PM, Salz, Rich wrote:
One idea about addressing this is that we change the AD job to not be so full-time. Or add more AD's.
IESG is not embarrassingly parallel - adding more processing elements (ADs) creates more overhead for everyone. For that and other reasons, neither IESG nor IETF as a whole scales very well.
At a time when there was much more participation in IETF than there is now, I looked at what it would take to make IETF scale better, and didn't find any good answers. If we could reliably separate the issues that require broad consensus from those which don't, that might help. If we could reliably distinguish between issues that IETF needs to take on from those that don't, without making it really easy for competing organizations to circumvent IETF on the important issues, that might help.
But these are some of the problems for which the obvious solutions have a lot of potential to create worse problems. (The cause of problems is solutions, after all.) That doesn't mean don't look for solutions. It does mean don't presume that a "new" approach is better simply because it's new or different, and the problems therefore aren't as evident.
Keith
