[Spud] A few comments on Brian's proposal

["Fossati, Thomas (Nokia - GB)" <[e-mail]>]

Hi,

One thing that I like about the PLUS co-operative game, as currently
proposed, is that you can set a precise goal for it, and subverting the
game is going to be very hard because users are in control of what they
leak to the path.
Another pleasant thing, is that opt-in is not a boolean flag, but rather a
multi-dimensional variable that users can slice the way they want at any
time depending on the game they want to play (even reducing it to zero
dimensions if they want to stay out of any game).
This looks like very good design to me.

Now, the goal we are looking at is direct/indirect ways to improve users'
QoE.  (Mobile NwOs and users are aligned on this -- for different reasons
of course, yet aligned, and this is the only thing that matters.)
PLUS would give us a solid platform to build the needed signalling in a
pretty straightforward way.
All the use cases we are working on (i.e., adaptive video flows through
the mobile network, radio traffic scheduling) depend on in-band
signalling, so if we had to do that ourselves we'd probably reinvent PLUS.

Other people have use cases that are more on the side of "give me back the
ability to do proper network management" which certainly need to be
tackled.  But I would not narrow PLUS' scope down to those use cases only.
There is a great line that the minutes attribute to Mark that starts with:
"By standardizing this, we might make new things possible.", which I
wholeheartedly agree with: the innovation potential in PLUS is exactly
what we want to explore and are excited about.  And then continues with:
"That worries me, because I don't understand how this impacts end users."
Well, that doesn't worry me at all, and for a couple of good reasons:
1. PLUS' architecture gives me enough guarantees that the QoE game that I
want to enable is not going to be easily subverted;
2. The vocabulary is not yet defined, and this is where the discussion
about users' privacy would certainly happen involving the relevant
stakeholders.

If it's not yet clear: I support the creation of this working group :-)
and also plan to contribute with drafts, reviews, and running code.

Cheers, t