Telechat Review of draft-ietf-codec-opus-

Request Review of draft-ietf-codec-opus
Requested rev. no specific revision (document currently at 16)
Type Telechat Review
Team General Area Review Team (Gen-ART) (genart)
Deadline 2012-06-05
Requested 2012-05-24
Authors Jean-Marc Valin, Koen Vos, Tim Terriberry
Draft last updated 2012-07-05
Completed reviews Genart Last Call review of -?? by Elwyn Davies
Genart Telechat review of -?? by Elwyn Davies
Assignment Reviewer Elwyn Davies
State Completed
Review review-ietf-codec-opus-genart-telechat-davies-2012-07-05
Review completed: 2012-07-05


I am the assigned Gen-ART reviewer for this draft. For background on
Gen-ART, please see the FAQ at

Please wait for direction from your document shepherd
or AD before posting a new version of the draft.

Document: draft-ietf-codec-opus-14
Reviewer: Elwyn Davies
Review Date: 3 June 2012
IETF LC End Date:
IESG Telechat date: 7 June 2012

As far as the textual part of the document is concerned this version
addressed my last call comments and apparently adequately describes the
structure of the codec.

My review at last call was concentrated on this textual part and I did
not have time to significantly dip into the large body of code that
makes up the normative part of the draft.  I cannot therefore say that
there is an exact correspondence between the description and the code,
but the description does seem to give a good view (significantly better
then before Last Call IMO) of the way the codec operates.  I would like
to thank the authors for the speedy and effective way that they
addressed the comments that I made - to an uninformed neophyte in codecs
this seems to be an excellent piece of work.

Having had a little time to dip into code since doing the last call, the
one serious comment that I would make is that it is a great pity that
the code did not incorporate Doxygen or similar code documentation
comments.  Being able to extract a Doxygen comment document would
significantly enhance the understandability of some very complex code
and potentially allow the code to be tied back to the textual comments
in a way that would (I suspect) help future implementors trying to
generate alternative versions.  I appreciate that is a lot of work at
this late stage but it would  help a lot.

Elwyn Davies