Some thoughts in-line on the latest version of the draft...
Best wishes,
Gorry
* TFRC is within the Charter of the DCCP WG, so that is fine. My question is
to whether the authors intend this to be also proposed as a method for DCCP
(which I think is also applicable). If so, should the Introduction also say
this?
* I like the new discussion of the "meaning/behaviour of the idle period" -
but perhaps this thinking still needs to be further developed? - Perhaps
something we could start in the face-to-face WG meeting and then take on the
list?
* Page 9
- Capacity can also change quickly when the L2 network allocates capacity
based on traffic conditions (BoD). Such methods are common across a range of
wireless technologies, and can take several RTTs to adapt to changes in
traffic conditions.
-----------------=====-----------------
NiTs for consideration by authors in the next rev:
*page 4, X_active_rec
/receive reported/receive rate reported/
^^^^^
*page 7
/to adjust feedback packets' Receive Rates/
- English could be improved.
*page 8
/and the window in re-opened/and the window is re-opened/
^^^
*page 8
/And..../
- English could be improved... "In addition,"?
*page 9
/to make everything look just like lovely white noise/
- English could be improved.
*page 9
/controlled of a sending rate/controlled by the sending rate/
^^^^^^
* section 5
/and idle periods of hours/
- To me, this seems rather understated. Can we be slightly bolder?
A sender that is idle for 30 minutes seems to me to be one that is no longer
actively part of a session. This period is of the order of routing updates.
* [JSA05]
- Please update reference (unless you mean to cite the pre-WG version of the
document.
