Advice from Magnus Westerlund on a successful BOF

This is from a message by Magnus Westerlund to the rtp-congestion mailing list, April 11, 2012.


I think it is excellent that people are interested in discussing
possible solutions and obstacles for a congestion control algorithm
suitable for real-time interactive media over best-effort Internet paths.

However, I think there is more important things to discuss at this
stage, namely the planning for the BOF in Vancouver. Time is actually
much shorter than people may realize. The request must be sent to the AD
no later than: 2012-06-18 (Monday) by at 17:00 PT (UTC -7). For this the
BOF procedures applies and the following rules applies:

If you go look at what information the IESG expect at their discussion
teleconf shortly after the cut-off
(  it is informations such as
proposed charter, Internet-drafts, existence of mailing list, existence
of both proponents and hopefully intended BOF chairs.

As a former IESG member a reasonable solid charter, existence of related
Internet drafts and a clear and active community (as shown by activity
on the mailing list) is the most important to get the BOF approved. I
also recommend that people read RFC 5434 "Considerations for Having a
Successful Birds-of-a-Feather (BOF) Session"

Thus I think it is important that we focus on these matters that leads
to a successful BOF.

Thus I think we need to focus on two things. Ensure that we have an
agreeable charter. My general understanding of the idea for the WG was
the following:

1. A WG with the long term goal of eventually be able to recommend a
congestion control algorithm suitable for real-time interactive media.

2. To get to that we need a WG that does the following:
 a) Discuss what requirements and consideration real-time media imposes
 b) Determine how we can evaluate and discuss different proposals in a
constructive way
 c) Allows people to present, discuss and most likely publish
experimental proposals for solution.

I believe the WG charter should be solution neutral as I am not willing
to bet on a particular horse this early.

The ADs attempt on a charter is documented here:

I don't know if it has been updated with the feedback provided. I think
everyone should review and provide feedback on the charter proposal anyway.

For a successful BOF at the Vancouver IETF meeting I think we need the
following things being discussed at the BOF:

1) The high level requirements that Interactive Real-time media imposes

2) Why existing solutions such as TFRC(-SP), AIMD, LEDBAT aren't working
well enough and need more work.

3) Some short presentation of ideas for solutions. No more than 10
minutes preferably 2 to show multiple interest and multiple solutions.

4) Discussion of the charter and how a WG would work addressing these

The BOF also needs sufficient discussion time of at least 1 hour.

This as you see would need that we have a number of presenters preparing
different presentations. We also need several people with sufficient
aligned views to actively argue in the same general direction. We also
need some one taking the responsibility for driving the process and
ensuring that the right things happens in a sufficiently timely fashion.

To make it clear, that one is not me! I have WGs to run.