Re: [codec] [IAB] WG Review: Internet Wideband Audio Codec (codec)

["Gregory M. Lebovitz" <[e-mail]>]

At 07:05 AM 1/13/2010, Russ Housley wrote:

>>>I see absolutely no good reason not to start the work and do
>>>negotiations with other SDOs on the side.
>>
>>That is thw way these joint bodies are usually formed (at least the
>>MPEG/ITU-T ones).  The group(s) form, and begin their
>>work.(independently)  In parallel the chairs and SDO management work out
>>the joint body organization.  It is easiest if the discussions on joint
>>body organization begin as soon as possible.
>
>My experience is different.  When the IETF works with another SDO, 
>the final steps in approval are extremely painful.  The IETF process 
>does not meld well with others, and the result is that getting the 
>exact same words published by both organizations is a near deadlock 
>experience.  It has been done, but only because of heroic efforts by 
>chairs and editors.  For this reason, I prefer a situation where one 
>SDO runs their own process and the result is submitted to partners 
>after the words are final.  Of course, collaboration during the 
>development of the document is most welcome, but the process rules 
>of one SDO are governing the development.

Agreed. And now we are continuing to re-trace old steps repeatedly. 
We had this discussion already before the Hiroshma BoF, and in the 
Hiroshima BoF. We concluded that likely the most positive form of 
collaboration we could have with the other SDO's on this topic was 
for the IETF work to run it's course, -- while (a) inviting the 
individuals and companies and experts from the other SDOs to 
participate and contribute openly, and (b) liaising regularly on our 
progress and inviting input -- publish it's first set of RFC's (and 
remember, RFC's are documents that may be standards track, but new 
work doesn't start as a standard, only a proposed standard, i.e. we 
can revise them over time, or kill them later, if we so choose), then 
work with other SDOs on modifications or joint promotion or whatever. 
The key here is that the IETF first needs to run its own process so 
that we know clearly what we want technically and business-wise, and 
have an example of something we think fits the description, and we 
document those in an RFC or two. Then we can more effectively engage 
with other SDOs at a textual level. This is my take away from the 
lessons learned to which Russ refers.

Hth,
Gregory




>Russ