The numbers are very interesting -- thank you Bob, Donald and others who
did the work to find out what the situation really is.
I would like to repeat my comment from the plenary that there are many
different ways to crunch the numbers. You can look at participants,
weigh any particular subset of them, or look at contributors in some
other way. The bottom line is that all of us can select the statistics
that support our preferred distribution :-)
For contributor statistics, RFC publication author set
(http://www.arkko.com/tools/recrfcstats/d-contdistr.html) is probably
not the one that we should look at. It lags behind the work in the IETF
due to RFC Editor and IESG delays. It is also biased because some set of
people have learned how to use the IETF and can publish RFCs more easily
than others.
The draft publication author set is probably more representative
(http://www.arkko.com/tools/stats/d-contdistr.html), but also does not
account for people who are only observing for now but would like to
write specifications.
The attendee stats are biased according to where we held the meeting
last, though as Bob and Donald have shown we can try to eliminate this
bias. In any case, attendance statistics include tourists as well as
people who do or plan to contribute. How should we weigh tourists?
Pick your poison.
I think we should not over-analyze the selection process too much. I
support 1-1-1 because its a simple model, it feels right, and because I
believe general IETF participation is headed towards the 1-1-1 model
even if we are not there by all measures yet.
Jari
_______________________________________________
Ietf mailing list
Ietf@xxxxxxxx
https://www.ietf.org/mailman/listinfo/ietf